Re: [Discuss] Mastodon?

[Mike Small]

Eric Chadbourne  writes:

> Anybody else playing around on Mastodon? Come say hi. 
> https://securitymastod.one/@sillystring See if this thing works.
>
> It feels a bit like twitter but I know even less people.

I got on sdf's instance for a bit and read some "toots." Yes, it does
seem a lot like twitter, in the sense for me that it was mostly nonsense
and on the whole of no value.  Yet for awhile, despite that, I felt an
odd compulsion to keep up with the postings.  This would keep me doing
basically nothing for up to an hour at a time and feeling slightly sick
about it afterwards.

I first checked it out after sensing the enthusiasm from the hosts of
Libre Lounge for their own technical work on activityPub and for the
idea that this would be different by being de-centralized (and involving
more free software?). But I wondered if mastadon hadn't inadvertently
pulled in some serious undesirability by aping twitter. People who hate
tinfoil hat stuff should stop reading at the next punctuation mark, but
I was imagining a team of unscrupulous behavioural psychologists working
at Twitter Inc. to make their site maximally addicative, by having it
tweak but never fulfill whatever lack sends people like me to the
computer in search of connection or whatever else. That's what I thought
mastadon inherited and why it's something to avoid.

-- 
Mike Small
sma...@sdf.org
___
Discuss mailing list
Discuss@blu.org
http://lists.blu.org/mailman/listinfo/discuss

Re: [Discuss] Trying to install Tor on Fedora

[Mike Small]

Nancy Allison  writes:

> Hi, all. I sent this reply to Rich but forgot to include the list.
>
> I am your Test Case par excellence: someone trying to use Fedora who is not
> terribly technical. Can Linux be used by people for whom it is really a
> challenge, even things that to technical people are obvious? Here is a
> distillation of the later discussion with Rich.

The Tor project may be interested in your feedback. Maybe once you get
torbrowser going you could let them know where their instructions weren't
clear. My impression is they want their stuff to be accessible to
everyone, but it's not a huge project and they wouldn't have the money
to check if their instructions are useful to as large a cross section of
people as they would hope.

Also, it's not easy on Linux to give this kind of direction since we all
get to choose (as we should) what programs we'll use to access and run
other programs, e.g. what desktop environment, whether we like nautilus,
kde's file manager, or if we instead will run with something more spare
like twm and use rox-filer as the file manager. Or maybe someone prefers
to use a plan 9 text editor named acme, which makes a pretty decent file
manager as well, or to use emacs's dired mode, or only to use the
terminal to navigate. Point is, it's not possible to give the kind of
step by step instruction with screen shots you'll see in those very fat
books in the computer section and Barnes and Noble, because you can't
know exactly what people are running. The fallback is often to give
instructions one can run at a terminal program using the command
line. That's all that can be assumed to be universally available.

And Windows, in as much as it's simpler by foisting their mall
kioskesque disaster of a UI on everyone, still manages not to be that
great, in fact. Just the other day, on the laptop my employer makes
me run a proprietary operating system on I had to do this:

https://www.tenforums.com/browsers-email/101100-make-firefox-default-app-web-browser.html

I suppose myself to be fairly technical, but I could not figure that out
on my own. (Btw. for all the talk about Microsoft being Linux friendly
with WSL or whatever, I noted in this configuration screen a message
suggesting that by using firefox or anything other than MS's new browser
as the default that I may not be doing what's best for my system. That
would have provoked roars of disapproval back in the day. Maybe
Microsoft hasn't gotten better. Maybe we've just gotten used to even
worse treatement by the new 800 pound gorillas out there, the Apples and
the Google/Samsung/Verizon/Android "who does this phone belong to
anyway?" style environments.)

>
> I downloaded the Linux file from the Tor site and it opened automatically.
> The next step I need to take is

In case you want to do the verification step, I'll try to give
instructions to use at a command prompt (run a program named terminal or
xterm or gnome-terminal, whatever you can find along those lines in your
menus). There may be a UI to do this, but I'm not familiar with what's
out there like that. Once you've got a prompt up in such a terminal
emulator program you can type the commands below:

1. change directories to where the tor software archive and the
corresponding signature (.asc) was downloaded.
e.g. ...
cd Download

2. try running gpg to verify the file:

$ gpg --verify tor-browser-linux64-8.0.8_en-US.tar.xz.asc 
tor-browser-linux64-8.0.8_en-US.tar.xz
gpg: Signature made Fri 22 Mar 2019 07:47:17 PM EDT using RSA key ID D9FF06E2
gpg: Can't check signature: public key not found

3. Since you probably also don't have the public key from the tor
project in your key ring, get that. It kind of defeats the purpose of
this whole check, this fact, except that at least once you get the key
once your later checks will have it, so you'll at least narrow your
exposure to being fooled the first time you downloaded torbrowser, its
signature, and the public key needed to verify instead of having the
potential to be fooled every time you download torbrowser (future
upgrades).  To be safer, in theory, you could look across the signatures
of that public key until you arrive at someone's public key who you
recognize and trust.  I dunno, I tried this with the tor key the other
day and ran out of steam before reaching anyone I'd heard of to where I
had any kind of meaningful trust in the public key I retrieved. I mean,
I felt like is was close to linking up with Poul-Henning Kamp, a well
known FreeBSD developer with a known email, but I could only match up
one of the tor developer's key to his and not the one actually used to
sign the archive. PGP's web of trust kind of breaks down in cases like
these I think.

$ gpg --keyserver pgp.mit.edu --recv-keys D9FF06E2
gpg: requesting key D9FF06E2 from hkp server pgp.mit.edu
gpg: key 93298290: public key "Tor Browser Developers (signing key) 
" imported
gpg: no ultimately trusted keys found
gpg: Total number processed: 1
gpg:

Re: [Discuss] mtpfs question

[Mike Small]

Dan Ritter  writes:

> dan moylan wrote: 
>> 
>> great -- i've got termux installed on the tablet and sshd
>> running, but where are the kindle files on the tablet?
>> 
>> i've poked around a bit and found nothing -- of course, /data
>> and /data/data are inaccessable.
>>  
>> and when i ssh in:
>> 
>>   u0_a5 ~[40] ls
>>   The program 'ls' is not installed. Install it by executing:
>>pkg install busybox
>>   or
>>pkg install coreutils
>>   u0_a5 ~[41] pkg install coreutils
>>   The program 'pkg' is not installed. Install it by executing:
>>pkg install termux-tools
>>   u0_a5 ~[42] pkg install termux-tools
>>   The program 'pkg' is not installed. Install it by executing:
>>pkg install termux-tools

It's been awhile since I installed termux, but I don't remember having
to install coreutils myself. Plainly something is wrong if you can't run
pkg. apt-get should be there too, which works just as well, but if pkg
is not runnable my guess is that isn't working either. What does pwd and
what is your $PATH here?  /usr/bin for termux on my phone is
/data/data/com.termux/files/usr/bin/. Can you cd to that and run ./ls?

Or another thought, could you have another sshd application
running. i.e. you think you're sshing to termux's sshd when in fact
you're connecting to another one, so your uid and PATH are not termux's,
meaning you can't run anything termux. (GNU stands for GNU is not Unix
but clearly Android really is not Unix. ;) I think I had to run termux
on the phone, install sshd, and then type sshd at the shell prompt
there. Periodically I have to start it up again from the phone when
termux completely goes away.

That doesn't entirely fit either. If another application gives you sshd,
how can it not give you ls? And clearly you are in a termux shell if
it's telling you to run pkg. But maybe the above will give you debugging
ideas even if it's not on the mark.

I'm not at the machine I do my sshing from. I might be sshing as the
termux user. Could that be all it is? Wrong user?

>> 
>> on the tablet, ls and pkg work as expected -- i must be sshing
>> in wrong somehow.  hmmm ... where do i go from here?
>
> Very likely /sdcard.

With any luck.  My android phone has an /sdcard/Books directory that I
didn't create and the Cool Reader ebook application I use will look
in. So you'd think Kindle would put things in there or somewhere else
accessible, but who knows.

The security model on Android, IIRC, is that every application runs as a
different user (and using other Linux features for isolation,
capabilities? cgroups? selinux?) and has a directory somewhere under
/data/data (which Dan seems aware of) that's not (neccessarily?)
accessible to you as the termux user or the user your file manager runs
as.

Maybe less so than iOS (but I don't know first hand), but Android seems
very much to make your phone not entirely your phone. This is why I'd
like to switch to debian or postmarketOS, or at least LineageOS,
whenever I figure out how to do that on my model and find a big block of
time to fiddle.

Rooting might help, but I don't understand the implications of doing
that fully enough yet to dare doing it on my main phone. I have a spare
I'm going to try it on one of these days.

-- 
Mike Small
sma...@sdf.org
___
Discuss mailing list
Discuss@blu.org
http://lists.blu.org/mailman/listinfo/discuss

Re: [Discuss] mtpfs question

[Mike Small]

Can you install termux on this thing and run sshd there? I've had good
luck that way using ssh and scp to get files off my phone. rsync works
too. Probably sshfs would work, but I never bother with that.

- Mike

dan moylan  writes:

> ole dan wrots:
>> moylan ~[1024] simple-mtpfs -l
>> 1: AsusMemoPad 7 (ME572CL)
>> moylan ~[1025] simple-mtpfs --device 1 ~/mnt
>> moylan ~[1026] ls /mnt
>
>> sorry -- i missed the tilde -- all is well.
>
> and indeed it was -- at least the android tablet directories
> showed up under ~/mnt, though ~/mnt/kindle did not show in
> any readable fashion the various books that showed in the
> kindle reader.
>
> however, today i started in again and got this:
>
>   moylan ~[1065] simple-mtpfs -l
>   1: AsusMemoPad 7 (ME572CL)
>   moylan [1066] simple-mtpfs --device 1 ~/mnt
>   LIBMTP PANIC: Trying to dump the error stack of a NULL device!
>
> now what brought THAT on?  what can i have done to deserve
> this?
>
> ole dan
>
> j. daniel moylan
> 84 harvard ave
> brookline, ma 02446-6202
> 617-777-0207 (cel)
> j...@moylan.us
> www.moylan.us
> [no html pls]
> _______
> Discuss mailing list
> Discuss@blu.org
> http://lists.blu.org/mailman/listinfo/discuss
>

-- 
Mike Small
sma...@sdf.org
___
Discuss mailing list
Discuss@blu.org
http://lists.blu.org/mailman/listinfo/discuss

Re: [Discuss] Discuss Digest, Vol 94, Issue 3

[Mike Small]

Mayuresh Rajwadkar  writes:

> Ubuntu Install & Swap & Partitions.
>
> I understand that the typical Linux kernel takes up about 1GB of RAM, so
> assume it will run easily in 2GB of RAM
> If your host station has  16GB or more of RAM, for most applications you
> really dont need any swap partition.
>

16GB is a lot of RAM, but if I had it I'd still create a swap partition
of some size. Linux will put it to good use if you have it:

The casual reader may think that with a sufficient amount of memory,
swap is unnecessary but this brings us to the second reason. A
significant number of the pages referenced by a process early in its
life may only be used for initialisation and then never used
again. It is better to swap out those pages and create more disk
buffers than leave them resident and unused.

-- 
https://www.kernel.org/doc/gorman/html/understand/understand014.html#chap:%20Swap%20Management

An interesting bug report on this topic:

https://github.com/kubernetes/kubernetes/issues/53533

Scroll down to anguslees's comment on Oct 5, 2018. He discusses similar
reasons to what I quoted above for why it's good to get the inactive
pages out to swap.

But then the question is what size should the swap partition be?  I
recall seeing advice that the 2 x RAM recommendation is outdated, but
when I was looking up the other info above, I noticed this blog about
SAP and Linux. According to the author, SAP also feels that 2 x RAM is
obsolete, but their cut off is way higher than I would have thought
(granted, I'm not trying to run SAP software), recommending 2 x RAM if
you have 32 GiB or less:

http://sapbasiskishore.blogspot.com/2014/08/note-1597355-swap-space-recommandation.html

Oh, in case you didn't follow the link and need to know, 320 GiB of swap
is the most you should ever need, even if you have more than 8 TiB of
RAM. Phew.

- Mike

>
>
> On Fri, Mar 8, 2019 at 12:00 PM  wrote:
>
>> Send Discuss mailing list submissions to
>> discuss@blu.org
>>
>> To subscribe or unsubscribe via the World Wide Web, visit
>> http://lists.blu.org/mailman/listinfo/discuss
>> or, via email, send a message with subject or body 'help' to
>> discuss-requ...@blu.org
>>
>> You can reach the person managing the list at
>> discuss-ow...@blu.org
>>
>> When replying, please edit your Subject line so it is more specific
>> than "Re: Contents of Discuss digest..."
>>
>>
>> Today's Topics:
>>
>>1. Re: Ubuntu Install Question (Mike Small)
>>2. Re: Ubuntu Install Question (Rich Pieri)
>>
>>
>> --
>>
>> Message: 1
>> Date: Thu, 07 Mar 2019 20:50:03 +
>> From: Mike Small 
>> To: Jerry Feldman 
>> Cc: Boston Linux and Unix 
>> Subject: Re: [Discuss] Ubuntu Install Question
>> Message-ID: 
>> Content-Type: text/plain; charset=utf-8
>>
>>
>> That's good to hear. The parted developers certainly don't lack in
>> bravery or ambition when it comes to what features to try to implement
>> and maintain. In a different context while tracking down a bug I noticed
>> that parted probes filesystem superblocks checking for everything from
>> nilfs2 to amiga's filesystem to Apple's HFS. I'd never even heard of
>> nilfs. All useful features to someone I guess.
>>
>> Jerry Feldman  writes:
>>
>> > Over the years running installfests I have used a number of partitioning
>> > tools. Today, GPARTED is included in most Linux distros and is used by
>> the
>> > installers, I generally trust both the Fedora installers (anaconda) and
>> the
>> > Ubuntu installers to shrink windows partitions. In recent years I have
>> not
>> > had an issue. Sometimes I may boot live USB and use gparted to resize the
>> > partitions. Then I run Windows to double check.
>> >
>> > On Thu, Mar 7, 2019 at 3:28 PM Mike Small  wrote:
>> >
>> >>
>> >> Hi Ivan,
>> >>
>> >> The Linux partitioning tools (that the installation programs use) claim
>> >> to be able to shrink existing partitions to make room for a Linux
>> >> partition, but I've never trusted that. Could be paranoia on my
>> >> part. I'd be especially careful when Windows is involved, but I have no
>> >> recent experience with it. If it were me I'd put Ubuntu on the 2nd drive
>> >> and leave Windows with what it originally had. I haven't installed
>> >> Ubuntu in this situation, so I don't know how much help it gives you,
>> >> but I believe the boot loader

Re: [Discuss] Ubuntu Install Question

[Mike Small]

That's good to hear. The parted developers certainly don't lack in
bravery or ambition when it comes to what features to try to implement
and maintain. In a different context while tracking down a bug I noticed
that parted probes filesystem superblocks checking for everything from
nilfs2 to amiga's filesystem to Apple's HFS. I'd never even heard of
nilfs. All useful features to someone I guess.

Jerry Feldman  writes:

> Over the years running installfests I have used a number of partitioning
> tools. Today, GPARTED is included in most Linux distros and is used by the
> installers, I generally trust both the Fedora installers (anaconda) and the
> Ubuntu installers to shrink windows partitions. In recent years I have not
> had an issue. Sometimes I may boot live USB and use gparted to resize the
> partitions. Then I run Windows to double check.
>
> On Thu, Mar 7, 2019 at 3:28 PM Mike Small  wrote:
>
>>
>> Hi Ivan,
>>
>> The Linux partitioning tools (that the installation programs use) claim
>> to be able to shrink existing partitions to make room for a Linux
>> partition, but I've never trusted that. Could be paranoia on my
>> part. I'd be especially careful when Windows is involved, but I have no
>> recent experience with it. If it were me I'd put Ubuntu on the 2nd drive
>> and leave Windows with what it originally had. I haven't installed
>> Ubuntu in this situation, so I don't know how much help it gives you,
>> but I believe the boot loader (Grub) has the ability to do this. Someone
>> else could give better advice here.
>>
>> As a general comment, whenever you partition, make filesystems or
>> install 2nd systems, you want to know in advance that if you make a
>> mistake what you lose is no big deal. For most people that means that
>> everything you care about (including what you didn't remember you had
>> and would care about) needs to be backed up and the back up confirmed to
>> some degree to be recoverable. Another answer I often flirt with is to
>> know in advance that I'll be able to shrug off accidents thinking,
>> "well, I didn't really need that anyway -- how many times do I need to
>> listen to Cortez the Killer or all those other old songs anyway," but
>> now that I have child photos that's a less interesting disaster recovery
>> strategy than it used to be.
>>
>> - Mike
>>
>> Ivan Klimov  writes:
>>
>> > Hi Mike,
>> >
>> > Thank you for the clarification.
>> >
>> > I thought of swap partition to be more for temp file storage/exchange.
>> > Thank you for introducing the concept. When you said shuffle parts of
>> > programs... I have not done it and do not envision how it is done. Maybe
>> it
>> > is an advanced concept for me;) I have installed FULL programs on
>> different
>> > partitions -never parted programs out.
>> >
>> > Do you think it is a good idea to partition 50/50 SSD where windows 10 is
>> > already installed ?.. and install Ubuntu on the second partition ? Is it
>> > safe ?
>> >
>> > Best,
>> > Ivan
>> >
>> >
>> >
>> > On Thu, Mar 7, 2019, 12:10 PM Mike Small  wrote:
>> >
>> >> Ivan Klimov  writes:
>> >> ...
>> >> > Excerpt from the post:
>> >> > On this step we’ll create our custom partition layout for Ubuntu
>> 18.04.
>> >> On
>> >> > this guide will recommend that you create two partitions, one for root
>> >> and
>> >> > the other for home accounts data and a partition for swap (use a swap
>> >> > partition only if you have limited RAM resources or you use a fast
>> SSD).
>> >> >
>> >> > My thought was
>> >> > DISC C (SSD):
>> >> > Partition 0:
>> >> > Win10 : ~60 GB SSD
>> >> > Partition 1:
>> >> > UBuntu: ~ 200 GB SSD
>> >> >
>> >> > Disc D:
>> >> > Partition 2:
>> >> > 650 GB HD
>> >> > Partition 3:
>> >> > 350 GB HD
>> >> >
>> >> > home accounts data and a partition for swap (use a swap partition)
>> >> > What is swap partition ?
>> >> >
>> >>
>> >> A swap partition is disk space the operating system can use to
>> >> temporarily shuffle parts of programs out of the way when memory gets
>> >> low or when it can think of better ways of using memory than holding a
>> >> part of a program that isn't used much. Sometimes the OS

Re: [Discuss] Ubuntu Install Question

[Mike Small]

Hi Ivan,

The Linux partitioning tools (that the installation programs use) claim
to be able to shrink existing partitions to make room for a Linux
partition, but I've never trusted that. Could be paranoia on my
part. I'd be especially careful when Windows is involved, but I have no
recent experience with it. If it were me I'd put Ubuntu on the 2nd drive
and leave Windows with what it originally had. I haven't installed
Ubuntu in this situation, so I don't know how much help it gives you,
but I believe the boot loader (Grub) has the ability to do this. Someone
else could give better advice here.

As a general comment, whenever you partition, make filesystems or
install 2nd systems, you want to know in advance that if you make a
mistake what you lose is no big deal. For most people that means that
everything you care about (including what you didn't remember you had
and would care about) needs to be backed up and the back up confirmed to
some degree to be recoverable. Another answer I often flirt with is to
know in advance that I'll be able to shrug off accidents thinking,
"well, I didn't really need that anyway -- how many times do I need to
listen to Cortez the Killer or all those other old songs anyway," but
now that I have child photos that's a less interesting disaster recovery
strategy than it used to be.

- Mike

Ivan Klimov  writes:

> Hi Mike,
>
> Thank you for the clarification.
>
> I thought of swap partition to be more for temp file storage/exchange.
> Thank you for introducing the concept. When you said shuffle parts of
> programs... I have not done it and do not envision how it is done. Maybe it
> is an advanced concept for me;) I have installed FULL programs on different
> partitions -never parted programs out.
>
> Do you think it is a good idea to partition 50/50 SSD where windows 10 is
> already installed ?.. and install Ubuntu on the second partition ? Is it
> safe ?
>
> Best,
> Ivan
>
>
>
> On Thu, Mar 7, 2019, 12:10 PM Mike Small  wrote:
>
>> Ivan Klimov  writes:
>> ...
>> > Excerpt from the post:
>> > On this step we’ll create our custom partition layout for Ubuntu 18.04.
>> On
>> > this guide will recommend that you create two partitions, one for root
>> and
>> > the other for home accounts data and a partition for swap (use a swap
>> > partition only if you have limited RAM resources or you use a fast SSD).
>> >
>> > My thought was
>> > DISC C (SSD):
>> > Partition 0:
>> > Win10 : ~60 GB SSD
>> > Partition 1:
>> > UBuntu: ~ 200 GB SSD
>> >
>> > Disc D:
>> > Partition 2:
>> > 650 GB HD
>> > Partition 3:
>> > 350 GB HD
>> >
>> > home accounts data and a partition for swap (use a swap partition)
>> > What is swap partition ?
>> >
>>
>> A swap partition is disk space the operating system can use to
>> temporarily shuffle parts of programs out of the way when memory gets
>> low or when it can think of better ways of using memory than holding a
>> part of a program that isn't used much. Sometimes the OS will even swap
>> out whole programs when you're really pushing your memory usage.  The
>> suggestion used to be to make it twice the size of your RAM, but I've
>> seen advice more recently to cap it at some amount smaller than
>> that. New machines have so much ram that if you ever activated that much
>> swap, well the experience would not be pleasant, waiting for all that
>> I/O.
>>
>> >
>> > Any input is welcome.
>>
>> I'm not going to try to give much advice on how to lay out your
>> partitions, since I always regret my choices later. One nice piece of
>> advice I read on an openbsd mailing list once was to only create
>> partitions holding the space you imagine right now that you'll need. The
>> rest leave to create partitions with later, when you'll know what it
>> should be used for. This advice maybe applies more to OpenBSD since it
>> has no options like LVM, ZFS, or btrfs. On the other hand, maybe you
>> don't feel like learning how to exercise those options right now, plus
>> if you're splitting with Windows, something like LVM won't help you if
>> you wanted that space instead for Windows.
>>
>> >
>> > Best,
>> > Ivan
>> >
>> >
>> >
>> >
>> >
>> > On Tue, Nov 20, 2018, 12:39 AM Ivan Klimov  wrote:
>> >
>> >> Eric,
>> >>
>> >> Greatly appreciate it !
>> >>
>> >> I will refer to the guide once will have some time to go through the
>> >> process.
>> >>
>> 

Re: [Discuss] Ubuntu Install Question

[Mike Small]

Ivan Klimov  writes:
...
> Excerpt from the post:
> On this step we’ll create our custom partition layout for Ubuntu 18.04. On
> this guide will recommend that you create two partitions, one for root and
> the other for home accounts data and a partition for swap (use a swap
> partition only if you have limited RAM resources or you use a fast SSD).
>
> My thought was
> DISC C (SSD):
> Partition 0:
> Win10 : ~60 GB SSD
> Partition 1:
> UBuntu: ~ 200 GB SSD
>
> Disc D:
> Partition 2:
> 650 GB HD
> Partition 3:
> 350 GB HD
>
> home accounts data and a partition for swap (use a swap partition)
> What is swap partition ?
>

A swap partition is disk space the operating system can use to
temporarily shuffle parts of programs out of the way when memory gets
low or when it can think of better ways of using memory than holding a
part of a program that isn't used much. Sometimes the OS will even swap
out whole programs when you're really pushing your memory usage.  The
suggestion used to be to make it twice the size of your RAM, but I've
seen advice more recently to cap it at some amount smaller than
that. New machines have so much ram that if you ever activated that much
swap, well the experience would not be pleasant, waiting for all that
I/O.

>
> Any input is welcome.

I'm not going to try to give much advice on how to lay out your
partitions, since I always regret my choices later. One nice piece of
advice I read on an openbsd mailing list once was to only create
partitions holding the space you imagine right now that you'll need. The
rest leave to create partitions with later, when you'll know what it
should be used for. This advice maybe applies more to OpenBSD since it
has no options like LVM, ZFS, or btrfs. On the other hand, maybe you
don't feel like learning how to exercise those options right now, plus
if you're splitting with Windows, something like LVM won't help you if
you wanted that space instead for Windows.

>
> Best,
> Ivan
>
>
>
>
>
> On Tue, Nov 20, 2018, 12:39 AM Ivan Klimov  wrote:
>
>> Eric,
>>
>> Greatly appreciate it !
>>
>> I will refer to the guide once will have some time to go through the
>> process.
>>
>> Best regards,
>> Ivan
>>
>> On Mon, Nov 19, 2018 at 10:39 PM Eric Luther  wrote:
>>
>>> I found this comprehensive and up to date walkthrough of the steps needed
>>> which includes pictures.
>>>
>>> https://askubuntu.com/a/1031994
>>>
>>> On Mon, Nov 19, 2018 at 10:34 PM Ivan Klimov  wrote:
>>>
>>>> Eric and Greg,
>>>>
>>>> Many thanks for your reply.
>>>>
>>>> I went to Microcenter and grabbed $1000 machine with 6 cores (12 threads)
>>>> 16 Gb RAM SSD -- awesome specs for the money.
>>>> If someone shops for a laptop, it is good time now.
>>>>
>>>> Now, I need to install Linux. The machine has 1 TB + 256 SSD (with
>>>> Win10).
>>>> Any recommendations on how to prepare for Ubuntu install ? How to divide
>>>> HD/SSD, etc.?  I would like to make it dual-boot.
>>>>
>>>> FYI.
>>>> I clarified with my colleague and compatibility problem was with the
>>>> motherboard - not the chip itself. Let's hope that this machine will not
>>>> have the same issue.
>>>>
>>>> Best,
>>>> Ivan
>>>>
>>>>
>>>>
>>>>
>>>>
>>>> On Mon, Nov 19, 2018 at 8:47 AM Greg Rundlett (freephile) <
>>>> g...@freephile.com> wrote:
>>>>
>>>> > On Sun, Nov 18, 2018 at 9:58 PM Ivan Klimov 
>>>> wrote:
>>>> >
>>>> >> Dear Linux Group,
>>>> >>
>>>> >> Background:
>>>> >> Friend of mine upgraded PC with new motherboard and RAM some time ago
>>>> and
>>>> >> could not install Ubuntu due to some conflicts with hardware.
>>>> >>
>>>> >> I am just about to buy a laptop with the following spec. Do you see
>>>> any
>>>> >> problems/conflicts  installing Ubuntu on this machine as dual boot ?
>>>> >
>>>> >
>>>> >
>>>> > You could compare details of your hardware with the official
>>>> "certified"
>>>> > hardware for Ubuntu desktops listed at
>>>> > https://certification.ubuntu.com/certification/
>>>> >
>>>> >
>>>> >> Is
>>>> >> there a way to check it beforehand ?
>>>> >>
>>>> >
>>>> > If you can put a LIVE CD into the product, you can boot it without
>>>> > installing Ubuntu.  However, I'll assume you don't have physical
>>>> access to
>>>> > the machine (ie. buying online).
>>>> >
>>>> > HTH. Some others on the list have more experience with installations,
>>>> so
>>>> > perhaps they'll offer advice too.
>>>> >
>>>> > Greg Rundlett
>>>> > https://eQuality-Tech.com <https://equality-tech.com/>
>>>> > https://freephile.org
>>>> >
>>>> >
>>>> ___
>>>> Discuss mailing list
>>>> Discuss@blu.org
>>>> http://lists.blu.org/mailman/listinfo/discuss
>>>>
>>>
> ___
> Discuss mailing list
> Discuss@blu.org
> http://lists.blu.org/mailman/listinfo/discuss
>
>

-- 
Mike Small
sma...@sdf.org
___
Discuss mailing list
Discuss@blu.org
http://lists.blu.org/mailman/listinfo/discuss

Re: [Discuss] Backing up evernote

[Mike Small]

Jerry Feldman  writes:
...
> My main concern with evernote is the rumors it may go out of business. So,
> I have two desires
> 1. Backup on my system (or Dropbox or drive) to preserve the notes.
> 2. Viable replacement for evernote that supports Android and Linux.

emacs seems to have a mode available for doing something or other with
evernote files: https://github.com/pymander/evernote-mode

But I guess if you were an emacs user you'd be using org-mode, plus
emacs is painful on Android. A person wanting a Lisp on Android should
probably instead use picolisp and some simpler editor.

This reminds me, does anyone have a way to enter text into an android
phone at a shell or editor prompt that gives the shell or editor the
whole display rather wasting half of it showing a soft keyboard? The
picolisp author has a solution called pentikeyboard, but I haven't been
able to figure out how to install it on my phone yet:
https://picolisp.com/wiki/?TermuxPentiPicoLisp
I wonder if it's not more suited for tablets than phones anyway. 
My hand is too large to drape fingers and thumb across the phone's screen.

-- 
Mike Small
sma...@sdf.org
___
Discuss mailing list
Discuss@blu.org
http://lists.blu.org/mailman/listinfo/discuss

Re: [Discuss] Linux on a Chromebook

[Mike Small]

Kent Borg  writes:

> Has anyone here run native Linux on a Chromebook? I don't mean running
> under Chrome OS (I don't trust Chrome OS), but a native Linux boot.
>
> I realize these things are low-powered, that's okay. I basically want
> a Raspberry PI-calibre machine, but with nice packaging, to use as
> notebook, for narrow purposes.
>
> Is this a sensible thing to do? (What should I look for in selecting
> hardware?)
>
> Thanks,
>
> -kb
>

The guys who did the podcast Garbage were into running OpenBSD on them
(and porting it to them). Sounded like a headache if you don't want an
exercise in driver development, but usually Linux has better hardware
support earlier, so maybe using it would be simpler. Interesting topic
for podcast episodes, though, even if you've already decided against it:

Episodes 37 and earlier here have some mention of Chromebooks:
https://garbage.fm/
And there's this linked from one of the episodes:
https://jcs.org/notaweblog/2016/08/26/openbsd_chromebook/

-- 
Mike Small
sma...@sdf.org
___
Discuss mailing list
Discuss@blu.org
http://lists.blu.org/mailman/listinfo/discuss

Re: [Discuss] How Daylight Saving Time Messes With Hospitals

[Mike Small]

Bill Ricker  writes:
>  If we wanted to join India and Newfoundland as rebels, we could even
> split off N.E. as a half-hour timezone based on our local meridian, or
> adopt permanent DST by joining NB/NS/PEI as AST (but without ADT/DST).

If we join Newfoundland in taking a half hour offset we should also
start celebrating Guy Fawkes day like they do. I'm sure people around
here would love a good bonfire.

> 20 years ago either permanent DST (=AST year round) or half-hour zone
> would not have been practical because network TV only recognized 2 TZ,
> ET & PT, but with streaming/binging/DVR today, would anyone care if
> GoT came on an hour "later" in Boston just as it is an hour "earlier"
> in Chicago today?

Well, there's still the late night shows. Sure you can time shift them,
but somehow it's just not the same. I grew up on Atlantic time and
staying up for Letterman was a special little rebellious indulgence. And
then there's sports. People like to watch with others in bars and so
on. It's hard in Boston to stay late in bars, what with transit not
running very late and with so many people having to live so far away
from the affordability problems. On the other hand maybe fans would like
the idea of being in a different time zone than NYC just because.

-- 
Mike Small
sma...@sdf.org
___
Discuss mailing list
Discuss@blu.org
http://lists.blu.org/mailman/listinfo/discuss

Re: [Discuss] How Daylight Saving Time Messes With Hospitals

[Mike Small]

Rich Pieri  writes:

> Lots of articles about this over the weekend, and they all quote the
> same line about how handling a time change is simple and easy.
>
> Fact is, it's anything but simple or easy when it comes to medical
> records. Here's an example: what time were you born? If your birthday
> for a given year falls between the second Sunday of March and the first
> Sunday of April or between the last Sunday of October and the first
> Sunday of November then you're probably wrong because the days for
> daylight savings changed in 2007. Your times are off by about 1 hour
> unless you accomodate that change or your place of birth and your
> current location do not honor daylight savings time.
>
> Multiply that by many hundreds of millions of patient records across
> many years of patients' lives, many locales and time zones and
> timekeeping changes, and inconsistencies across different
> record-keeping procedures, and you have a bonafide nightmare.
>
> Is there a solution? I don't think so short of doing away with daylight
> savings. Epic have been at this for almost 40 years. If it were
> possible and viable I think they'd have figured it out by now.

If database programmers would only always use UTC for their storage
format and translate as necessary for presentation it wouldn't be so
bad, but I guess in your example case it's too late now.

Maybe you'd like this article if you haven't seen it already:
http://naggum.no/lugm-time.html

-- 
Mike Small
sma...@sdf.org
___
Discuss mailing list
Discuss@blu.org
http://lists.blu.org/mailman/listinfo/discuss

Re: [Discuss] Feedspot and other RSS Readers

[Mike Small]

Nancy Allison  writes:

> Hi, all.
>
> What do you use to aggregate the things you read? I've stumbled upon
> Feedspot, which costs $$, and I'm wondering if it is necesssary.
>

Other than a few news websites I visit directly in a browser, most of
what I read from the Internet I aggregate with Gnus, which is a USENET
reading package for the emacs text editor that also handles email and
some RSS formats. There's a snippet in the emacs wiki to make it handle
atom feeds, which improves its feed support somewhat. I'm uncertain now
because I don't read many RSS feeds and am not motivated to make them
work when they don't, but I want to say it won't accept some (newer?)
RSS formats too, RSS 2.0 vs. 1.0 maybe, which can be
disappointing. Still, way more than RSS I use gmane and gwene USENET
views of mailing lists and a few websites and it works very well for
those: http://www.gmane.org/

I've looked at separate RSS readers in the past, but over time what
isn't displayed in the same list of "groups" along with my mail and
USENET groups I'll eventually ignore. Two master lists of things to read
doesn't work for me, particularly with Gnus having this helpful feature
of pushing most frequently read groups to the top of my list.

As it is I have way more unread articles than time to read them, so not
getting an RSS feed working is not the worst thing that could
happen. Also, writing emacs lisp is really fun, so if I did find myself
with all kinds of time, I could extend Gnus with backends to handle
various kinds of information I want included in my list.

-- 
Mike Small
sma...@sdf.org
___
Discuss mailing list
Discuss@blu.org
http://lists.blu.org/mailman/listinfo/discuss

Re: [Discuss] New document on Unbound caching DNS server

[Mike Small]

Derek Martin  writes:

> On Thu, Sep 13, 2018 at 07:36:26PM -0400, Steve Litt wrote:
>> Hi all,
>> 
>> The Unbound DNS server is the new kid on the block. A lot of admins are
>> replacing BIND9 with Unbound, perhaps plus an authoritative DNS server
>> for their domain. 
>
> Why?  BIND9, for whatever flaws it may have, is robust,
> well-understood software.  What advantages does Unbound offer that
> outweigh the benefit of running well established code?
>

My impression is that unbound and nsd are not new or experimental code.

I'm not a system admin only a user but I'll take a shot at some
justifications...

BIND9's source code is no joy to read. Anyone who's tried to maintain a
patch against it has my sympathy.  I'd guess the number of people for
whom this software is well understood at a source code level is actually
quite small. I haven't looked at unbound's code, but I suspect if OpenBSD
was willing to take it in (they commit to auditing what they include in
base) that it's probably an improvement as far a readability goes at the
very least.

Second, I'll give the diversity argument. There will continue to be
security holes in bind9 (and in unbound and nsd). Some people running
other things may mitigate the global risk of one severe incident.

Third, possibly someone has special requirements or perceptions of the
different projects that make unbound and nsd more attractive to them. At
least in 2012 (and apparently long before), OpenBSD felt unbound fit
their needs better than Bind9:
https://marc.info/?l=openbsd-misc=132921194328662=2

>> More interesting still, a lot of laptop owners are installing Unbound
>> to replace their old 8.8.8.8 or per-accesspoint resolvers with a full
>> caching DNS, which is more secure, faster, and makes for much faster
>> browsing.
>
> FWIW, this is often a bad idea.  On average, you will typically get
> the best overall performance by using your ISP's DNS servers (unless
> you know they're bad).  If you care about why, the short answer is
> CDNs, but here's a somewhat lengthy explanation:

If you set your resolver to be both caching and forwarding (meaning when
it doesn't have the record in cache it goes to your ISP's server or
whatever substitute you use) this isn't an issue I think.  Whether it's
worth the bother to set it up on a home network is another question. It
might be fun if you're into that sort of thing, or it might be good for
practice.

-- 
Mike Small
sma...@sdf.org
___
Discuss mailing list
Discuss@blu.org
http://lists.blu.org/mailman/listinfo/discuss

Re: [Discuss] New document on Unbound caching DNS server

[Mike Small]

OpenBSD 5.6 switched to unbound instead of bind9 in their base system
back in fall of 2014:  https://www.openbsd.org/plus56.html

For that reason, I had it as a forwarding caching nameserver on my
router on an old soekris board before I switched to getting internet
only from my phone's data plan and stopped using a router. Seemed to
work well enough, not that caching my home network is any kind of
test. Now that I think about it I'm not sure why I thought running a
caching nameserver was worth doing (thinking I had some muddled idea
it would increase privacy). But that board only had 128 MiB and
a chip that showed up in dmesg as a 486, in case that's worth anything
to people wondering about its resource requirements.

Steve Litt  writes:

> Hi all,
>
> The Unbound DNS server is the new kid on the block. A lot of admins are
> replacing BIND9 with Unbound, perhaps plus an authoritative DNS server
> for their domain. 
>
> More interesting still, a lot of laptop owners are installing Unbound
> to replace their old 8.8.8.8 or per-accesspoint resolvers with a full
> caching DNS, which is more secure, faster, and makes for much faster
> browsing.
>
> At  http://troubleshooters.com/linux/unbound_nsd/unbound.htm  I've 
> created a new document detailing the installation and setup of Unbound,
> including:
>
> * Making it useable all across your LAN
>
> * Optimizing for lookup speed with pre-priming and other techniques
>
> * Enabling remote control
>
> * Making DNS forward and reverse resolution of LAN local computers
>   available across the LAN, without using a traditional Authoritative
>   DNS.
>
> * Landmines and gotchas.
>
> * Forwarding to a traditional authoritative DNS server.
>
> Hope you like it.
>
> SteveT
>
> Steve Litt 
> September 2018 featured book: Quit Joblessness: Start Your Own Business
> http://www.troubleshooters.com/startbiz
> ___
> Discuss mailing list
> Discuss@blu.org
> http://lists.blu.org/mailman/listinfo/discuss
>

-- 
Mike Small
sma...@sdf.org
___
Discuss mailing list
Discuss@blu.org
http://lists.blu.org/mailman/listinfo/discuss

Re: [Discuss] OT - new legislation limiting use of "non-competition agreements"

[Mike Small]

Steve Litt  writes:
> I wish everyone would ask, ahead of time, to read all employment papers
> they're required to sign, and before resigning their old position, walk
> away from anyone trying to ram non-compete down their throats. The way
> employers get away with this stuff is most people sign stuff without
> reading it, and also, by the time they're presented with the
> non-compete, they've resigned their old job.

That implies we feel we have alternatives and some sense of a bargaining
position. Heck, I wouldn't even be seen publicly commenting on this
topic.  Oh darn.

-- 
Mike Small
sma...@sdf.org
___
Discuss mailing list
Discuss@blu.org
http://lists.blu.org/mailman/listinfo/discuss

Re: [Discuss] Can't play videos in Firefox

[Mike Small]

Is everything checked in this page?

https://www.youtube.com/html5

I don't recall now what to do if it's not, but that site helped me once
when trying to play a DRMed youtube video I bought for my son. It at
least helps zero in on what's not working.

Nancy Allison  writes:

> Hi, John and all.
>
> Thanks for these suggestions. My goodness, what a messy situation. I'm
> copying Jerry Feldman directly, because I know he uses Fedora and knows it
> well. Jerry, is there any better way that John and I can both benefit from?
>
> Thanks.
>
> --Nancy
>
> On Mon, Jul 23, 2018 at 1:02 AM, John Abreau  wrote:
>
>> I've had the same problem for several years. My workaround is to use a
>> command-line tool to download the videos, then play them with mplayer.
>>
>> The command-line tool is called "youtube-dl". You can install it with
>> dnf, but the version in dnf doesn't get updated very often, so I
>> prefer to install it manually so I can easily fetch the latest update
>> as needed.
>>
>> youtube-dl can dowload videos from a large number of sites, including
>> youtube, facebook, and vimeo. I just tried it on a washington post
>> page, and while it didn't work on the page directly, it worked when I
>> viewed the page source and then used youtube-dl on the embedded video
>> url, which was from abc.com.
>>
>> https://rg3.github.io/youtube-dl/
>>
>>
>>
>> On Sun, Jul 22, 2018 at 9:23 PM, Nancy Allison
>>  wrote:
>> > Follow-up: I've followed steps 1 and 1 (that's how they're numbered) on
>> > this page and installed RPM Fusion for Fedora 28. Then I rebooted. Still
>> no
>> > luckc playing videos in Facebook, Washington Post, or Vimeo (so far).
>> > Suggestions gratefully received.
>> >
>> > https://rpmfusion.org/
>> >
>> >
>> >
>> > On Sun, Jul 22, 2018 at 6:02 PM, Nancy Allison <
>> nancythewrit...@gmail.com>
>> > wrote:
>> >
>> >> For what it's worth, I also can't play videos from the Washington Post
>> >> site, so it's not just Facebook.
>> >>
>> >> On Sun, Jul 22, 2018 at 4:48 PM, Nancy Allison <
>> nancythewrit...@gmail.com>
>> >> wrote:
>> >>
>> >>> Hi, all.
>> >>>
>> >>> I have just switched to Fedora 28 from Ubuntu, and now Facebook videos
>> do
>> >>> not run in Firefox.
>> >>>
>> >>> I googled this issue and found a topic about it:
>> >>>
>> >>> https://ask.fedoraproject.org/en/question/120605/video-probl
>> >>> ems-on-firefox-on-fedora-28/
>> >>>
>> >>> I entered the command suggested and got this result:
>> >>>
>> >>> [theauthor@new-host ~]$ sudo dnf groupupdate multimedia
>> >>> [sudo] password for theauthor:
>> >>> Last metadata expiration check: 2:28:43 ago on Sun 22 Jul 2018 02:16:22
>> >>> PM EDT.
>> >>> 13Group 'Multimedia' is already installed.
>> >>> No match for group package "gstreamer1-plugin-mpg123"
>> >>> Dependencies resolved.
>> >>> Nothing to do.
>> >>> Complete!
>> >>>
>> >>> I then entered the second command as suggested:
>> >>>
>> >>> [theauthor@new-host ~]$ sudo dnf install compat-ffmpeg28
>> >>> Last metadata expiration check: 2:30:13 ago on Sun 22 Jul 2018 02:16:22
>> >>> PM EDT.
>> >>> No match for argument: compat-ffmpeg28
>> >>> Error: Unable to find a match
>> >>>
>> >>> I don't know what to do next to fix this problem. All suggestions
>> >>> gratefully received.
>> >>>
>> >>
>> >>
>> > ___
>> > Discuss mailing list
>> > Discuss@blu.org
>> > http://lists.blu.org/mailman/listinfo/discuss
>>
>>
>>
>> --
>> John Abreau / Executive Director, Boston Linux & Unix
>> Email: abre...@gmail.com / WWW http://www.abreau.net / PGP-Key-ID
>> 0x920063C6
>> PGP-Key-Fingerprint A5AD 6BE1 FEFE 8E4F 5C23  C2D0 E885 E17C 9200 63C6
>>
> ___
> Discuss mailing list
> Discuss@blu.org
> http://lists.blu.org/mailman/listinfo/discuss
>

-- 
Mike Small
sma...@sdf.org
___
Discuss mailing list
Discuss@blu.org
http://lists.blu.org/mailman/listinfo/discuss

Re: [Discuss] ugly problem

[Mike Small]

dan moylan  writes:

> since yesterday i have been having irregular freezeups on my
> computer.  for instance, today i was reading the nytimes via
> chrome, and suddenly the mouse ceased to function, nor would
> the touchpad on my keyboard.  the screen content remained
> unchanged, the mouse pointer visible but stationary.
> neither ctl-alt-del nor ctl-alt-f2 had any effect.  i
> attempted to ssh in from another computer and got "no route
> to host", nor could i ping.  a hard reset restores apparent
> normalcy -- at least for a while.  times vary.  i have yet
> to experience a freeze w/o the browser running (which of
> course is seldom the case).
>
> the computer is an intel nuc-10 running fc27.
>
> any suggestions would be appreciated.

What you describe sounds consistent with running low on memory and the
system swapping a lot or downright thrashing. If you walk away and come
back several minutes later do things come back to normal at all?  You
might also try running top in a terminal or some other memory monitoring
software to verify the pig process hypothesis.

If other ideas in this thread don't work for you, chrome may have an
extension to disable javascript selectively on parts of sites (like
noscript in firefox can). You'll need at least some of the nytimes' own
urls enabled for javascript so that you see a login button and pressing
it does something, but perhaps without javascript on some of the other
domains that load along with their pages the memory use will be small
enough to suit your machine.

When people criticize firefox and chrome for bloat they sometimes don't
give proper "credit" to the authors of the websites themselves. I run
firefox on very modest hardware (a 2 GiB RAM Core Duo based laptop made
in 2006). When I have to use the web I prefer dillo where possible, but
sadly, unlike the Guardian, NYTimes is not a site you can read with
dillo. Their web developers seem not as skilled as their reporters. I
manage to at least keep up with Paul Krugman's column using Firefox,
though. I can only explain the lack of problems I have by my use of
Firefox extensions like noscript, privacy badger, and uBlock
Origin. Well, perhaps using twm as my Window manager doesn't hurt
either, but my guess is that noscript is helping me more than that. With
javascript enabled there is no limit to what sites can throw at your
machine (particularly if you aren't up to date with meltdown and spectre
mitigations ;)).


-- 
Mike Small
sma...@sdf.org
___
Discuss mailing list
Discuss@blu.org
http://lists.blu.org/mailman/listinfo/discuss

Re: [Discuss] Guido van Rossum steps down

[Mike Small]

Kent Borg  writes:

> I think there is turmoil ahead for Python.

On the other hand I feel like Python has millions of users who don't
even know who Guido van Rossum is. I don't know if that's a sad or happy
thought for Python enthusiasts.

> In hindsight, they made a mistake to break compatibility in 3.0, yet

Is this bothering people very much? My impression was that the
adjustments tend to be minimal and that there are tools to
help. Granted, the person who told me this is very gung ho on the
language.

> didn't take the opportunity to fix the global interpreter lock that
> keeps Python from doing multithreading very well. Does not bode well.
>
> -kb, the Kent who is learning Rust, a language that does
> multithreading well.

People now expect good multithreading in scripting languages? Or have
the notational and/or library improvements in the C++ style languages
come along enough to where the scripting language conveniences don't
stand out as much? I always assumed Rust was sort of the next D (and
might top out around the same userbase size?) and wouldn't be considered
a substitute for the same kind of program where Perl, Python or Ruby
seemed to make sense. It's staticly typed and relatively verbose, right?
Big change from Python.

Well, let me know when it really is on the downward trajectory, so I can
finally get around to properly learning it. Work is sometimes easier to
find when you claim to know languages that were once very popular but
have since been damned with the "legacy" label.

-- 
Mike Small
sma...@sdf.org
___
Discuss mailing list
Discuss@blu.org
http://lists.blu.org/mailman/listinfo/discuss

Re: [Discuss] Running a mail server, or not

[Mike Small]

Dan Ritter  writes:
>
> So when the very first line of the entry is:
>
>   The nnml spool mail format isn’t compatible with any other known
>   format. It should be used with some caution. 
>
> you should read that as "don't use this, it was an experiment in
> being better than everybody else that didn't work out".

Oh it's not as bad as all that. nnml is actually the format the manual
auther gives in his quick getting started section before documenting
that there are many backend options. I think it's more or less the
default or most commonly used backend for people not needing IMAP.  If
you read more of the manual you'll see that it's just the author's style
to be open about such things. E.g. if you read the section, "Comparing
Mail Back Ends," you'll see lots of caveats across formats (though true
enough, nnmaildir seems not to have them, except see later note about
nnml's speed). I did have some doubts once about the inode usage
(nnmaildir would have same issue), but a quick check of what sdf has for
inodes free put that concern to bed quite quickly. In that same section
there's also a statement that nnml is probably the fastest format for
reading mail. Gnus is not a racehorse like mutt so this is a factor to
consider.  I look now and also notice that the emacswiki page
corresponding to the manual's, "which backend," section had no one chime
in on nnmaildir. More suggestion that it's less exercised/known than
nnml.

As far as compatibility with standard formats used by other clients I
don't care. I've both used Gnus and nnml for a long time but have also
had the odd "forget about emacs and use mutt and vi" phases. When you
change you either write or find something to convert, or (more likely in
my case) just leave the old stuff where it is and fire up Gnus if you
ever want to read old email (which I don't tend to outside work). It's a
bit like using a new source control system. There's the deluxe and
correct way to convert and then there's the way your company will do
it. That's it, I'm not lazy only cost conscious. Aside from that,
future wavering may be less likely now. I'm a lot more tied to Gnus and
emacs ever since I started using the feature of org-mode where you can
make (emacs specific) hyperlinks to emails very easily in notes and
planning files. At best I might try another emacs client, and I've
already gotten my rmail phase out of the way. How many emacs mail
clients can there be?  No wait, don't answer that.

> The point of having IMAP access on your phone is not to have
> every feature from your desktop available on your phone. The
> point is to be able to read new messages which are important to
> you, search for a message that you need right now, and compose a
> short message right now.

I don't get any important messages at the email address in question. But
I'll use IMAP eventually probably.

>> What I'd really like is if someone made a mobile version of emacs,
>> somehow, maybe with some complicated gesture scheme for input. There's
>> some emacs person, I think, who's done something to make it possible to
>> keep two Gnusae's set of folders in sync across two machines. So if I
>> could run Gnus on the phone and use that person's scripts, that would be
>> the ideal. Probably will never happen.
>
> If Gnus read IMAP, you would get this for free.

Someone emailed me off list with some good advice about splitting
(automatically filing into folders) using IMAP and Gnus. I'll probably
look into that eventually. I'd be more likely to switch to that backend
thatn try taking nnmaildir for a ride.

This still leaves the issue that emacs doesn't have a good phone
interface yet. I'd still be using k9 or whatever on the phone, so the
Gnus IMAP support would only be a way of addressing the caveat I raised
about how my normal email checking on real computers messes up the IMAP
on the phone checking. A good thing, but not a huge problem I need to
solve right now.

I may be pessimistic, but I suspect using Gnus with IMAP will have
interesting corner cases I'll find out about over time. emacs is nice
and featureful, but lots of features and many options usually means
flakiness across some permutations. I wonder a bit about the size and
voice of the IMAP section in the manual (not original author?). It
starts out with, "The most popular mail backend is probably ‘nnimap’,"
but the rest of the section is spare and functional. The author didn't
feel the urge to get chatty like he did in the next mail section or in
the USENET sections, so I wonder whether nnimap is quite as much in the
strike zone for Gnus as is nnml. It's probably a newer feature grafted
on later vs. the old spool based stuff. Maybe it works as well maybe it
doesn't.

-- 
Mike Small
sma...@sdf.org
___
Discuss mailing list
Discuss@blu.org
http://lists.blu.org/mailman/listinfo/discuss

Re: [Discuss] Running a mail server, or not

[Mike Small]

Richard Pieri  writes:

> On 6/28/2018 4:03 PM, Mike Small wrote:
>> client side I think it can be made bearable. Or probably I should just
>> go find the instructions on sdf for setting up an IMAP client and
>> install one on the phone. One of these days.
>
> K-9 Mail. Get it. Setup is straightforward for reading and retrieving,
> maybe not so straightforward for sending depending on what account types
> you have at SDF.

Eventually, sure. sdf has IMAP set up and can also be used as an smtp
smarthost if you pay for the right account features (very reasonable,
his fees), so I assume it will be relatively simple:
http://sdf.org/?faq?EMAIL?03

But it's not something I frequently want to do, check email on my phone,
so I'm not in a hurry. Generally speaking, I'm in front of a good
computer all day, one with a monitor and keyboard. Then I leave and the
only person I want to communicate with will soon be sitting nearby. And
besides I've got a TODO list with 10 or 20 similarly sized tasks I care
more about.

An issue I expect to run into is that when I read my email at a real
computer I'll ssh in, start emacs with it loading Gnus, and that takes
email out of my spool and splits it off into different folders based on
a list of regular expressions (there are various options Gnus supports
for mail storage but I chose nnml:
http://www.gnus.org/manual/gnus_84.html#Mail-Spool). I'm thinking IMAP
won't pick up the mails where Gnus put them. Doesn't IMAP have it's own
idea of what a folder is and how that's to be set up?

And then the IMAP client wouldn't have Gnus's killer feature, the
ability to "expire" a mail so that it 1. isn't visible again unless I
open the folder to show read articles and articles with similar kinds of
marks and 2. in some number of weeks, but not the day before tomorrow
when I decide I want to keep it after all, it will automatically age out
and really delete those expired mails. With that and some other things,
I'm half thinking featurewise, the IMAP client might be such a
relatively impoverished interface to email that I might even prefer,
often, still reading it under termux/ssh, as clunky as that is.

On the other hand, it will be useful sometimes, like that time in the
airport where I was on vacation and hadn't started Gnus so the mail
would still be in the spool for IMAP to see. So eventually, sure, I'll
try IMAP.

What I'd really like is if someone made a mobile version of emacs,
somehow, maybe with some complicated gesture scheme for input. There's
some emacs person, I think, who's done something to make it possible to
keep two Gnusae's set of folders in sync across two machines. So if I
could run Gnus on the phone and use that person's scripts, that would be
the ideal. Probably will never happen.

-- 
Mike Small
sma...@sdf.org
___
Discuss mailing list
Discuss@blu.org
http://lists.blu.org/mailman/listinfo/discuss

Re: [Discuss] Running a mail server, or not

[Mike Small]

Derek Martin  writes:
> The trick is usually access.  Like I have no way to SSH into my
> server at the moment...  Technically I can do it from my phone, but
> I've found trying to do anything non-trivial on the phone is extremely
> tedious and time consuming, so while it can be done, not in the amount
> of time that wouldn't be extremely awkward while you're dealing with a
> sales clerk or whatever...

This reminds me of having to dig out an airline email at a Jet Blue gate
in DC. The woman at the gate needed a confirmation code for a
rescheduled flight before she could assign me a seat. She seemed very
bright but was also really impatient (two that often go together). Here
I was fumbling around sshing to my shell account, starting emacs + Gnus
and trying to get to the email with the code, when she gets fed up and
makes me hand over the phone. I started to try to warn her that it was
different, unix, mumble mumble, but she brushes that aside saying, "what
were you doing, space to scroll? Okay here it is... type, type, type,
have a nice day... next."

It is awkward. But with the right shell aliases on the phone side and
the right emacs macros (to quickly save off attachments) on the email
client side I think it can be made bearable. Or probably I should just
go find the instructions on sdf for setting up an IMAP client and
install one on the phone. One of these days.

-- 
Mike Small
sma...@sdf.org
___
Discuss mailing list
Discuss@blu.org
http://lists.blu.org/mailman/listinfo/discuss

Re: [Discuss] through the looking glass

[Mike Small]

Richard Pieri  writes:

> On 6/2/2018 12:42 PM, dan moylan wrote:
>> not portable?  where would that be?
>
> Not portable in that "[[" is specific to Bash. You may not find it in
> other Bourne-derived shells.

ksh, either the real thing or one of the pdksh derived ones, has it the
same way. I'm starting to get the impression that some large number of
what people call bashisms come from David Korn (the feature not the
implemention). I was reading a book of his and he actually deprecates [
in favour of [[, mostly because, as I think you implied further up, [[
is syntax and [ just a command, so [[ ]] avoids that awful kind of error
you'll get when a variable is empty and you use it in a comparison
without the idiomatic hack to make sure an operand doesn't become .

Which are the the shells that don't give you [[?  Looks like dash
doesn't have it, from the man page (so ash too I assume). What's the
expression, "posix me harder?"

-- 
Mike Small
sma...@sdf.org
___
Discuss mailing list
Discuss@blu.org
http://lists.blu.org/mailman/listinfo/discuss

Re: [Discuss] I Hate Ubuntu

[Mike Small]

Bill Ricker <bill.n1...@gmail.com> writes:
...
> *BSD is great for servers. As a "former" Security type, I respect OpenBSD's
> commitment to security out of the Box. A recent client had a mix of Centos
> and *BSD servers, worked fine. I might have notes to remind me whether it
> was FreeBSD or OpenBSD.
>
> What I'm curious about is which *BSD has good APCI support? I do want my
> laptop to act like a laptop. One advantage of Commercial distros - Ubuntu
> and RedHat - is they test new big brand laptops early. (OTOH I really
> prefer older Lenovos now, so maybe that matters less for me now?) Or should
> I give up and embrace Darwin/BSD (aka MacOSX) for the laptop?

OpenBSD's ACPI support is interesting, because they wrote their own ACPI
code whereas FreeBSD (and everyone else?) uses the reference
implementation from Intel:

https://www.freebsd.org/projects/acpi/
https://man.openbsd.org/acpi.4

I'd guess OpenBSD would work well on your older Lenovos, since that
seems a popular choice of people on the mailing lists, but perhaps not
immediatly on the newer laptops for the reason you name. The perception
out there may be that OpenBSD is for servers not laptops, but their
developers don't think that way:

https://www.tedunangst.com/flak/post/openbsd-laptops

My own experience is pretty limited since I've only tried it on two
laptops in the last 10 years or so. It works great on an Asus Z35Fm
(from 2006) but last I tried had sound and X problems on an HP Elitebook
8440 (from 2010).

Best to avoid NVidia video cards if you might end up wanting to run
OpenBSD. That's a big problem with that second machine, and the usual
response on the mailing lists to people having a bad time is that few
developers have the time to figure out Nvidia cards, what with there
being no decent developer documentation, so don't buy their
stuff. Though I think someone in NetBSD or maybe even DragonFlyBSD is
working on porting the Nouveau drivers over, so maybe someday someone
from OpenBSD will take that in. OpenBSD's kernel now imports Linux code
for AMD/ATI and Intel video cards, but not NVidia's.

Also, while I don't run it myself so have no first hand impressions, a
couple of French developers do a lot of work to make sure recent Gnome
runs well. There is or was apparently a large multinational using Gnome
on OpenBSD as the desktop environment for some large swath of their
users.

-- 
Mike Small
sma...@sdf.org
___
Discuss mailing list
Discuss@blu.org
http://lists.blu.org/mailman/listinfo/discuss

Re: [Discuss] I Hate Ubuntu

[Mike Small]

Robert Krawitz <r...@alum.mit.edu> writes:

> On Wed, 09 May 2018 12:40:38 +0000, Mike Small wrote:
>> These descriptions from https://netplan.io are a bit funny:
>>
>> "The network configuration abstraction renderer
>> Netplan is a utility for easily configuring..."
>>
>> abstraction renderer? Pfft. What's so abstract about network interfaces?
>
> Try running OpenShift Origin, say, on a system that also supports
> multiple guests (for example, an OpenShift cluster on a single
> domain!).

So maybe the problem is with wanting one operating system to fit all
problems. Was listening to a John Maddog Hall interview where he
described the "bad old days" when there were 7 or 8 operating system on
PDP _'s (forget the model he named), each for different purposes,
e.g. real time, real time but not so much as the last one, ...  So now
maybe the time is ripe to swing back a little.

No doubt the people coming up with this stuff are bright people, but
their problems are in tension with what I want when I'm a home user and
hobbiest. I'd think the people doing cloud work are in large
organizations and ought to settle on some specialized OS, or set of
OSes, or at least, distributions rather than make the rest of the world
walk along with them. But practically, I guess the more reasonable
solution is for me to budge and treat GNU/Linux, or at least the most
mainstream distros, the way I used to treat Windows (i.e. the stuff need
to know for work) and settle on a different OS for home use, where
possible.

And the world would be much more interesting if there were other heavily
used free software operating systems. Maybe there'd even be ones not
coming out of Unix heritage. E.g. it's a shame ReactOS isn't further
along so Win32 experts could hack on it and make it their own. Also
would be very nice to have things based on the old successes the
Smalltalk and Lisp people had.

-- 
Mike Small
sma...@sdf.org
___
Discuss mailing list
Discuss@blu.org
http://lists.blu.org/mailman/listinfo/discuss

Re: [Discuss] I Hate Ubuntu

[Mike Small]

Richard Pieri <richard.pi...@gmail.com> writes:

> On 5/8/2018 4:32 PM, A. Richard Miller wrote:
>> Here, Rich, try this:
>> http://www.ubuntugeek.com/disable-netplan-on-ubuntu-17-10.html
>> 
>> Then you can save your hate for more deserving targets.
>
> I think you're missing the point. Points.
>
> First, these Ubuntu installs are for product testing. For paying
> customers. Who won't be disabling Netplan. Which means disabling Netplan
> in the test environments DOES. NOT. HAPPEN.
>
> Second, the YAML version forces dependencies on NetworkManager, systemd,
> and a(nother) YAML parser without making management of network
> interfaces any better or easier by hand and only minimally by automation
> tools like Ansible.
>
> Netplan is vendor gratuitous changes and I will continue to hate Ubuntu
> for engaging in the practice.

These descriptions from https://netplan.io are a bit funny:

"The network configuration abstraction renderer
Netplan is a utility for easily configuring..."

abstraction renderer? Pfft. What's so abstract about network interfaces?

"Netplan reads network configuration from /etc/netplan/*.yaml which are
written by administrators, installers, cloud image instantiations, or
other OS deployments."

"The most useful configuration snippet (to bring up things via dhcp) is as 
follows:"

 network:
   version: 2
   renderer: NetworkManager

Okay, and how does NetworkManager know to choose dhcp from that?  Is
dhcp "version 2?" Probably not, eh? This maybe is the YAML "API" version
I'm going to guess.

Well then, aside from the problem of having to learn what YAML keys are
available or mandatory, and the unpleasantness of typing such things
vs. a well devised custom text format intuitively matching ifconfig (or
whatever command we're supposed to be using for that now on Linux, ip?),
or the near certainty that YAML will be considered passé in 5 years or
so and replaced with whatever the one true persistance format of the day
is, seems like we're not configuring anything here so much as telling
the system that we're using something else to configure the
interface.

Here's what the equivalent looks like on the machine I'm using right
now. urndis0 is the network interface device name for ethernet over USB
to my phone, which gives me my internet:

$ cat /etc/hostname.urndis0
dhcp
<<<>>>
Plus I can do the thing below here in a way I think you all
can understand without reading OpenBSD docs, but for a normal
ethernet interface this isn't necessary. (Since I wrote that
comment below I found out that Android is smart enough to remember when
you pressed the USB tethering slider and somehow slide it over when OpenBSD
is ready for it, but it's still handy to have this for when I forget.)
<<<>>
# Without turning on USB Tethering at exactly the right time
# during boot, urndis0 won't be up and rl0 will instead become 
# the default gateway. Undo that for when I manually restart networking
# for this interface.

!route del default 10.0.0.1
!route add default 192.168.42.129

<>

Here is the very difficult parsing code for that from /etc/netstart:

# Parse the hostname.if(5) file and fill _cmds array with interface
# configuration commands.
set -o noglob
while IFS= read -- _line; do
parse_hn_line $_line
done <$_hn

... and the relevant part of the case statement (all of parse_hn_line is
short enough to fit in one screen in case you doubt the other cases) in
parse_hn_line ...


dhcp)   _c[0]=
_cmds[${#_cmds[*]}]="ifconfig $_if ${_c[@]} down;dhclient $_if"
V4_DHCPCONF=true
;;

   '!'*)   _cmd=$(print -- "${_c[@]}" | sed 's/\$if/'$_if'/g')
_cmds[${#_cmds[*]}]="${_cmd#!}"
;;

So it's not like everything that isn't xml, yaml, or json has to be some
horrendous journey into natural language parsing, like say...
https://lists.debian.org/debian-user/2018/02/msg00095.html

Now, I'm not saying I've lost enthusiasm for GNU/Linux in general, but
when you run the two systems side by side, you start to notice how the
people developing the non-GNU parts of Linux userland, or perhaps the
more influential ones, are from a quite different development culture
than the more tradional BSD one, and in a different way from how the GNU
people were/are. It reminds me more of the .NET developer culture, which
isn't a bad thing necessarily, but not always to my first preference,
what with the intimidating terminology like abstraction renderers,
provisioners, providers or what have you.

-- 
Mike Small
sma...@sdf.org
___
Discuss mailing list
Discuss@blu.org
http://lists.blu.org/mailman/listinfo/discuss

Re: [Discuss] I Hate Ubuntu

[Mike Small]

David Kramer <da...@thekramers.net> writes:

> Yes and no.  As a Software Engineer and as someone who loves
> automation, the move to files that are more easily parsed by software
> allows us to get closer to the linux mentality of making small
> components you glue together.  It also allows for pre-written
> libraries to be used to parse the file into data objects in the code.
>
> At least it's not XML.

Still not much fun if your language is bourne shell, which I'm going to
guess would be the most common choice for getting at this kind of file.

But it's good. There are systems going in either direction to suit
either of our preferences.

-- 
Mike Small
sma...@sdf.org
___
Discuss mailing list
Discuss@blu.org
http://lists.blu.org/mailman/listinfo/discuss

Re: [Discuss] I Hate Ubuntu

[Mike Small]

Richard Pieri <richard.pi...@gmail.com> writes:

> And here is the rough equivalent for Netplan:
>
> network:
>   version 2:
>   ethernets:
> enp0s31f6:
>   addresses:
>   - 192.168.1.202/24
>   gateway4: 192.168.1.2
>   nameservers:
> addresses:
> - 192.168.1.202
> search:
> - rgo.gweep.net
>

Excessive use of YAML and JSON seem one of the current maladies among
developers. I'm starting to notice command line utilities with JSON
output options too, e.g. lsblk.  Hmmm, well, maybe that would be handy
to certain people who use certain languages, but it seems like bloat.

OpenBSD remains quite usable. Can't beat them if the question is one of
good design taste, IMO. At least I don't find myself constantly
scratching my head wondering who had an issue with the old incarnation
of a revamped (for the 2nd or 3rd time in some cases) subsystem there
like I do with Linux. You won't find as many device drivers of
course.

-- 
Mike Small
sma...@sdf.org
___
Discuss mailing list
Discuss@blu.org
http://lists.blu.org/mailman/listinfo/discuss

Re: [Discuss] Post issue with bootable linux

[Mike Small]

Jerry Feldman <gaf.li...@gmail.com> writes:

> I have a guy with an older system
> Core2 duo CPU
> 4GB ddr2 scramble
>
> Dell motherboard
> The problem is when I boot with a known good bootable Linux usb

Known good bootable on BIOS only systems too, or would you have always
UEFI booted this stick before?

> Ubuntu, Fedora the system fails to post.

Fails to post as in doesn't get past the POST? Any beeps? If it's really
failing during or before POST that would rule out anything software
related, from boot loader on down.

> 1. Dell logo comes up
> 2. Press F12 (boot) or F2 (setup) the light flashes on the usb and the
> system is otherwise frozen.
> 3. Neither the setup nor boot menu come up
>
> However, when I use a bootable gparted USB it comes up fine. When I go to
> the boot menu and select the usb, it boots.

This would contradict it being a POST or pre-POST failure unless it was
by chance.

I dunno, I've had an old laptop whose keyboard would fail unpredictably
in ways that seemed to be related to total power draw. It wouldn't get
far trying to run Windows 2000, but OpenBSD or Slackware with a
lightweight wm would run for hours at a time with all or almost all the
keys working. I supposed that some component or circuit had gotten where
voltages were in some ambiguous area that was somehow a function of
current or else heat was changing physical distances between parts in
tricky little ways that started to matter over time.  Maybe your
friend's machine has some kind of power budget issue that it's getting
crankier about over time, and one of your USB sticks draws current a bit
differently than the other.

-- 
Mike Small
sma...@sdf.org
___
Discuss mailing list
Discuss@blu.org
http://lists.blu.org/mailman/listinfo/discuss

Re: [Discuss] Phone maker settles charges it let partner collect customers' text messages

[Mike Small]

"Greg Rundlett (freephile)" <g...@freephile.com> writes:

> On Tue, May 1, 2018 at 9:57 AM, Bill Horne <b...@horne.net> wrote:
>
> [snip]
>
> Phone maker settles charges it let partner collect customers' text
>> messages
>>
>> BLU phones sent a massive amount of data to firmware and data-mining
>> provider.
>>
>> https://arstechnica.com/tech-policy/2018/04/phone-maker-
>> settles-charges-it-let-partner-collect-customers-text-messages/
...
> Generals, and organizations like "Consumer Reports" should join the FSF and
> EFF be issuing statements about the need for consumers to seek out
> completely free software alternatives that respect privacy
> because “Sunlight is said to be the best of disinfectants.” – Louis Brandeis

What alternatives do you like best or find most promising for phones?
These are the efforts I've been following off and on:

1. LineageOS (successor to CyanogenMod)
   - doesn't support my model yet. An Android clone. Perhaps the most
   reasonable first point of entry for an alternative phone OS?
   
2. Replicant - similar to above but more conscientious about everything
   being free software. Kind of the Trisquel of Phone OSes.

Options 1. and 2. are inferior options in my opinion. I'd prefer not to
have Android at all, neither it's one language development model nor its
massive monolith build approach.

3. Debian: https://wiki.debian.org/Mobile
4. PostMarketOS:  https://www.postmarketos.org/

Something like 3. or 4. seems like the holy grail to me.  You'd have a
proper GNU/Linux distro on the thing and be able to run the usual Unix
software without dealing with Bionic's incompleteness or other
stangeness about the middleware/userland parts of Android. But I'm
guessing Debian on any old phone is a long way off and may never reach
some models.

There are companies working on some nice things, but in principle I'm
against buying dedicated free software hardware while I have working
hardware already. It's nice that there are people working on such things
and people willing to put their money down supporting that work, but one
important part of what I want from a truly free phone OS would be that
it allow mopping up all the old e-junk and putting it to good use (or at
least support or be close to supporting what I have already), as
GNU/Linux and the BSDs do so well with PCs and some other odd old junk.

For now I'm making do on my Android phone with termux (a debian based
chroot that doesn't require rooting) and gradually learning about it and
Android's limitations by trying to install all the dependencies of the
Perl module WWW::YouTube::Download using cpan (stuck for now on
Net::SSLeay). It's fun and interesting but also makes me wish I had the
real thing on there.

-- 
Mike Small
sma...@sdf.org
___
Discuss mailing list
Discuss@blu.org
http://lists.blu.org/mailman/listinfo/discuss

Re: [Discuss] Echos from the fediverse

[Mike Small]

There's a Mastodon instance on SDF I was debating joining, but I'm
holding off for now.

"Greg Rundlett (freephile)"  writes:

> Thanks for the nudge Eric, I've been meaning to join Mastodon since
> LibrePlanet. I'm @freephile@mastodon.technology
>
> Greg Rundlett
> https://eQuality-Tech.com
> https://freephile.org
> ___
> Discuss mailing list
> Discuss@blu.org
> http://lists.blu.org/mailman/listinfo/discuss
>
___
Discuss mailing list
Discuss@blu.org
http://lists.blu.org/mailman/listinfo/discuss

Re: [Discuss] systemd reboot

[Mike Small]

Richard Pieri <richard.pi...@gmail.com> writes:

> The old "sync;sync;sync;halt" mantra is folklore from the days before we
> had a shutdown/reboot command which does this for us. The first sync
> flushes any dirty buffers, the second blocks waiting for the first to
> complete ensuring that there are no dirty buffers when the system goes
> down, and the third... makes us feel good (it has no technical benefit).
>
> This doesn't work as expected today because most drives lie about
> committing writes to permanent storage. The second sync won't block
> unless the size of data in dirty kernel buffers exceeds the drives'
> write cache capacity and then it will block only long enough for that
> ratio to flip. If the system restarts, loses power, whatever, when the
> drives' on-board caches have not been committed then there will be data
> loss. The Linux kernel code which guarantees that writes are committed
> doesn't actually work because it relies on drives not lying about their
> cache commits.

Thanks, this is great info. Curiously, in my local reproduction of the
issue the lying hardware involved is qemu's virtio simulated disk. So
maybe their simulation is super realistic, eh?

-- 
Mike Small
sma...@sdf.org
___
Discuss mailing list
Discuss@blu.org
http://lists.blu.org/mailman/listinfo/discuss

Re: [Discuss] systemd reboot

[Mike Small]

jbk <j...@kjkelra.com> writes:

> On 03/03/2018 08:20 AM, Richard Pieri wrote:
>> On 3/2/2018 9:09 PM, Mike Small wrote:
>>> I see behaviour where if I change something under /etc/grub.d/, run
>>> update-grub and then immediately run /sbin/reboot, upon start up grub
>>> sees the old grub.cfg not the new one. This is a Ubuntu Xenial based
>> I don't think systemd has anything to do with it. My guess is that you
>> have more than one /boot/grub on the system (perhaps a replica, perhaps
>> a dual-boot system), possibly more than one grub2 installed, and the
>> active loader is reading from one of those alternate /boot/grub points.
>>
> I'd have to agree with Rich that it is something to do with the path
> to the active grub.cfg.
>
> On Fedora I use this command to effect grub updates:
>
> grub2-mkconfig --output=/boot/grub2/grub.cfg

update-grub does exactly that.

>
> The update-grub command you are using is probably a plain text script
> in /bin or /sbin that issues the same as above.
>
> The other place you might look is /var/log/grubby which on my system
> is a record of every manual or scripted update of
> /boot/grub(2)/grub.cfg.
>
> Look at the UUID's for the root partition it is pointing to if you
> multi boot distro's or versions.

Multiboot isn't involved.

-- 
Mike Small
sma...@sdf.org
___
Discuss mailing list
Discuss@blu.org
http://lists.blu.org/mailman/listinfo/discuss

Re: [Discuss] systemd reboot

[Mike Small]

Steve Litt <sl...@troubleshooters.com> writes:

> On Sat, 03 Mar 2018 02:09:21 +
>> I see behaviour where if I change something under /etc/grub.d/, run
>> update-grub and then immediately run /sbin/reboot, upon start up grub
>> sees the old grub.cfg not the new one. 
>
> How do you know that? Did you back up the old grub.cfg, run update-grub
> and reboot, and then do a diff on the current and backed up grub.cfg
> and find them the same? If not, perhaps there are other reasons
> alterations of contents of /etc/grub.d don't appear to be recognized.

This is what made me think shutdown write cache syncing was an issue:

- what's involved is a script that updates a file under /etc/grub.d with
a new menu entry, runs update-grub and sets the new entry as the default
using grub-reboot.
- Intermittently (on some runs and not others, but usually within 4 or 5
tries, if I have my environment just so) the grub.cfg on reboot lacks
that menu entry. I.e. it looks like it did before.
- a debugging message after update-grub is run shows an md5sum for
/boot/grub/grub.cfg different from that which I see for the file after
rebooting.
- when run with additional debugging messages added later in the script the
problem no longer happens. i.e. grub.cfg will have the new entry and
across 90 tries.
- making the script run sync(1) just before it runs reboot similarly makes the
problem go away.
- this script had previously run on a Ubuntu Trusty derived system (but
with sysvinit used in place of upstart) without this issue as far as I
know.

It's curious though, systemctl does seem to call sync(2). When run as
the reboot symlink it eventually hits this code path:

static int halt_now(enum action a) {
int r;

/* The kernel will automaticall flush ATA disks and suchlike
 * on reboot(), but the file systems need to be synce'd
 * explicitly in advance. */
if (!arg_no_sync)
(void) sync();
...

It's not quite clear to me yet how arg_no_sync gets set. The code
suggests -n but the man page says differently. Regardless, I'm not seeing
how running reboot gets systemctl any extra args.

Or at least that's how it is in systemd 232 (what's in Debian
Stretch). At work I would have a different version perhaps, though I
think Xenial and Stretch are closely related.

I had read sync(2) on some systems only queues up the writes and
returns, but Linux seems to give stronger guarantees:

  According  to  the  standard specification (e.g., POSIX.1-2001), sync()
  schedules the writes, but may return before the actual writing is done.
  However  Linux  waits  for I/O completions, and thus sync() or syncfs()
  provide the same guarantees as fsync called on every file in the system
  or filesystem respectively.

So it's all a bit puzzling.  sync; sync; is better than sync? Or what,
should I do what was suggested in olden time and sync; sync; sync?

-- 
Mike Small
sma...@sdf.org
___
Discuss mailing list
Discuss@blu.org
http://lists.blu.org/mailman/listinfo/discuss

[Discuss] systemd reboot

[Mike Small]

I should be able to find my answers in the man pages or other systemd
documentation but I'm getting fatigued reading them. So if you have the
time...


I see behaviour where if I change something under /etc/grub.d/, run
update-grub and then immediately run /sbin/reboot, upon start up grub
sees the old grub.cfg not the new one. This is a Ubuntu Xenial based
system where /sbin/reboot is a symlink to systemctl. I'm not at the
machine but I think /boot is on the root filesystem instead of having
its own partition. Questions...

1. Presumably reboot should itself or via systemd unit dependencies sync
buffer cache to disk or at least try to. E.g. on OpenBSD the docs are
straightforward, from reboot(8): "The halt and reboot utilities flush
the file system cache to disk, execute the rc.d(8) scripts specified by
the pkg_scripts variable defined in rc.conf(8) in a reverse order, run
the system shutdown script, send all running processes a SIGTERM (and
subsequently a SIGKILL), and, respectively, halt or restart the system."
So what are the assurances her for Linux with systemd, and is there a
way it can go wrong?

2. What is the mechanism? So far I'm thinking it's not in reboot,
a.k.a. systemctl itself.  I see a dependency graph at the end of
bootup(7) that suggests systemd-reboot.service precedes the
reboot.target. Is that what's responsible for the syncing?

3. systemd-reboot.service(7) (or maybe it's systemd-halt.service(7), I'm
looking online at html and the labeling is poor) says, "immediately
before executing the actual system halt/poweroff/reboot/kexec
systemd-shutdown will run all executables in
/usr/lib/systemd/system-shutdown/" So that maybe is where syncing could
happen? If when I'm back at the machine I look in there I might expect
to see script or executables that do this, or if not the machine is
incorrectly configured?

4. Somewhere I thought I read something to the effect that the root file
system was a special case. I can't find it now, but I thought I read
that systemd won't unmount it (but maybe could sync it?) but instead
will leave that to initrd scripts that it will return to when systemd
terminates. Does that even make sense? If so can you think where I might
have read this?

5. /sbin/reboot is a symlink to systemctl. Does that mean running it is
equivalent to running systemctl with a certain argument, e.g. reboot?
I've somehow missed where this is documented. There seem to be various
ways to run systemctl to do a reboot: systemctl reboot or systemctl
start reboot.target with varous args. systemctl(1) says reassuring
things under its reboot command so long as you don't give the force
argument twice, but I can only guess or read the source code to know if
that's what running the reboot symlink does. That section of the man
page also says something alarming about a command being asynchronous,
but I'm hoping that means the wall message command and not the reboot
command itself or some part of it like syncing to disk.

- Mike
___
Discuss mailing list
Discuss@blu.org
http://lists.blu.org/mailman/listinfo/discuss

Re: [Discuss] ten more years

[Mike Small]

Kent Borg <kentb...@borg.org> writes:

> On 02/14/2018 02:10 PM, Mike Small wrote:
>> I'm hoping postmarketos (or ideally debian mobile + their arm porting
>> teams) goes somewhere so I don't have to discard my Samsung
>
> I just recently did the Kickstarter, er, Indigogo, for the Gemini PDA.
>
> https://www.indiegogo.com/projects/gemini-pda-android-linux-keyboard-mobile-device-phone#/
>
> I hope it is decent. Donno.
>
> I did not order the phone version, I did wifi-only.
>

It's nice to see them offering a real Linux as a secondary boot
option. If that were an expectation for all such devices we'd be in a
much better place than we are now. But myself I can't get excited about
new hardware and kickstarters like this. Exciting to me rather are the
people trying make all the junk floating around out there, going several
years back, just another device that Linux and one or more of the BSDs
are ported to. The blurbs on postmarketOS's pages capture my feelings
perfectly. But it probably won't happen for a lot of it (e.g. how long
will armel be a supported Debian architecture?).

-- 
Mike Small
sma...@sdf.org
___
Discuss mailing list
Discuss@blu.org
http://lists.blu.org/mailman/listinfo/discuss

Re: [Discuss] ten more years

[Mike Small]

Kent Borg <kentb...@borg.org> writes:
> * Irony alert: a 2013 Nexus 7 shouldn't have to be discarded as
> pathetic. This calibre machine used to be supercomputer territory.

I'm hoping postmarketos (or ideally debian mobile + their arm porting
teams) goes somewhere so I don't have to discard my Samsung (which being
more than 1.5 years old obviously can't be expected to be patched any
further). Or, really, I'm not going to discard it, but I hope one day to
run something that will make it slightly less like carrying windows 98
in my pocket but with wifi.

Anyone have any general advice on how to start learning the fundamentals
you need to start porting non-Android Linux to a smart phone? My model
isn't among those that LineageOS has a port for. That seems to be the
starting point for all such efforts. Without that I'm not sure where to
begin.

Even just a book about the architecture of android based phones might be
nice purely for interest sake. Is there a good reference? Every book
I've looked at is just how to program Android Apps. Is there nothing
that describes a typical device architecture (or maybe there's not much
common to the various devices?) the way there used to be for the
PC. I.e. something that describes the bus topology, the interrupts,
etc. Oh, and good information about device start up and the boot loaders
used or available, probably would need that. BIOS or UEFI in the PC is
to _ on a phone, for instance? 

It's such a tease having Android on a phone, it being almost like a real
GNU/Linux environment. E.g. I'd assumed the other day I could easily
install youtube-download (Perl) or youtube-dl (Python) under Termux, but
no such luck, not in their apt repository. Then I try cpan and that
seems not to work at all under Termux. Blah.

-- 
Mike Small
sma...@sdf.org
___
Discuss mailing list
Discuss@blu.org
http://lists.blu.org/mailman/listinfo/discuss

Re: [Discuss] node.js and npm on Debian?

[Mike Small]

Kent Borg <kentb...@borg.org> writes:

> [Warning, philosophical tangent.]
>
> The thing that strikes me is that it feels like we are doing something
> fundamentally wrong if a sensible way to write "Hello, World!" (or
> "Hello, I'm a GUI widget!") could possibly be: "Fire up a Linux
> VM.". (Or a Docker container...)
>
> I'm not saying the answers in this thread are wrong (I appreciate
> them, I have learned a lot, and I think I even have npm
> installed--though an unrelated regression stalled my node.js playing
> today). I am saying the so-called "full stack" that is trendy these
> days smells bad and feels wrong headed.
>
> The binary for a modern-day IRC-type program (Slack) is over
> 80MB. Sure, the original IRC didn't have pictures. But 80MB!? I have
> an internet radio program (Tunein Radio) that has an install of 65MB.
>
> The Linux kernel--arguably festooned with too many features--is only a
> 4MB binary on my notebook. And can be trimmed down to about half that
> (last I looked) yet still do powerful stuff. It is crazy that Slack is
> twenty-times the size of the big version of the kernel.
>
> You might ask: "What's the harm? Storage is cheap!"
>
> I'm not sure, but it feels like this is an infinite supply of security
> vulnerabilities, plus a lot of regular feature bugs and much of the
> general "computerized things don't work very well".

For the specific issue of the bloatification of Unix software there's
suckless: https://suckless.org/

Not sure they have a substitute for node.js, though. And I've found for
some of what they offer they're a bit too reductionist. My sense is that
what OpenBSD packages sucks less than suckless. Better sense of balance.

There are people still trying to make some new start out of plan 9. And
then there are still some people, I guess, making Smalltalk systems
(Pharo?) and probably some people somewhere trying something with
Lisp. I wonder if Linux today is ripe to be supplanted by something
old but new that catches people's imaginations the way Linux did when it
was a hobby and Windows was the job.

Or we could, for therapy, try to get Oberon running in a vm when we're
feeling down. Actually, I'm not even sure it necessarily even still
needs a vm. There are enthusiasts playing with that too.

-- 
Mike Small
sma...@sdf.org
___
Discuss mailing list
Discuss@blu.org
http://lists.blu.org/mailman/listinfo/discuss

Re: [Discuss] node.js and npm on Debian?

[Mike Small]

Mike Small <sma...@sdf.org> writes:

> Dan Ritter <d...@randomstring.org> writes:
>
>> On Tue, Feb 13, 2018 at 10:51:41AM -0800, Rich Braun wrote:
>>> Kent Borg <kentb...@borg.org> asks:
>>> > But I can't figure out how to install npm. When I search for
>>> > installation instructions they all seem to want me to pipe a curl
>>> > command into a sudo bash. Huh? That's scary as hell.
>>> 
>>> Let others do the installation for you: my go-to technology for this is
>>> Docker. First get docker installed
>>
>> And transfers those headaches to your security and ops teams.
>>
>
> Would Nix or Guix (run from within Debian -- guessing the former would

In case you were thinking of this as a random suggestion from an
ignoramus (which would be entirely appropriate) here's a FOSDEM talk on
the subject:
https://archive.fosdem.org/2017/schedule/event/deploying_npm_packages_with_nix/

-- 
Mike Small
sma...@sdf.org
___
Discuss mailing list
Discuss@blu.org
http://lists.blu.org/mailman/listinfo/discuss

Re: [Discuss] node.js and npm on Debian?

[Mike Small]

Dan Ritter <d...@randomstring.org> writes:

> On Tue, Feb 13, 2018 at 10:51:41AM -0800, Rich Braun wrote:
>> Kent Borg <kentb...@borg.org> asks:
>> > But I can't figure out how to install npm. When I search for
>> > installation instructions they all seem to want me to pipe a curl
>> > command into a sudo bash. Huh? That's scary as hell.
>> 
>> Let others do the installation for you: my go-to technology for this is
>> Docker. First get docker installed
>
> And transfers those headaches to your security and ops teams.
>

Would Nix or Guix (run from within Debian -- guessing the former would
be more likely to have npm) be a better substitute? This was one of the
issues mentioned against things like Docker when we had that Guix talk a
couple years back. I have no first hand experience myself, only curious.

In fact I started to try to install Guix this weekend and ran into the
issue that Stretch doesn't have gnutls's guile bindings. I looked at the
history there, and it seems some gnutls and guile test failed. The Guix
maintainer, who also is a big guile hacker, looked at it and it appeared
to be something deep in garbage collection, so I have no harsh words for
debian -- they did what they day to do. But it was a little fustrating,
given I'd finally gotten around to actually wanting to try guix after
noticing CLISP didn't make Stretch due to a little slip up not fixing
something in time for the release. So maybe that's how it is with
Debian? You either have Woody/Sarge-like release cycles or you have
releases missing certain things.

(It does look like it will be simple enough for me to make a local
version of gnutls with guile-gnutls, except last I was at it I hit some
issue in dh_makeshlibs I don't understand yet. Is there an easier to
understand reference to shlibs in regards to debian packaging than their
policy document?)

-- 
Mike Small
sma...@sdf.org
___
Discuss mailing list
Discuss@blu.org
http://lists.blu.org/mailman/listinfo/discuss

Re: [Discuss] docker Re: Corralling Processes on Linux

[Mike Small]

"Rich Braun" <ri...@pioneer.ci.net> writes:

> Kent Borg <kentb...@borg.org> wrote:
>>> I am playing with lots of different processes
>>> communicating with each other, maybe some coming and going
>>> incrementally. I want the ability occasionally kill them all and
>>> start from a clean slate.
>
> Sure sounds like what you really want is Docker and/or Kubernetes. Cgroups is
> part of the mechanism used by containers (such as the original LXC) to isolate
> process groups but there's a whole open-source infrastructure that provide the
> tools that abstract out a lot of the difficult parts of what you want to do.

At what point does it make sense to go to the cgroup level or even
container level and at what point are traditional Unix abstractions like
process groups and sessions adequate?  If Kent is creating all the
processes himself and they all fall in one process group then kill(2) on
the negative of the process group leader should kill them all. To me
that would be simpler (but not necessarily easier? I have much to learn
about container tech.) than working with these higher level abstractions
and toolkits. If plain old process groups are adequate he also gets the
benefit that his software might run on a BSD as is.

-- 
Mike Small
sma...@sdf.org
___
Discuss mailing list
Discuss@blu.org
http://lists.blu.org/mailman/listinfo/discuss

Re: [Discuss] Things in Hidden, Magic . Files

[Mike Small]

Kent Borg <kentb...@borg.org> writes:

> On 01/17/2018 12:31 PM, Kent Borg wrote:
>> It has its advantages, it has its drawbacks, but then I stumbled on
>> a *big* problem: source code debugging with pdb quit working. Emacs
>> used to open another pane and put an arrow next to the line about to
>> be executed. No longer. Naturally I suspect elpy.
>
> I think I have a workaround...
>
> I usually invoke pdb by dropping in my code:
>
>   import pdb ; pdb.set_trace()
>
> But I tried meta-X pdb and got an error, something about a string
> being nil? (I didn't write it down.) I tried tracing in emacs...and
> that make it work. I reboot and "import pdb ; pdb.set_trace()" fails,
> but meta-X 
> pdb works...but thereafter it seems "import pdb ; pdb.set_trace()" now
> works.
>
> Weird.

So emacs has two debuggers. Even if you don't want to learn edebug
(which is too bad because it's my favourite debugger -- seriously, I'd
rather write elisp than common lisp or scheme cause I know no debugger
that compares to edebug -- way better than gdb and Perl's debugger), you
could at least get a stack trace by setting debug-on-error to true
before you reproduce the issue.

What you describe sounds like a module not loading early
enough. Something loads when you m-X pdb because of an autoload or
whatever, something that the other way can't make happen since that way
is only between, uh, what?, your python interpreter and the
comint/gud/gdb buffer.

-- 
Mike Small
sma...@sdf.org
___
Discuss mailing list
Discuss@blu.org
http://lists.blu.org/mailman/listinfo/discuss

Re: [Discuss] Things in Hidden, Magic . Files

[Mike Small]

Did you do an apt purge? I wouldn't think emacs package source files
would be considered configuration but who knows.

As an aside, with emacs packages I'd be inclined not to use the distro
repo. It's unlikely you'd have a distro package that depends on an elisp
package outside of emacs's guts, so using emacs's native package manager
won't cause you the trouble using Perl's or Python's might (without some
wrapper thing insulating you from the distro). IMO the native
interpreter/language packager would always be better to use and distros
should stay out of it, except for that pesky problem that distro
software might also need some version of the modules you use.

What about using emacs's own symbolic debugger to step through the elpy
elisp code until things get weird? It's pretty easy to get started
with. Only a handful of key presses to learn until you can do the basics
with it.

The emacs manual section "The Init File, `~/.emacs'" describes what
emacs loads on start up. The load-path variable shows the places elisp
files can come from (or most of the places? maybe some of the init files
or package manager stuff isn't in that list). That section also shows how
to start emacs without loading the usual init files. You could try that
and manually load the things you need for a basic Python debugging
session. Then if that works rebuild the death star from the inside out
until the weirdness reasserts.

Kent Borg <kentb...@borg.org> writes:

> Short version: Where are all the places to put things to change
> emacs's behavior?
>
> Longer version: I am an emacs person, programming in Python mostly
> these days. So the other day I thought I should get modern (always a
> risk) and get emacs to do clever tab-completion and stuff. I tried
> installing elpy.
>
> It has its advantages, it has its drawbacks, but then I stumbled on a
> *big* problem: source code debugging with pdb quit working. Emacs used
> to open another pane and put an arrow next to the line about to be
> executed. No longer. Naturally I suspect elpy.
>
> So I did my best to uninstall it. And uninstall the apt-get stuff I
> installed to make elpy happy.
>
> My source code debugging is still not fixed.
>
> I try debugging as another user and that does work. I try hiding my
> .emacs file and .emacs.d directory in another directory, and it is
> still not working.
>
> Running Debian 9.
>
> Ideas?
>
> Thanks,
>
> -kb
> ___
> Discuss mailing list
> Discuss@blu.org
> http://lists.blu.org/mailman/listinfo/discuss
>

-- 
Mike Small
sma...@sdf.org
___
Discuss mailing list
Discuss@blu.org
http://lists.blu.org/mailman/listinfo/discuss

Re: [Discuss] kmplayer

[Mike Small]

dan moylan <j...@moylan.us> writes:

> i downloaded kmplayer in an attempt to listen to a file sent
> by my son of his new bagpipe music.  when i tried to use it
> i had odd results:
>
>   moylan 2017[1039] kmplayer pom-test.wma
>
>   KGlobal::locale(): Warning your global KLocale is being
>   recreated with a valid main component instead of a fake
>   component, this usually means you tried to call i18n related
>   functions before your main component was created. You should
>   not do that since it most likely will not work "sni-qt/1831"
>   WARN  14:28:53.794 void
>   StatusNotifierItemFactory::connectToSnw() Invalid interface
>   to SNW_SERVICE kphononplayer -cb :1.915/master_0
>
>   Terminated
>
> the computer locked up and i was only able to recover by
> ssh-ing in from another and killing kmplayer.  i don't
> understand the warning i received.
>
> any suggestions?

Run mplayer directly rather than this seemingly broken attempt at
wrapping it in kde stuff. Or maybe run mpv instead. Some prefer that to
mplayer. It has a nice feature where you can keep track of where you
left off in a movie by pressing `Q' instead of `q' to exit, which I
don't think mplayer has implemented yet.

I've never understood the value of a fancy UI for a video player when
you're usually going to want to watch your movie full screen.

-- 
Mike Small
sma...@sdf.org
___
Discuss mailing list
Discuss@blu.org
http://lists.blu.org/mailman/listinfo/discuss

Re: [Discuss] php question. change directory to executing script.

[Mike Small]

Eric Chadbourne <sillystr...@protonmail.com> writes:

> Hi All,
>
> I have a little function that needs to delete some files.  I don't
> want it to delete them from the directory of my user, but in the
> directory of the script.
>
> For example from terminal:
>
> cd $HOME
> php /path/to/script/foo.php  // will execute foo in home
>
> But I want it to execute foo in /script.
>
> So in the function I told it to change working directory to script path.
>
> chdir(dirname(__FILE__));
>
> Am I missing something or is this OK?  Is it evil to do such a thing?

Could your script ever be a symlink? Seems to me I've seen this go wrong
or at least get confusing (yes, it seems not uncommon to want to run
things in the script's directory - each place I've worked at the last 20
years has done it, both under Linux and under Windows) when symlinks are
involved.

What about passing an argument with the directory to execute in. Then at
least you have outside control. It's kind of the moral equivalent of a
local (or really a function argument) vs. global variable, in the sense
that your script's directory is like a global in that it comes from the
surrounding environment and isn't stated explicitly vs. a command
argument which would be more like (and end up in) a local variable.

It it's clumsy to run that way you can always use an alias.

On the other hand if its your own thing make it as simple as works for
you. Maybe you don't care about symlinks, and it's small enough that
"global variables" are appropriate. I just recall at the Windows job
when the program got big new programmers were a bit puzzled about which
directory it ran in and where it found its configuration files. It was
yet another bit of annoying incidental knowledge to be passed on that
the "working directory" wasn't the normal working directory but the
program's directory, or subdirectories relative to that.

-- 
Mike Small
sma...@sdf.org
___
Discuss mailing list
Discuss@blu.org
http://lists.blu.org/mailman/listinfo/discuss

Re: [Discuss] LibreOffice and .docx files

[Mike Small]

Steve Litt <sl...@troubleshooters.com> writes:

> Mike Small mentioned
> http://obsd.si/papers/eurobsdcon2014-mandoc-paper.pdf , which is an
> improved man page authoring tool. From what I learned skimming the
> link, mandoc has been created exclusively to create man pages, to the
> extent that its styles have a 1 to 1 correspondance with the semantics
> of a man page. This is great for man pages, but it would probably be
> easier to complete Stylz or use Docbook/XSLT than to broaden the usage
> of the very specific mandoc.

That's true. Their focus is man pages.

However, the author does sometimes express opinions on Docbook and the
minimarkup formats and has worked quite a bit in this area. I didn't
link to his more opinionated posts before, thinking his paper on mandoc
a better, non-controversial intro to his ideas, but here are some
below. As someone working in this area yourself, you may find his
thoughts useful for comparison to your own:

https://lists.gnu.org/archive/html/groff/2014-02/msg00106.html
https://lists.gnu.org/archive/html/groff/2014-02/msg00109.html

-- 
Mike Small
sma...@sdf.org
___
Discuss mailing list
Discuss@blu.org
http://lists.blu.org/mailman/listinfo/discuss

Re: [Discuss] LibreOffice and .docx files

[Mike Small]

Theodore Ruegsegger <grun...@gmail.com> writes:

> On Tue, Dec 12, 2017 at 2:13 PM, Steve Litt wrote:
>> [Rationale for separating document content from presentation, as embodied in 
>> Stylz]
>
> Isn't this the raison-d'etre for DocBook?
>
> Last I looked, that seemed to be working pretty well, though it was
> difficult to find tutorials that didn't lose me in short order.
>
> On the other hand, for the relatively simple documents I write from
> day to day, asciidoc is insanely easy to manage, and converts into
> TeX, HTML, PDF, or whatever with the flip of an option or two.
>
> How does Stylz differ from DocBook and its derivatives?
>
> This won't help the OP with her problem, but might be of interest to others.
>

I'm not up on this stuff, since everything I write now is without style
in every sense of the word, but some might find Inge Schwartze of
OpenBSD's opinions on this topic interesting. He's focused on the
problem of software documentation, but there's some overlap with what
you're discussing:

http://obsd.si/papers/eurobsdcon2014-mandoc-paper.pdf

-- 
Mike Small
sma...@sdf.org
___
Discuss mailing list
Discuss@blu.org
http://lists.blu.org/mailman/listinfo/discuss

Re: [Discuss] LibreOffice and .docx files

[Mike Small]

Nancy Allison <nancythewrit...@gmail.com> writes:

> Hello, all.
>
> I painfully conclude that I must install Word on my Ubuntu 16.04 machine. I
> need to be able to handle Word files consistently.
>
> So I opened Synaptic Package Manager and searched for Microsoft Word. No
> luck.
>
> What do I do? How do I get it installed?
>
> I know, it pains me! But I don't see a way around it. All suggestions
> gratefully received!
>
> --Nancy
>

Okay, you said "all suggestions," remember that.

So here goes: The sooner you (and whatever network of people forcing
your hand here) can migrate away from Microsoft Word AND Libreoffice or
Openoffice and over to something like LaTeX the better off you will
be. You'll need to learn LaTeX of course (a wrapper for it named Lyx may
help smooth the learning curve, and there are various quite well written
books about it) but once you master it you should be much happier and
won't have to deal with these sorts of headaches, or with a number of
other headaches intrinsic to the PC wordprocessing WYSIWIG
approach. That was kind of a temporary abberration of the 20th century,
like Reaganomics, urban renewal or the cold war. WYSIWIG word processors
and their free software clones belong in a museum of popular culture
next to the hoola hoop, the Apple II and Barry Manilow albums.

-- 
Mike Small
sma...@sdf.org
___
Discuss mailing list
Discuss@blu.org
http://lists.blu.org/mailman/listinfo/discuss

Re: [Discuss] Question regarding parking at MIT for BLU meetings

[Mike Small]

John Abreau <abre...@gmail.com> writes:

> We've had unusually low turnout at the past couple of BLU meetings, and I'm
> wondering how much of it is due to the changes in MIT's parking policy.
>
> How many of you have stopped attending BLU meetings because of the new
> parking restrictions?

For me parking is never an issue as I only have my behind to park but
this was a month when Boston Perl Mongers and BLU had the same week, so
I was less inclined to listen to a second computer related talk within
two days over going to an odd Chinese movie at the Brattle and listening
to a talk about James Madison at Harvard Coop. Also, no offense meant to
the speaker or arrangers, but 3d printing is probably one of those
topics a free software enthusiast has either already got into big time
or one that he will never have much interest in.

-- 
Mike Small
sma...@sdf.org
___
Discuss mailing list
Discuss@blu.org
http://lists.blu.org/mailman/listinfo/discuss

Re: [Discuss] perl and xml

[Mike Small]

"Rich Braun" <ri...@pioneer.ci.net> writes:
> solutions in other widely-available languages. And yes, I realize perl is just
> as widely available as python; a lot of people don't realize that the converse
> is true (that python is as widely available as perl: they go hand-in-hand like
> vim and bash).

So a python vs. perl war wasn't enough for you? Now you want an emacs
vs. vim one ;)

-- 
Mike Small
sma...@sdf.org
___
Discuss mailing list
Discuss@blu.org
http://lists.blu.org/mailman/listinfo/discuss

Re: [Discuss] Crashplan is discontinued

[Mike Small]

Bill Bogstad <bogs...@pobox.com> writes:

> On Thu, Aug 31, 2017 at 10:02 PM, Mike Small <sma...@sdf.org> wrote:
> Does git only compare the checksum or does it also look at file size as well?
> I would think that comparing file size might make it even harder to
> get a collision.
> The only duplicate checksum that I've ever seen in practice was on 0
> length files.
> Zero length files are, of course, all perfect duplicates of each other... :-)

Ah, git plumbing. Not really my specialty, but I think the answer is
implied by some of the docs, kind of. I'll add some guess work and if
someone knows better he or she can correct me.

Zero length file collisions are not an issue in git because the stuff in
its store (.git/object/{first two letters of SHA1 hash}/{rest of SHA1
hash} includes both the file contents themselves (blobs - check me in
gitglossary(7)) and tree objects which have capture file and directory
names and reference the content blobs. Here's some of my
.emacs.d/.git/objects contents (not a great use of git I'm finding,
btw. I should have done it down at the level where I only have files I
treat as my source code as opposed to stuff emacs changes behind my
back.):

8613r0:.git$ du -a objects/ | head  
 
4   objects/af/2ef3b97a02a0cdc859c59e4d39d6a7aa01116c
4   objects/af/ef5e0daed0ecdf0d51dcc347149ae2e1f0e998
12  objects/af
4   objects/d7/2834524cad924ea210b41920293a6fcc5d72ff
8   objects/d7
4   objects/17/dc6f4f501ce4ee0f3488d246b825d0c3ad63fe
4   objects/17/62e9cf542a661f15351b3bb2c50e1a1d26a1cd
12  objects/17
4   objects/pack
4   objects/8e/709560a5a09f69f8be7665ad66e3c394620123
...

So if I'm understanding rightly you could have 10 zero length files in
git with different names and that's not a problem. You'd have 10 tree
objects in the store, i.e. directories and files matching the SHA1 hash
involved, perhaps that all reference one blob object with a different
SHA1 directory and file name for the contents (or lack thereof).

I think so far I don't see an actual compare, necessarily, just it
creates these tree objects and creates the blob object. Maybe it
overwrites the blob object for each file or maybe it sees it already
exists and just references it, I don't know. Kind of doesn't matter
except for performance or whatever. Or does it?

Let's take the malicious case. You want to get a file into the store
that has the same hash as an existing blob file, so that existing
references now have your contents instead of the original stuff. So
you'd be creating whatever tree object in the store, no hash collision
on that, but you'd want your file blob object to overwrite an existing
one. Unless my guesswork here is totally off I'm going to say git must
simply overwrite a blob file if you succeed in getting a hash
collision. If it did a compare to see if a path with the sha1 number was
already under .git/objects and didn't bother to write the new contents
then a hash collision couldn't be a real vulnerability and there
shouldn't have been a thread discussing it.

But I could be way off here. If you really want to know probably you
want to start by reading gitcore-tutorial(7), gitrepository-layout(7),
and maybe the source of git-hash-object or some other plumbing
command. Oh wait, git-hash-object I see now is a link to git, so you'd
have to read the top of the source which looks at what the execed
filename was, assuming I have indeed picked the right command here. The
plumbing man pages are pretty thin. Maybe higher level commands are
relevant here too.

-- 
Mike Small
sma...@sdf.org
___
Discuss mailing list
Discuss@blu.org
http://lists.blu.org/mailman/listinfo/discuss

Re: [Discuss] Crashplan is discontinued

[Mike Small]

Mike Small <sma...@sdf.org> writes:

> Mike Small <sma...@sdf.org> writes:
>
>> John Abreau <abre...@gmail.com> writes:
>>
>>> I've heard of tools using MD5 or SHA1 hashes to identify duplicates, and
>>> potential issues with hash collisions causing false positives.
>>
>> By accident or maliciously? The numbers seem off for accidental
>> collisions. An md5 sum is a 16 digit hex number. That gives
>
> Typo here. It's a 32 bit hex number.

Grrr, a 32 digit hex number. No more emailing for me tonight.

-- 
Mike Small
sma...@sdf.org
___
Discuss mailing list
Discuss@blu.org
http://lists.blu.org/mailman/listinfo/discuss

Re: [Discuss] Crashplan is discontinued

[Mike Small]

Mike Small <sma...@sdf.org> writes:

> John Abreau <abre...@gmail.com> writes:
>
>> I've heard of tools using MD5 or SHA1 hashes to identify duplicates, and
>> potential issues with hash collisions causing false positives.
>
> By accident or maliciously? The numbers seem off for accidental
> collisions. An md5 sum is a 16 digit hex number. That gives

Typo here. It's a 32 bit hex number.

-- 
Mike Small
sma...@sdf.org
___
Discuss mailing list
Discuss@blu.org
http://lists.blu.org/mailman/listinfo/discuss

Re: [Discuss] Crashplan is discontinued

[Mike Small]

John Abreau <abre...@gmail.com> writes:

> I've heard of tools using MD5 or SHA1 hashes to identify duplicates, and
> potential issues with hash collisions causing false positives.

By accident or maliciously? The numbers seem off for accidental
collisions. An md5 sum is a 16 digit hex number. That gives
340282366920938463463374607431768211456 potential hash sums (or does the
algorithm offer only a smaller subset?). I'm not going to bother to
compute the probability of a collision. It's a very remote possiblity,
yes? For the malicious case, if someone's able to mess with the hashes
used by deduplication code in your file system or in your hopefully
almost as good userland equivalent (which of course must use git in some
way or another for reasons that are not clear to me) you have unsolvable
problems.

I once saw a pointer to a thread speculating about the problem for git
(when actually used for source code, go figure), but the hypothetical
attack needed a hostile committer. It wasn't an easy attack even then,
though that might have been in part from the social engineering
challenges of having other committers not notice what you've
done.

-- 
Mike Small
sma...@sdf.org
___
Discuss mailing list
Discuss@blu.org
http://lists.blu.org/mailman/listinfo/discuss

Re: [Discuss] CrashPlan Home is discontinued - what's next?

[Mike Small]

wor...@alum.mit.edu (Dale R. Worley) writes:

> I have a cron job which commits my home directory into a Git repository
> every minute.  Surprisingly, this puts no noticeable load on the
> computer.

How do you handle file permissions? E.g. .ssh directory contents or PGP
key files having restricted permissions and a git checkout pulling them
out of the repository with more lax default permissions based on your
umask (at least I think that's what it does). IIRC Joey Hess wrote some
kind of tool to use git to track /etc changes and had to add something
on top to deal with permissions and file ownership. I'd think you'd need
to do something similar.

-- 
Mike Small
sma...@sdf.org
___
Discuss mailing list
Discuss@blu.org
http://lists.blu.org/mailman/listinfo/discuss

Re: [Discuss] Eclipses Re: Great talks last night, however...

[Mike Small]

Bill Bogstad <bogs...@pobox.com> writes:
> So lets say that I accept everything you say about both the
> inefficiency and unclean
> characteristics of solar PV + battery storage.   Are the current
> incumbent solutions (Oil, Coal, Natural
> Gas) any better on either characteristic?  When doing your efficiency
> calculations,
> please don't cheat.   i.e. Do total life cycle back to when the
> material was first buried
> underground.  I suspect that even turning corn into ethanol is more energy
> efficient then the process that created fossil fuels.

This book is helpful on this topic:
http://withouthotair.com/

It's a little old, but the way he calculates makes me think his numbers
should hold up pretty well over time. He avoid economics, so I'm
thinking the recent advancements in solar (mostly concerning price per
unit energy?) don't change things. He does concentrate on the picture
from a U.K. perspective, though. It's kind of like a Doctor Who episode,
where the aliens somehow always come to England, but with renewables
instead. Still, people on this list probably can take his calculations
and apply them to Massachusetts easily enough.

I'm wondering maybe for MA if the answer shouldn't be buying lots of
Quebec and Labrador hydro (my old country is not paying me to say this,
honest) to fill in the power troughs left by solar and wind. I mean,
what, you can turn on gas power quickly too, I think, but that's a
greenhouse gas, and, jesus, if we ever leave this era of low natural gas
prices anyone with a few panels up will quickly find a way to appreciate
every last kwh they generate even if it's not coming quite at the right
time. What's NE ISO's mix at now, like 60% natural gas? And the next
proposed major step (with the first being largely taking advantage of
gas prices dropping to be less than coal as I understand it? or perhaps
that's unfair) for MA's climate change policy is to switch to electric
cars. Yikes.

Nuclear still sounds like a needed thing, but I have a hard time
imagining a new plant around here anytime soon, and the ones we have are
winding down. More reason to fear a gas hike, not to mention the
difficulty of building sufficient renewables quickly enough, their
disappointingly low power generation per unit land area numbers, or
their intermittency.

And then there's stuff like this:
http://www.bostonglobe.com/news/nation/2017/07/20/dakota-access-developer-new-pipeline-rankling-regulators/LpMzzvtTpFT3KJH6wb7WJO/story.html

-- 
Mike Small
sma...@sdf.org
___
Discuss mailing list
Discuss@blu.org
http://lists.blu.org/mailman/listinfo/discuss

Re: [Discuss] Docker and Own^h^h^h Nextcloud

[Mike Small]

Rich Braun <ri...@pioneer.ci.net> writes:
> never trusted in the first place). No fear of loss. And they can be shared
> online with friends the same way one would share a Dropbox link. Voila, the
> whole vacation photo-album problem is solved without having to resort to lame
> Facebook or Google Groups solutions that I've always found pretty tedious.

Well that's a feature I could use. I was going to make a website for all
my pictures but this would be less work if there's not much needed from
others to get them. It's a high bar though. My brother-in-law has long
since given up trying to get people to follow Dropbox links to photos and
home videos, for instance.

-- 
Mike Small
sma...@sdf.org
___
Discuss mailing list
Discuss@blu.org
http://lists.blu.org/mailman/listinfo/discuss

Re: [Discuss] Docker and Own^h^h^h Nextcloud

[Mike Small]

David Kramer <da...@thekramers.net> writes:

> I have a family. But if I didn't, there's still the problem of
> multiple computers and Android devices.  rsync is great for keeping
> two computers up to date but what if you have a desktop and a laptop
> and a tablet and a phone and there are some documents you want to
> read/write on all of those devices?  At the very least I would use
> unison for bidirectional syncing, not rsync.

I guess that could be handy. Personally if I had documents like that I'd
put them in the one area I control that no one ever turns off -- that
being my home directory on SDF's disk cluster -- and then I'd get it
from there, maybe looking into sshfs if getting at them in emacs (via
its tramp emacs lisp package, which, if you're unfamiliar with it, is
something like gnome vfs) isn't adequate.  Auto-syncing might be
something I want at some point, for my music files and family
photos/videos especially, but for the near future the storage sizes are
so different on my three devices that I need to think a bit about what
gets copied where, so something more manual or custom written makes more
sense for me. Besides, I could use the practice at writing shell scripts
and should learn more about udev, which is all probably more fun than
whatever I'd need to learn to get going with nextcloud.

Maybe I'd be more curious about it if nextcloud wrapped bittorrent in an
easy way others would like, letting me use my friends' and relatives'
computers to share, backup, and distribute pictures and videos of my
son, done in a way where no central server (e.g. SDF) gets hit too
much. Then again I'm not sure any of these people leave their computers
on that long, except their phones and the storage on those isn't
adequate yet, nor could I assume people would have the kind of data plan
where they'd want to participate. On that topic, can you use non-web
services like bittorrent, ssh and tor okay on t-mobile's network?  Was
wondering how general an ISP they are and if I'd be happy dropping RCN
in favour of only using mobile hotspotting for home networking. My
available network capacity is about 10 times what I seem to need, so I
should buy less of it, to the degree there exist product options between
too much and none at all.

I had started to write something about Android devices in my first
email, since I just got my first one a month or so ago, but thought I
shouldn't take things on a tangent with my beginner questions (so feel
free to skip this paragraph). Android seems to have some annoying
issues. Anyone successfully run an alternative like Replicant or
something Mer based?  SailfishOS sounds kind of nice. Some of it may be
solveable by rooting the phone but because of some software Samsung puts
on this model, something called Knox, it sounds as if that's not such a
good idea. According to threads online Samsung does some kind of remote
attestation on the system software and will set a hardware flag, in a
way that can't be unset, if you do certain things like root it or
install another operating system. Oh, and btw. how did it come to be
that phone operating systems are called firmwares or even ROMs?  That
strikes me as a massive piece of manipulative b.s. by someone or other,
not to mention this term "sideloading." Can we thank Apple for this
terminology?

...

> On 06/19/2017 04:03 PM, Mike Small wrote:
>> "Rich Braun" <ri...@pioneer.ci.net> writes:
>> ...
>>> Even with all that, though, this looks like something I should've pursued 
>>> back
>>> in 2013 when I first heard the software title Owncloud. If you've got trust
>>> issues and "don't love the Cloud", read about it!
>> SDF (the shell provider not the Syrian resistance group) offers owncloud
>> or maybe its successor, but I haven't yet had a problem I thought it
>> would be the solution for, so haven't looked into it. The trouble I have
>> is that the only people I can think of I'd share files with this way
>> would only go along with it if I used Dropbox instead. It's a similar
>> kind of problem to what prevents me from ever encrypting any of my
>> email.
>>
>> For files I don't share but sync between machines I figured rsync (using
>> rsync directly? do Dropbox and owncloud use rsync under the covers? do
>> the rsync authors ever get a $ or stock tip for their work backing
>> Dropbox if so?) has a richer set of options and is more flexible.
>>

-- 
Mike Small
sma...@sdf.org
___
Discuss mailing list
Discuss@blu.org
http://lists.blu.org/mailman/listinfo/discuss

Re: [Discuss] Docker and Own^h^h^h Nextcloud

[Mike Small]

"Rich Braun" <ri...@pioneer.ci.net> writes:
...
> Even with all that, though, this looks like something I should've pursued back
> in 2013 when I first heard the software title Owncloud. If you've got trust
> issues and "don't love the Cloud", read about it!

SDF (the shell provider not the Syrian resistance group) offers owncloud
or maybe its successor, but I haven't yet had a problem I thought it
would be the solution for, so haven't looked into it. The trouble I have
is that the only people I can think of I'd share files with this way
would only go along with it if I used Dropbox instead. It's a similar
kind of problem to what prevents me from ever encrypting any of my
email.

For files I don't share but sync between machines I figured rsync (using
rsync directly? do Dropbox and owncloud use rsync under the covers? do
the rsync authors ever get a $ or stock tip for their work backing
Dropbox if so?) has a richer set of options and is more flexible.

-- 
Mike Small
sma...@sdf.org
___
Discuss mailing list
Discuss@blu.org
http://lists.blu.org/mailman/listinfo/discuss

Re: [Discuss] emoji in my url

[Mike Small]

Eric Chadbourne <sillystr...@protonmail.com> writes:

> I just noticed that you can have an emoji URL. I'm I just old or is this 
> moronic?
>
> The url bar should contain plain text and obscure nothing, else how do you 
> know where you are?

Is this a URL with UCS characters? This is what RFC 3986 has to say:

   When a new URI scheme defines a component that represents textual
   data consisting of characters from the Universal Character Set
   [UCS], the data should first be encoded as octets according to the
   UTF-8 character encoding [STD63]; then only those octets that do
   not correspond to characters in the unreserved set should be
   percent- encoded.  For example, the character A would be
   represented as "A", the character LATIN CAPITAL LETTER A WITH
   GRAVE would be represented as "%C3%80", and the character KATAKANA
   LETTER A would be represented as "%E3%82%A2".

This is what it considers unreserved:

   unreserved  = ALPHA / DIGIT / "-" / "." / "_" / "~"

It also says this:

   A URI is a sequence of characters from a
   very limited set: the letters of the basic Latin alphabet, digits,
   and a few special characters.

So I'd say the URI with the emoji is supposed to be encoded (assuming
it's a standard UCS emoji).

But which is more obscure, %01%F6%3C or a little cat face with a wry
smile? I might like a way to get the UCS code point and long description
from the glyph, but I think I'd rather see the kitty by default even if
the character in the actual HTTP stream has to be encoded. Actually,
there is a way outside the browser to find out the codepoint. You could
copy and paste the glyph to the command line and run a command named uni
(included with the Perl module App::Uni on CPAN) on it. So yeah, if your
browser gets %01%F6%3C in a URI and shows you a face instead of the
standard URI encoding I think that's great (if there aren't security
implications from doing that, and if it lets you set this to your
preference).  But if it's some stupid thing like what Pidgin does to
certain character pairs then I'm with you. That would be awful.

-- 
Mike Small
sma...@sdf.org
___
Discuss mailing list
Discuss@blu.org
http://lists.blu.org/mailman/listinfo/discuss

Re: [Discuss] Ubuntu 16.04 Wifi Connection Disappeared

[Mike Small]

Nancy Allison <nancythewrit...@gmail.com> writes:

> Preface:  I decided to use Ubuntu on my laptop as a way of thumbing my nose
> at Microsoft and Apple. Most Linux people are pretty technically savvy. Me,
> not so much.

Cool.

>
> Suddenly, my 16.04 installation doesn't connect to wifi. I am writing to
> you on my Windows 7 machine, which does successfully connect to the
> self-same wifi network.
>
> In the Ubuntu interface, I tried  disabling my home wifi network, foolishly
> thinking that Ubuntu 16.04 would successfully find it when I rebooted. No
> dice.
>
> I found a website with a long discussion of this Ubuntu 16.04 problem. I
> would type all that stuff beginning with "sudo" if I knew where to type it,
> and if I weren't afraid that I might misunderstand something, type
> something wrong, and make the situation worse.

I'll only answer this part since the Ubuntu machine I use at work is on
a wire and I'm not using its default user interface (Unity). At home I
use other things (Slackware and OpenBSD) and again not with wireless
only with an ethernet cable. So someone else will be better at directing
you to where to find error messages in the UI to diagnose better.

You can type that sudo command by first running a program named
gnome-terminal (or xterm or maybe some other programs with terminal in
the name). If I remember rightly in Ubuntu/Unity it's in the menu, maybe
only labeled as "terminal" or "gterminal". If you've ever used the
MS-DOS prompt in Windows it's kind of like that. When it comes up you
can type the command there. I assume you mean the command...

sudo service network-manager restart

That's pretty safe to type in. While the "sudo" part does make the
command run as root (with superuser or admin privileges), a plausible
typo in the rest of the command isn't going to cause anything beyond an
error message being spat out. There are things you could type after the
sudo part that would cause severe damage but you wouldn't realistically
do that by accident trying to type this command. After you type it in
you'll be asked for a password. You enter the password of your Ubuntu
user.

>
> http://askubuntu.com/questions/762198/16-04-lts-wifi-connection-issues
>
> I would be very grateful for suggestions about how to fix the problem.

The problem the people in that thread were having may be your problem
but then again yours may be something else. They saw their issue waking
up the computer from sleep. Did your wireless work initially under
Ubuntu but then go away after putting the machine to sleep? Other
possibilities...

1. Linux, or Ubuntu in particular, doesn't support your wireless card.
If you can tell us more about the hardware you're using someone may be
able to help determine this or rule it out as a cause.

2. Some cards Linux only supports with the assistance of something
called firmware that gets loaded when the system starts up. I don't have
much experience with wireless firmware under Linux and none with Ubuntu
so maybe someone else can help here if they consider it a potential
sticking point.

3. Something more pedestrian like your wireless password not being typed
in correctly?

-- 
Mike Small
sma...@sdf.org
___
Discuss mailing list
Discuss@blu.org
http://lists.blu.org/mailman/listinfo/discuss

Re: [Discuss] Systemd & not Re: I, uh, deleted the wrong kernel....

[Mike Small]

Jerry Feldman <gaf.li...@gmail.com> writes:
...
>> The only thing I kind of dislike about slackware is the directory
>> structure in the 64 bit version. Libraries go into /lib64 and
>> /usr/lib64.  Obviously this is a pretty petty criticism, but not really
>> needing multilib I find it annoying to have these directory names. I
>> prefer OpenBSD's approach, which is to not bother with multilib and have
>> simple expected names like /usr/lib and /lib.
>>
> Fedora  and RHEL are the same way. That is because they fully support
> both 64-bit and 32 bit libraries and applications.

And Debian/Ubuntu too, though I almost like their solution better even
though it's actually more complex.  It just seems backward looking to do
the rename on the 64 bit directory. I guess in 2006 or whenever it
probably seemed reasonable, but in retrospect lib32 for x86 and lib for
x86_64 would have made a lot more sense.

Like a lot of things in Linux, for the sake of someone who needs a
feature you have this added piece of clutter/complexity you may
eventually need to become aware of that you would never hit on a simpler
less feature laden O/S. (But of course if you actually need to run 64
bit and 32 bit processes side by side OpenBSD's answer of, "if you need
32 bit install the 32 version of the O/S," might be frustrating, at
least until they finish the vm hypervisor they've started maybe.)

-- 
Mike Small
sma...@sdf.org
___
Discuss mailing list
Discuss@blu.org
http://lists.blu.org/mailman/listinfo/discuss

Re: [Discuss] I, uh, deleted the wrong kernel....

[Mike Small]

One other thing. The boot command edit area has tab completion on path
names. Use it. In this case it's not just saving typing but will save
you time and puzzlement from missing by a letter.

Mike Small <sma...@sdf.org> writes:

> David kramer <da...@thekramers.net> writes:
>
>> No, I'm not building my own kernels.  I have the 34 and 36 kernels, so
>> one path (and it would be a great one) would be to somehow get grub to
>> look for those instead of 28.  How to do that after booting off USB
>> though.  That would be my favored option, and I'll have to google in
>> that direction some more.
>>
>> Thanks.
>
> Start with the grub manual. You're going to be typing something
> in that looks like the entry did in your config file for the lost kernel
> but with different version numbers. Sadly in Ubuntu the config file
> looks like a little program itself so it may be a little hard to find,
> but something like this...
>
> linux /vmlinux-4.4.0-34-generic root=/dev/sda1 
> initrd/initrd.img-4.4.0-34-generic
>
> Maybe you'll need some other commands first if that doesn't work alone,
> e.g. load_video, insmod or other crud. You'll see this stuff in each
> menuentry block in /boot/grub/grub.cfg.

-- 
Mike Small
sma...@sdf.org
___
Discuss mailing list
Discuss@blu.org
http://lists.blu.org/mailman/listinfo/discuss

Re: [Discuss] I, uh, deleted the wrong kernel....

[Mike Small]

David kramer <da...@thekramers.net> writes:

> No, I'm not building my own kernels.  I have the 34 and 36 kernels, so
> one path (and it would be a great one) would be to somehow get grub to
> look for those instead of 28.  How to do that after booting off USB
> though.  That would be my favored option, and I'll have to google in
> that direction some more.
>
> Thanks.

Start with the grub manual. You're going to be typing something
in that looks like the entry did in your config file for the lost kernel
but with different version numbers. Sadly in Ubuntu the config file
looks like a little program itself so it may be a little hard to find,
but something like this...

linux /vmlinux-4.4.0-34-generic root=/dev/sda1 
initrd/initrd.img-4.4.0-34-generic

Maybe you'll need some other commands first if that doesn't work alone,
e.g. load_video, insmod or other crud. You'll see this stuff in each
menuentry block in /boot/grub/grub.cfg.

-- 
Mike Small
sma...@sdf.org
___
Discuss mailing list
Discuss@blu.org
http://lists.blu.org/mailman/listinfo/discuss

Re: [Discuss] EOMA68 Computer

[Mike Small]

Steve Litt <sl...@troubleshooters.com> writes:
>> consistent with his numbers in the white paper: "if the computer was
>> not connected to the Internet it could continue to be used for its
>> designated tasks until it suffered major component failure (possibly
>> in 8 to even 15 years time)," though I dislike the "if the computer
>> was not connected to the internet" part. At some point you could
>> imagine repair shops coming back because of modular design but it's a
>> distant point from where we are now.
>
> I'm guessing from the opinions expressed above, I'm guessing you're
> over 21. Once upon a time, before they grew up, I had college-age
> triplets, each with their own laptop. Between ages 18 thru 21, not one
> of those laptops lasted more than 1.5 years. Hard drives, power jacks,
> broken hinges, cracked screens, busted cases. Two have graduated
> college (they're 23 now), and both of those walk around with incredibly
> busted computers because I won't pay for another one, and neither will
> they.
>
> Also, the computers of 2000-2008 were much more robust than today's
> thin, light, cut glass ornaments that pass for laptops. I bought two
> laptops in 2006, and except for a bad wifi card caused by a bad
> replacement attempt by me, they work perfectly.

Yeah, I'm more than twice 21 and my son hasn't had his own computer
yet. I also am yet to experience this modern low quality phase. My spare
laptop is from 2006 and my new one, the one I use, from 2010. And then
before that I had a powermac that showed no signs of ever dying.  But
the slowness was too much even for me, it not being able to handle H.264
encoded mpegs. Even the group I gave it to could only really use it for
demonstrations of what's inside a computer rather than for someone to
actually use. But it was pretty cool for that the way it opens on a
hinge and all.

Good to know to avoid this era of laptops. Maybe I'll try desktops or
small form factor next time I need to get something. Part of me wants to
try to use a soekris card as a desktop. I splurged this winter, being
sick of the fan noise on my 2006 laptop when used as a router
(relatively large power consumption too, > 30W), and bought a Soekris
net4501 off ebay. It's pretty old and presumably had been in service
some of this time but it still chugs away only using a watt or two. A
net4501 couldn't be (much of) a desktop but when net6501-50's get below
$50 used... What do you think, a 1Gz Intel Atom E6xx should be able to
play dvds and youtube videos, right? Put a video card in one pci slot, a
sound card in the other or on USB (hey, what's not modular about this?)
and a medium size SSD on the SATA interface, should work (er, uh 30 W
limit -- have to watch for that):
http://soekris.com/products/net6501-50-board.html

Well, no doubt there's a better choice among small form factor devices,
but the soekris cards are supposed to be super reliable I hear, being
meant as networking equipment.

Hmmm, but I guess my prior argument should have me trying to use these
crappy laptops, getting by by making frequent backups, since there ought
to be little used market for them, while soekris boards are in demand
as long as they function.

-- 
Mike Small
sma...@sdf.org
___
Discuss mailing list
Discuss@blu.org
http://lists.blu.org/mailman/listinfo/discuss

Re: [Discuss] EOMA68 Computer

[Mike Small]

ng on a
   computer with 8 Gigabytes of memory and a Dual-Core Dual-Hyper-threaded
   2.4 Ghz processor, causing it to reach 100% CPU and lock up the entire
   machine.

It's not actually that bad I don't think (but I will reject a site that
rejects me -- and Street View? is it such a sacrifice not to have Street
View?). For instance, I notice banking and credit card sites are usable
enough on old machines. All sites hosting free software projects are
perfectly fine. BLU's site? Perfectly okay ;). Maybe there's some kind
of threshold for visitors that different kinds of companies are willing
to discard. Banking sites maybe that's some fraction of a
percent. Google I think it's at least 1% based on the obnoxious messages
they give when you don't whitelist them on noscript or librejs. So I
think there are a bunch of ways you can deal with the bloated sites some
web developers are producing these days. If we could spread tips and
(free) software to help here as we simultaneously save old computers
and get them used by people in substitution for new purchases then maybe
numbers in weblogs would force accomodation of slower machines with less
memory. You know, get the stubborn, difficult group who use things like
noscript or go away from a site early if it's slow up to 3% and see how
companies react to that.

IngeGNUe <ingeg...@riseup.net> writes:
> On 07/19/16 11:20, Mike Small wrote:
>> Having cheap upgrade options and having those
>> options publicized might also make me more tempted to consume more not
>> less.
>
> Actually, this could be a problem if big corporations take this project
> and run with it.
>
> But I think that's as issue with the profit motive taking precedence
> over the environment whatever you're selling.
>

I'm inclined to disagree here. My feeling is that environmental efforts
tend to be niche until a profit motive takes force, perhaps with the
help of a subsidy. Take Denmark and wind power for instance. I'm not
deeply familiar with their case but it sounds to me like the country
decided early to go to wind, the government introduced subsidies to
shift the supply curve as necessary to make it possible for companies to
create a wind industry, but then companies (whether big or small I don't
know) built up the equipment and produced the power, profit motive
driving all of that. Where there are not government subsidies we wait
until the thing stands on its own at market equilibrium. Maybe there's
some demand curve shift from environmental awareness, something similar
to and in the reverse direction (reduced consumption) of the demand
shift you get from advertising, but it's not an obvious effect. How many
times do you see protest and campaign to change behaviour but only
e.g. when the price of gasoline goes up do you see a noticeable change?

But I guess Free Geek is nonprofit. They also deal in the price system
to get sales to offset their costs, but as a nonprofit they could choose
reduced sales if that somehow better fulfills their mission statement.
Is that what you're getting at?  Hmmm, I guess now I think the onus for
not upgrading unnecessarily should be on the consumer in this case. Free
Geek should be selling as much as they can, constrained only by what's
available to divert. And if I feel like modularity in this new laptop
could induce me to upgrade unnecessarily I should deal with that myself
by not doing it (but I still won't buy one new).

___
Discuss mailing list
Discuss@blu.org
http://lists.blu.org/mailman/listinfo/discuss

Re: [Discuss] EOMA68 Computer

[Mike Small]

IngeGNUe <ingeg...@riseup.net> writes:

> This computer right here:
>
> https://www.crowdsupply.com/eoma68/micro-desktop
>
> Eco-friendly, upgradeable, portable, secure, affordable, no NDAs, no
> proprietary anything (except for MALI, if you insist on it), fully
> documented hardware. Can be used as phone (maybe phablet), tablet,
> laptop, desktop, "low-power co-located servers", anything you can fit it
> into.
>
> This is one of those projects which will change the way we do computing
> on GNU/Linux. Please read and if you like it, I hope you will support it!!!

Hmmm, I wonder if it would turn out to be as eco-friendly as intended in
practice. The people who want to reduce their footprint I think should
keep whatever they have now running and not buy anything new. Plus when
you need a computer their tonnes of old crap to grab cheap before it
hits landfill. I could see something like this letting me trick myself
into buying before my current machines are fully consumed or buying new
when I should look to used.

It would be nice if it influenced the industry to move to machines where
you can upgrade just the thing that's lacking maybe. Or would it? I find
my old laptops remain perfectly adequate and I haven't looked into
upgrade options: their upgradability certainly is not a feature anyone
has advertised to me. Having cheap upgrade options and having those
options publicized might also make me more tempted to consume more not
less. Of course, if it's really more fixable when one component fails
that's plainly good.

My own experience with hardware was to be an upgrade junky in the
90s. With some headaches you could get a new motherboard, change the
video card add SIMMs or DIMMs, upgrade the modem, etc. I'd end up
getting all this stuff I didn't really need cause you could do it
somewhat incrementally. Then I splurged and bought a powermac. Partly I
paid so damn much (at one time) for the foolish thing I didn't want to
buy new hardware for years. But also it lacked upgradability (or that
was my perception maybe combined with an irrational feeling that it was
a unit with a single identity less so than an aggregate of parts)
compared to the PC clones I'd previously dealt with. That helped me kick
the habit. And as mentioned above the limited upgradability I have now
with old laptops instead of desktops helps. Odd, somehow this old crap
just keeps running and running. Do computers still break? ;)

My cynical side fears this is a little like Macintosh or automobile
marketing, as in it's a computer whose first purpose, the purpose
sparking the sale, is to satisfy the soul who wants to express who he or
she is via purchases or in effect it working out that way despite the
good intentions of the founders. Hmmm, if that gets more people to use
GNU+Linux, okay, but maybe the way it works out in the market would not
end up being a net plus on the environmental or conflict mineral side.

-- 
Mike Small
sma...@sdf.org
___
Discuss mailing list
Discuss@blu.org
http://lists.blu.org/mailman/listinfo/discuss

Re: [Discuss] ssh keys question

[Mike Small]

Kent Borg <kentb...@borg.org> writes:
>>  passwd_timeoutNumber of minutes before the sudo password 
>> prompt times out, or 0 for no timeout. The timeout may include a 
>> fractional
>>component if minute granularity is 
>> insufficient, for example 2.5.  The default is 0.
>
> I want 0 to mean timeout immediately, not timeout never! If people want 
> never let them put 99 in there and let their heirs deal with the  
> eventual timeout.
>
> Guess I could try 0.001
>

If the password prompt timed out immediately how would you type in your
password? I think you mean to be discussing the other timeout option,
timestamp_timeout. Zero there gives you always ask for a password
behaviour.

I like very much having 0 to never timeout the password prompt. It's
annoying to come back to a build and find that make worked but make
install quietly failed after sudo timed out waiting for a response.
Especially bad in things like CPAN where it might fly by and move on to
something else without you seeing it.

-- 
Mike Small
sma...@sdf.org
___
Discuss mailing list
Discuss@blu.org
http://lists.blu.org/mailman/listinfo/discuss

Re: [Discuss] ssh keys question

[Mike Small]

"Rich Braun" <ri...@pioneer.ci.net> writes:
> I often wish sudo had functionality similar to ssh-agent: a way to require a
> token established at session start, rather than a password entered every time.

Is your timestamp_timeout option zero? That seems more or less the same
idea, with timeout set to taste or need. I miss the feature since
OpenBSD 5.8 because I uninstalled sudo when they created doas, and the
latter has no such feature last I looked. But I can't justify to myself
having doas, sudo, and su. doas and su are enough, at least for what I'm
using OpenBSD for now.

-- 
Mike Small
sma...@sdf.org
___
Discuss mailing list
Discuss@blu.org
http://lists.blu.org/mailman/listinfo/discuss

Re: [Discuss] ssh keys question

[Mike Small]

Kent Borg <kentb...@borg.org> writes:

> On 06/16/2016 06:37 PM, Dan Ritter wrote:
>> 1. You can assign passwords, but tell sshd to only allow access via 
>> keys. This is a Good Idea. 
>
> So for you--someone running your own machine--you use keys to login but 
> still use a password on sudo? (This is common? Seems part of going to 
> keys is to get rid of passwords.)

I hope it's common. Though personally I don't ssh back into my home
network, so my ssh keys are only for outgoing connections.

> But if you do not require a password on sudo it means that any program 
> you run runs with root privileges if it just bothers to ask for it. 
> Kinda the opposite of dropping privileges.

sudo is also handy for dropping privileges. E.g. when using Linux at
home I like using a 2nd user to run firefox (for general surfing) and to
run xterms to play movies or audio files I've downloaded from the web,
and when running a bittorrent client: i.e. when running complicated buggy
programs against untrusted data from the world at large.

otheruser below (names changed) is a 2nd user I created with little in
his home directory and mainuser is my normal user who has more
information in his directory.

alias xunpriv='xhost +si:localuser:otheruser; sudo -iu otheruser xterm'
alias ff='xhost +si:localuser:otheruser; sudo -iu otheruser firefox'

The following line in my sudoers lets my regular user sudo to the less
privileged user (but not vice versa):

# sudo -iu otheruser xterm   expands to /bin/bash -c xterm
mainuser ALL=(otheruser) NOPASSWD: /bin/bash -c *

(If someone sees any mistakes here please point them out.
Reading the sudoers man page is an endeavour.)

I also have something set up in su like BSD's wheel group so that only
the main user can use su to get to root. I'm not sure if this would be
the same on other distributions than Slackware, which is a holdout not
only on systemd but on pam.  If I remember rightly there's a different
way to emulate wheel when using pam.

1223r0:etc$ cat suauth
root:ALL EXCEPT GROUP wheel:DENY
mainuser:ALL:DENY

otheruser is not in group wheel so can't su to root, nor can he su to
mainuser due to the 2nd line.

Some people advocate making otheruser nologin. I haven't done that yet
cause I used to like to be able to login as that user when I knew I was
only going to listen to one podcast and then turn off the computer.

otheruser also lacks the ssh keys I use to connect to other machines.

-- 
Mike Small
sma...@sdf.org
___
Discuss mailing list
Discuss@blu.org
http://lists.blu.org/mailman/listinfo/discuss

Re: [Discuss] GNU xargs trick

[Mike Small]

Rich Pieri <richard.pi...@gmail.com> writes:

> On 4/27/2016 12:31 PM, Mike Small wrote:
>> New line characters in particular. -I changes matters to line at a
>> time so other space characters should be okay, right?
>
> Yes, line at a time, but you may need to quote the substitution string
> in the command arguments.

I wondered about that but when I tried a little experiment it seemed
like it's probably not necessary. When I "cat | xargs -I{} ls {}" and
type some two word lines the files ls complains of being missing do seem
to be the two word "files" I'm typing. Seems like maybe what xargs gets
(i.e. the expansion of {}) goes right to one of the exec library calls
without further shell intervention. Let me know if I'm wrong on this. I
continue to find the various Unix shell expansions too complicated to
keep in my head at once or retain over time. And unfortunately I'm not
getting any smarter as I age.

On this topic, the debian bug report where GNU Parallels came into being
as a debian package has some interesting discussion re. GNU parallel and
xargs and another thing called parallel from moreutils. Early in the
discussion there's the question of what parallel can do that xargs
cannot:

https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=518696

-- 
Mike Small
sma...@sdf.org
___
Discuss mailing list
Discuss@blu.org
http://lists.blu.org/mailman/listinfo/discuss

Re: [Discuss] GNU xargs trick

[Mike Small]

Daniel Barrett <dbarr...@blazemonger.com> writes:

> On April 27, 2016, Rich Pieri wrote:
>>find . -name "*.png" -print | xargs -I{} -P 4 optipng -o1 -preserve {}
>
> Nice trick. I'd recommend using "find -print0" and "xargs -0" for
> safety, in case any of your PNG filenames contain space characters:

New line characters in particular. -I changes matters to line at a time
so other space characters should be okay, right?

-- 
Mike Small
sma...@sdf.org
___
Discuss mailing list
Discuss@blu.org
http://lists.blu.org/mailman/listinfo/discuss

Re: [Discuss] Whence distributed operating systems?

[Mike Small]

David Rosenstrauch <dar...@darose.net> writes:

> On 04/21/2016 12:50 AM, Mike Small wrote:
>> "Sadly it seems that we now need to either wait for Linux or Windows to
>> catch up with the 1980s state of the art in distributed systems (think
>> Locus or AFS). What went wrong? Products like DataSynapse’s FabricServer
>> look like an interesting attempt to address the problem, at least for
>> the Java world, but it feels to me that mainstream operating systems
>> designers seem to have lost the plot somewhere along the way."
>>
>> http://discovery.bmc.com/community/blog-post/whatever-happened-to-distributed-operating-systems3/
>>
>> Is single system image still a thing?
>
> Aren't systems like Apache Mesos (which didn't exist back nearly 10 
> years ago in 2007 when the author wrote that post) the natural successor 
> to DataSynapse FabricServer, and an example of the "distibuted operating 
> system" he's talking about?  I.e., just a big pool of CPU cores, where 
> different portions of the pool can be utilized for different types of 
> distributed workloads.

Sounds more like what he's talking about. Are these kinds of systems
gaining much traction? I'd never heard of Mesos.

-- 
Mike Small
sma...@sdf.org
___
Discuss mailing list
Discuss@blu.org
http://lists.blu.org/mailman/listinfo/discuss

Re: [Discuss] Whence distributed operating systems?

[Mike Small]

Hmmm. I guess I meant single system image more in the sense of an
extension of the research work done in the 80s (and nineties a little?)
in the sense of there being a real network operating
system. i.e. something picking up from where Chorus or Amoeba left
off. (I haven't used these systems only read a little about them in
Tanenbaum's Distributed Operating Systems book.) If you'll allow me to
be a bit obnoxious, i.e. single system image not bundled together from
what we had combined with the requisite amount of bubble gum and baling
wire.

I'm thinking from the perspective of a person writing a program more
than an end user. It's easy to abstract details from a user (though
whether the result is completely palatable without them buying new
computers often may be in question). What would be interesting would be
to write to an operating system interface and have the system
transparently migrate processes by various criteria, do redundancy for
failure and load, etc. etc and have it accessible by friends and family
or for it to get popular and scale nicely without further work or too
high a cost. Then supposing the application writer is curious about how
all that works, it would be further interesting for him to be able to
read a nice book by Andrew Tanenbaum and, if not read all the code, at
least learn the next layer of abstractions and have it be somewhat
elegant, or at lease not something to make him slam the unscrewed panel
down in horror and paint a radioactive sign upon it. Further it would be
nice if reading said book helped supplement the application developer's
mental model in a way that makes writing programs more satisfying and
the resulting programs run yet better.

I guess I'm hoping for a new system programming hobbiest and follow on
entrepreneurs and a new inflection point for the industry. Something
feels deeply wrong about where we are right now, to the degree I get the
gist of it. Ditto twice over for this whole problem of the web browser
as client application delivery mechanism and defacto operating
system. If new Linuses come along whose design will they snarf?

Jack Coats <j...@coats.org> writes:

> [1:text/plain Hide]
>
> In many ways, we do have a single system signon, with a 'single system
> image' for well developed systems today.
>
> This is not true for everything, but within many companies web presence,
> the user (who we are in in business to support, right?) sees a 'single
> system image', whether it is implemented on a single system or a complex
> system or a 'cloud' (basically an obscured set of cooperating systems).
>
> Most often we just ask for users to sign in once to access all aspects of
> their 'user experience', even if the systems do re-authorization behind the
> curtains.
>
> For supporting the systems that provide this illusion to the users, we are
> still lacking on making that as smooth as it needs to be eventually.
>
> The cost of computing has kept going down for years.  And to make it
> economical to provide seamless experience for users, the cost of networking
> and computing in general has had to go way down. ... Even on a single
> platform, let's talk word processing.  Today Word or equivalent is the
> standard.  It generates a very flexible document.  Not just text anymore,
> but colors, multiple fonts, graphics, and much more, let alone the
> programability built into the macro type functions.
>
> This is all great, but it comes at a price.  Just look at the difference in
> size between a text only ASCII file that says "The little brown fox jumps
> over the lazy dog." and the word processing document that does that.
>
> The price?  Complexity in the programming, the size of files to store and
> shuffle from place to place (and associated network traffic).  Bigger
> systems (not individual but think all the aspects it takes to run things)
> take more administration, maintenance, power, people effort as developers,
> admins, system maintainers, let alone the overhead that comes with that and
> keeping things semi-organized.
>
> Is all this worth it?  Today the market has said it is.
>
> This whole computing thing is to provide users with a way for them to be
> more profitable in their lives.  Whether it is to lower stress, communicate
> easier, process information in a way than makes sense to them (not
> necessarily us), at a price that the end customer can tolerate and many of
> us 'middlemen' can still make a living (some better than others).
>
> Back toward the original subject, the reason that Single System Image was a
> big deal was to simplify life for customers and to reduce overhead for the
> customers.  Single Sign On was part of the whole thing too.
>
> So back to the question: Is SSI still a thing?  Yes. ... Just re-branded,
> revamped, re-released under a label saying it is al

[Discuss] Whence distributed operating systems?

[Mike Small]

After the meeting I was discussing this issue with a friend. It's not an
original criticism I didn't suppose, so I found someone with better
words to sum up my reaction:

"Sadly it seems that we now need to either wait for Linux or Windows to
catch up with the 1980s state of the art in distributed systems (think
Locus or AFS). What went wrong? Products like DataSynapse’s FabricServer
look like an interesting attempt to address the problem, at least for
the Java world, but it feels to me that mainstream operating systems
designers seem to have lost the plot somewhere along the way."

http://discovery.bmc.com/community/blog-post/whatever-happened-to-distributed-operating-systems3/

Is single system image still a thing?

-- 
Mike Small
sma...@sdf.org
___
Discuss mailing list
Discuss@blu.org
http://lists.blu.org/mailman/listinfo/discuss

Re: [Discuss] Govt Source Code Policy

[Mike Small]

Rich Pieri <richard.pi...@gmail.com> writes:

> ... Regardless, the GNU licenses
> prohibit adding terms that restrict the removal of attributions. That's
> a problem in a world where recognition is so highly prized.

I was reading part 7 of GPLv3 while composing my previous message and saw
exactly the opposite. Maybe you could quote what you're referring to.

This is what I'm referring to:

"Notwithstanding any other provision of this License, for material you
add to a covered work, you may (if authorized by the copyright holders
of that material) supplement the terms of this License with terms:
...
b) Requiring preservation of specified reasonable legal notices or
author attributions in that material or in the Appropriate Legal Notices
displayed by works containing it; or
c) Prohibiting misrepresentation of the origin of that material, or
requiring that modified versions of such material be marked in
reasonable ways as different from the original version; or
...
All other non-permissive additional terms are considered “further
restrictions” within the meaning of section 10. If the Program as you
received it, or any part of it, contains a notice stating that it is
governed by this License along with a term that is a further
restriction, you may remove that term."


-- 
Mike Small
sma...@sdf.org
___
Discuss mailing list
Discuss@blu.org
http://lists.blu.org/mailman/listinfo/discuss

Re: [Discuss] Govt Source Code Policy

[Mike Small]

Rich Pieri <richard.pi...@gmail.com> writes:
>> I don't think it shows what you say unless the argument is that these
>> OASCR people are bright as hell and they don't even mention rms's
>> arguments so therefore said arguments must be crap. [snip]
>
> DoE-funded scientists are the same scientists who work on Fermi Linux
> (GPL), Scientific Linux (GPL), Octave (GPL), ROOT (GPL) and so forth.
> They are no strangers to various FLOSS licenses. In fact, it was their
> activities which originally brought about the DoE's OSS policies, not
> the other way around.
>
> Also, the exclusion of RMS's arguments does not make their argument
> crap. It means they see his arguments as crap and not worth bringing up.
> Especially since the audience is already far more familiar with RMS than
> the people at the DoE.

You're misreading or my sentence is ambiguous. That's what I'm
saying. I'm not saying the DoE arguments are crap but that the only
argument you could be making (and in fact seem to be making in this
email) by citing that document is that they (rms's arguments) must be
crap since DoE never bothers to take them up in said document. I don't
consider that an example of someone considering his arguments and
working out how they're off base. Or at least it's not a useful example
since we're not exposed to their reasoning only its after effects.

>> This is puzzling. First in its specificity: there are other software
>> licenses that seem to more or less accomplish what CC BY does.
>
> The GPL does not. The GPL fails to meet one of the most important
> academic criteria: recognition. What RMS calls onerous advertising is
> treasured by the academic community. That the GNU licenses permit the
> removal of recognition notices makes them unsuitable for academic works.

Well, I meant the BSD, MIT, ISC, Apache, etc. licenses here. In other
words non-copyleft licenses. Those are what I was thinking of as roughly
functionally equivalent to CC BY in the main.

When rms discusses onerous advertising he's referring to the clause in
the original 4 clause BSD license (and if he were here he'd take me to
task for earlier typing "the BSD [license]" as if there were only
one). That issue wasn't one of mentioning people in the changes file or
source files but of requiring lists of organizations to appear in your
advertising materials.

In fact is there any GNU project at all that does not keep a list of
contributors which is sacrosanct and never elided from -- or for that
matter that does not use some kind of publicly accessible source control
system? e.g. I half followed emacs-devel when they converted to git and
getting history preserved was something they seemed extremely serious
about. The GPL did not seem to interfere with this so much as the fact
that people change their name or use different emails and tags to refer
to themselves over time.

Besides which, 4 clause BSD is rejected now by all the BSDs. What in the
GPL is different from the so called permissive licenses (the ones that
are still commonly used as opposed to 4 clause BSD) regarding keeping or
not keeping the list of copyright holders or contributors?  Taking the
ISC license, for instance, there is the requirement that the copyright
notice be retained but...

1. when I look in OpenBSD source control for instance, I don't tend to
see the complete list of contributors put in the copyright (and OpenBSD
is very conscientious about attribution, licensing and copyright so I
think a good example here). e.g. ...
http://cvsweb.openbsd.org/cgi-bin/cvsweb/src/usr.sbin/ntpd/control.c

2. GPLv3 part 7 seems to allow you to add a term that requires that
attribution be done strictly, and that that addition wouldn't qualify
for downstream stripping.

In practice you'd want to get attribution information from source
control with either license involved. So I don't know what you're
getting at regarding the GPL and scientists' need for attribution (but
not advertising, I wouldn't think, unless they happen to be selling
something).

-- 
Mike Small
sma...@sdf.org
___
Discuss mailing list
Discuss@blu.org
http://lists.blu.org/mailman/listinfo/discuss

Re: [Discuss] Govt Source Code Policy

[Mike Small]

Rich Pieri <richard.pi...@gmail.com> writes:

> On 4/7/2016 12:27 PM, Mike Small wrote:
>> Ah, then I agree with you that that seems like overreach unless and
>> until someone demonstrates rms's arguments are good enough to inform
>> general law.
>
> I have evidence that I believe strongly demonstrates that RMS's
> arguments should not be used to inform general law.

Interesting. I don't think it shows that but it does appear to be an
example of a government agency asserting copyright. And now I'm mixed up
again cause I thought they couldn't do that.

I don't think it shows what you say unless the argument is that these
OASCR people are bright as hell and they don't even mention rms's
arguments so therefore said arguments must be crap. Looking at the
benefits listed in the background it fits neatly into rms's usual
framing of the difference between the free software movement and open
source people in that it's a list of practical benefits to the OASCR,
with one practical benefit to the public at large, but makes no
reference to the rights a recipient of the software ought to have, rms's
bread and butter.

>
> The DoE added a rider on educational grants in 2008 (I believe). If
> released to the public, software funded by DoE educational grant money
> must be published with an Open Source Software (OSS) license or have
> copyright transferred to the DoE for publication:
>
> http://science.energy.gov/~/media/ascr/pdf/research/docs/Doe_lab_developed_software_policy.pdf
>
> The DoE more recently have demonstrated a preference for CC BY because,
> "requiring release under CC BY maximizes the public benefit of funding
> dollars, and ensures the creator retains copyright and the option to
> offer the work under other terms that benefit the particular business
> model or mission"

This is puzzling. First in its specificity: there are other software
licenses that seem to more or less accomplish what CC BY does. Further
what are they thinking about with the issue of "retaining copyright."
The GPL doesn't, as far as I know, prevent anyone from retaining
copyright. I thought in fact that it depended upon it.  You wonder if
they're conflating other licenses with projects that use them in
combination with contributor assignment agreements.

The question of whether allowing proprietary use of the code maximizes
public benefit, now that's an interesting question. I wonder how they
went about testing that assumption. It strikes me as something that
might be true or might instead be susceptible to an existence proof to
the contrary if things went really well for GNU and friends. And then
we might have to agree on what we mean by benefit.

>
> http://energy.gov/sites/prod/files/CCBY_9_3_2011.pdf
>
> The terms "Free Software" and "GPL" do not appear anywhere in the rider,
> nor do they appear in general DoE documentation regarding funded
> software. Despite this, the corporate handout state, which is the core
> of the pro-GPL rhetoric, has not come to pass.



___
Discuss mailing list
Discuss@blu.org
http://lists.blu.org/mailman/listinfo/discuss

Re: [Discuss] Govt Source Code Policy

[Mike Small]

Daniel Barrett  writes:

> On April 7, 2016, Rich Pieri wrote:
>>No. A tyrant with good intentions is the worst kind of tyrant.
>
> Erm... I thought the worst kind of tyrant is one who tortures and
> rapes his people, steals their goods, dismembers his enemies alive,
> burns the land, and hits puppies.
>

Yeah, you have to wonder if C.S. Lewis wrote the earlier referenced
quote before or after Mussolini and Hitler.

___
Discuss mailing list
Discuss@blu.org
http://lists.blu.org/mailman/listinfo/discuss

Re: [Discuss] Govt Source Code Policy

[Mike Small]

Rich Pieri <richard.pi...@gmail.com> writes:

> On 4/7/2016 11:23 AM, Mike Small wrote:
>> I haven't understood the thread, not seeing how it is that the U.S.
>> government now can hold copyright on its works and thus it be
>> possible for them to use the GPL.
>
> Because it's not about the government per se. It's about software funded
> by government money. For example, MIT LNS is funded by Department of
> Energy grants but programs written by scientists here are not owned by
> the DoE; they are owned by the scientists who write them.

Ah, then I agree with you that that seems like overreach unless and
until someone demonstrates rms's arguments are good enough to inform
general law. Failing that I'd think it would be better to have laws that
help scientists, or whoever, to use licenses like the GPL if they would
like to when their employers aren't agreeable to the idea. Or perhaps a
more general law making individual copyright less alienable. I wonder
how it would work if the standard employment agreement instead of
assigning copyright (in some cases even to work done off premises
unrelated to the day job) suggested a license under whose terms I agree
to release code. Then again, if I keep participating in these threads
fair use would probably be sufficient to take what little I manage to
write.

-- 
Mike Small
sma...@sdf.org
___
Discuss mailing list
Discuss@blu.org
http://lists.blu.org/mailman/listinfo/discuss

Re: [Discuss] Govt Source Code Policy

[Mike Small]

IngeGNUe <ingeg...@riseup.net> writes:
>> Of all tyrannies, a tyranny sincerely exercised for the good of its
>> victims may be the most oppressive. It would be better to live under
>> robber barons than under omnipotent moral busybodies. The robber baron's
>> cruelty may sometimes sleep, his cupidity may at some point be satiated;
>> but those who torment us for our own good will torment us without end
>> for they do so with the approval of their own conscience. They may be
>> more likely to go to Heaven yet at the same time likelier to make a Hell
>> of earth. This very kindness stings with intolerable insult. To be
>> "cured" against one's will and cured of states which we may not regard
>> as disease is to be put on a level of those who have not yet reached the
>> age of reason or those who never will; to be classed with infants,
>> imbeciles, and domestic animals.
>> 
>
> Is that a quote from like, John Stuart Mill? So, in other words a despot
> with good intentions is worse than bad intentions for the whole of the
> population? In any case, libre-lovers aren't despotic, the free software
> world is more of a lightly-regulated market.

It's a nice quote but I agree. Free software advocates could only wish
to have this kind of power now or in the foreseeable future. I haven't
understood the thread, not seeing how it is that the U.S. government now
can hold copyright on its works and thus it be possible for them to use
the GPL.  But even if they did, it's a fairly small thing IMO. The
slippery sloop would have to have superconducting magnets installed to
get to this dystopia Richard invokes via C.S. Lewis.

But instead suppose copyleft terms or some other way of ensuring each
recipient of software has rms's four freedoms were mandated by law for
all software in the country. Plainly this doesn't fit with the present
majority will, since not even 50% of this linux user group here seems
all that enthusiastic about copyleft. Do you or Greg have another
philosopher you'd invoke to support the argument that such a law is
justifiable without being popular? Or do you not think so.

-- 
Mike Small
sma...@sdf.org
___
Discuss mailing list
Discuss@blu.org
http://lists.blu.org/mailman/listinfo/discuss

Re: [Discuss] Govt Source Code Policy

[Mike Small]

Rich Pieri <richard.pi...@gmail.com> writes:

> On 4/4/2016 4:05 PM, Mike Small wrote:
>> That's a bit weak. You would only be liable if you in fact did not erase
>> your backups and had some. So their definition would stand on a
>> hypothetical present fact (that backups exist) and a hypothetical future
>> action (that the victim doesn't destroy them) for the hypothetical "you"
>> the license references.
>
> We know that backups do exist because we know the device had previously
> been synched with iCloud and those backups were not erased.

We'd need a lawyer here but I don't think you get to go specific when
interpreting the definitions of the license in general. i.e. I can't see
a legal interpretation of the GPL that makes the meaning of convey
conditional on the particular licensee's circumstances. You'd have a
kind of Schrodinger's license in that case.

>> But for the sake of argument let me concede the point. Let's say Syed
>> Farook's phone had had GPLed Apple system software on it. It would have
>> been Farook who would be breaking the GPL by not passing on the
>> "authorization info" as he "conveyed" this GPLv3ed iPhone software to
>> the FBI. This possibility wouldn't have prevented or disincentivized
>> Apple from using the GPLv3 + autodestruct.
>
> There are two possibilities here.
>
> One is that this information is generic to all of that device model. In
> this case all the FBI would need to do is have an agent buy an iPhone
> and request the information. In this case the FBI would not need Apple
> to sign their custom GovtOS in order to avoid wiping the device.

Generally people argue that in the long run, not depending on security
by obscurity forces people to make systems that work as intended even
when attackers have the benefit of source code. The FBI can buy versions
and be free to mess with them and see how they work, but so could
academic and industry security researches with the results being made
available so Apple could try again better next time.

>
> The other is that the DRM is uniquely keyed to the device. In this case
> the FBI might actually need Apple's intervention if said information
> were not in the employer's possession and not in Farook's effects, right?

>
> Wrong.
>
> Every or nearly every version of iOS, including the version on Farook's
> employer's iPhone, has vulnerabilities that can be exploited in order to
> run unsigned versions of the operating system. GPL Part 3 prohibits
> using laws like WIPO as protection which means the dissemination of
> exploits cannot be prevented or suppressed by those laws. In this case
> the FBI would legally have the information necessary to circumvent the
> DRM and thus still would not need Apple to sign their custom GovtOS in
> order to avoid wiping the device.
>

Not sure I'm understanding you. First off, the FBI as a criminal
enforcement agency is themself excempt from the DMCA:
https://www.law.cornell.edu/uscode/text/17/1201
Remember also again that Apple would not need to fear being out of
compliance with the GPL on software they're the sole copyright owner
of. You'd need some other copyright holder up the chain of what they're
distributing for that to matter when they violated that clause with a
DMCA suit. Then how much effect are DMCA civil suits really going to
have on dissemination of exploits?  About as much as copyright law has
had on the availability of movies with the copy protections stripped off
I should think. And besides, as was puzzling all along in this case, the
FBI is no doubt perfectly capable of coming up with their own exploits
or hiring someone to quietly do so.

> I have to admit: it's been entertaining watching you GPL adherents try
> to punch holes in your own favorite software license in order to prevent
> the FBI from hypothetically doing what it was carefully crafted to
> explicitly permit.

There's no irony here. I like the idea of a GPL with provisions not
granting equal rights to scumbags who spy on environmental orgs and
black lives matter activists or to people who manufacture weapons. But
it's easy to see the mess that would result if everyone had their pet
restriction added in. They struck the right balance, as usual IMO. So to
the degree the FBI exercises their right to mess with software they come
into possession of it's cool they're granted such rights.

And obviously the GPL wasn't "carefully crafted" to permit someone to
take your device and get at your data. I mean, maybe in the 80s rms had
said something that seems funny now about passwords, but today the FSF
is promoting use of encryption.

-- 
Mike Small
sma...@sdf.org
___
Discuss mailing list
Discuss@blu.org
http://lists.blu.org/mailman/listinfo/discuss

Re: [Discuss] Govt Source Code Policy

[Mike Small]

Rich Pieri <richard.pi...@gmail.com> writes:

> On 4/4/2016 12:50 PM, Mike Small wrote:
>> I don't know a lot about copyright, but I'm guessing being a victim of
>> theft or warranted seizure aren't among experiences that "would make you
>> directly or secondarily liable for infringement under applicable
>> copyright law."
>
> You are guessing incorrectly.
>
> If for any reason an original (the originally licensed copy of a work)
> is no longer in your possession you must destroy all archival copies of
> that original which remain in your possession. Failure to do so at your
> first opportunity makes you liable for copyright infringement.
>
> Copyright law in the US makes no exceptions as to why possession of the
> original is not retained so yes, theft or seizure would make victims
> liable for infringement.

That's a bit weak. You would only be liable if you in fact did not erase
your backups and had some. So their definition would stand on a
hypothetical present fact (that backups exist) and a hypothetical future
action (that the victim doesn't destroy them) for the hypothetical "you"
the license references.

But for the sake of argument let me concede the point. Let's say Syed
Farook's phone had had GPLed Apple system software on it. It would have
been Farook who would be breaking the GPL by not passing on the
"authorization info" as he "conveyed" this GPLv3ed iPhone software to
the FBI. This possibility wouldn't have prevented or disincentivized
Apple from using the GPLv3 + autodestruct.

For one thing if Apple were the only copyright holders section 6 and in
fact all of the GPL is irrelevant since the only one it could fear a
lawsuit from would be itself. But imagine there was an upstream to be
concerned about. Apple in fact is in compliance each time they "convey"
so long as they include the "installation information" that prevents the
thing turning into pixie dust when tampered with. If one of those folks
further "conveys" without "installation information," say by being shot
dead as they are "conveying" before getting a chance to yell out the
"installation information," Apple has nothing to fear. The GPL doesn't
open you to copyright suits from passing on software to someone who goes
on to violate the GPL. So Apple could happily use the GPLv3 to try to
ensure each user has full control of software they've received while
simultaneously using some kind of autodestruct to help the same users
evade disclosure of their info to the feds, all without fear of a
copyright lawsuit.

As an aside, the possibility of this kind of unintended side effect is a
good argument for licensing GPLv3 or later. If you've read or heard any
opinions from the FSF about this case you would imagine the "or later"
would attempt to snuff out any tip of the scales in favour of the Feds
if it actually existed.
___
Discuss mailing list
Discuss@blu.org
http://lists.blu.org/mailman/listinfo/discuss

Re: [Discuss] Govt Source Code Policy

[Mike Small]

Rich Pieri <richard.pi...@gmail.com> writes:

> On 4/4/2016 11:28 AM, Mike Small wrote:
>> on a device they keep), without any autodestruction happening.  When the
>> FBI takes possession of a person's, especially a late person's, device
>> and software they aren't being conveyed anything. They've simply taken
>> hold of someone's device and software.
>
> The GPL itself states otherwise:
>
>
> To “propagate” a work means to do anything with it that, without
> permission, would make you directly or secondarily liable for
> infringement under applicable copyright law, except executing it on a
> computer or modifying a private copy. Propagation includes copying,
> distribution (with or without modification), making available to the
> public, and in some countries other activities as well.
>
> To “convey” a work means any kind of propagation that enables other
> parties to make or receive copies. Mere interaction with a user through
> a computer network, with no transfer of a copy, is not conveying.
>
>
> The GPL makes no distinctions between gifts, thefts and warranted
> seizures. There may be a loophole in there but if there is then I for
> one don't see it.

I don't know a lot about copyright, but I'm guessing being a victim of
theft or warranted seizure aren't among experiences that "would make you
directly or secondarily liable for infringement under applicable
copyright law."

-- 
Mike Small
sma...@sdf.org
___
Discuss mailing list
Discuss@blu.org
http://lists.blu.org/mailman/listinfo/discuss

Re: [Discuss] Govt Source Code Policy

[Mike Small]

Matthew Gillen <m...@mattgillen.net> writes:

> On 4/3/2016 2:49 AM, Rich Pieri wrote:
>> On 4/2/2016 10:20 PM, Matthew Gillen wrote:
>>> That would satisfy the anti-tivoization and be within the limits of the
>>> GPLv3, while still causing a problem for the FBI in this particular
>>> instance.
>> 
>> Incorrect on both counts. Part 6 prohibits restrictions on consumer
>> devices that prevent or interfere with the continued operation of
>> modified software. Automatically wiping or factory resetting the device
>> definitely constitutes interference.
>
> That is quite debatable.  Auto-bricking the phone would definitely count
> as interfering with the device.  Erasing protected storage that does not
> render the device unusable (even if, for instance, it made it so you
> could never talk to iTunes again), would not necessarily constitute
> interference.

"“Installation Information” for a User Product means any methods,
procedures, authorization keys, or other information required to install
and execute modified versions of a covered work in that User Product
from a modified version of its Corresponding Source. The information
must suffice to ensure that the continued functioning of the modified
object code is in no case prevented or interfered with solely because
modification has been made.

If you convey an object code work under this section in, or with, or
specifically for use in, a User Product, and the conveying occurs as
part of a transaction in which the right of possession and use of the
User Product is transferred to the recipient in perpetuity or for a
fixed term (regardless of how the transaction is characterized), the
Corresponding Source conveyed under this section must be accompanied by
the Installation Information. But this requirement does not apply if
neither you nor any third party retains the ability to install modified
object code on the User Product (for example, the work has been
installed in ROM)."


My reading of this is that Apple has to give their customer control of
their keys or other authorization information in a way that the customer
could make changes and pass those along to someone else, along with
whatever keys and authorization are necessary to install and run the
distributed software (but not the same keys to get at that person's data
on a device they keep), without any autodestruction happening.  When the
FBI takes possession of a person's, especially a late person's, device
and software they aren't being conveyed anything. They've simply taken
hold of someone's device and software.

All Apple using GPLv3 would prevent would be having autodestruction of
the sofware or device when a user willingly (or unwillingly but
actively?) "conveys" the software (possibly modified) to the FBI along
with required authorization information, as they are required to do by
the license.  If the user were under duress he or she could simply
violate the terms of the GPL, give a wrong key and hope the drive gets
effectively erased or the device destroyed. No problem here, yeah? I
mean, what's the thought process, "do I incriminate myself to the feds
or do I not but risk a lawsuit from Apple or some upstream copyright
holder for violating their copyright, having fallen out of protection
from the terms of the GPL."  It's too ridiculous to even think about. I
wish I hadn't written it now.

-- 
Mike Small
sma...@sdf.org
___
Discuss mailing list
Discuss@blu.org
http://lists.blu.org/mailman/listinfo/discuss

Re: [Discuss] Apple FUD?

[Mike Small]

IngeGNUe  writes:

> I support some Macs in my day job and I was troubleshooting the Mac not
> writing to NTFS; then I thought, oh, let me just edit /etc/fstab, right?
> Well...apparently, Apple says this is "experimental" and unsupported and
> maybe maybe maybe will break something.
>
> That strikes me as weird because /etc/fstab has been around forever. Is
> this just FUD, is Apple prone to FUD? Or...have they broken
> compatibility with such things as you would see in GNU and BSD?

I don't use Macs currently but it doesn't sound like FUD to me so much
as a matter of their OS happening to have been built upon BSD, what,
cause that other thing didn't work out and the BeOS guy wanted more
money or something, but they're free, even more so than Redhat
employees, to do whatever the hell they want. If it suits them to keep a
particular aspect of the Unixy layer familiar to Unix users great. If
instead they want to hand disk layout over to a chunk of closed C++
system code instead, without troubling themselves to test if if said
chunk will choke on your GNU or BSD inspired edits, well, at least they
warned you about it.

But you must have missed a memo. Apparently we're not supposed to
question whether a Mac is good as a Unix system because if you don't
like anything about it you can always run a different operating system
in a virtual machine running on it. (Sorry, I seem to have left a nerve
exposed.)




___
Discuss mailing list
Discuss@blu.org
http://lists.blu.org/mailman/listinfo/discuss

Re: [Discuss] Local nonprofit that accepts hardware donations

[Mike Small]

Could also have been the Free Computer Labs announcement from last
spring he's thinking of. Julian, you still reading the list?

Stephen Ronan  writes:

> That was probably the South End Technology Center @ Tent City, 
> operated by Mr. Mel King, 359 Columbus Ave, Boston MA 02116. 617 
> 578-0597. At least for now, they've scaled back a lot on 
> their hardware refurbishing program. A key organizer is out of 
> the country for a while.
> This program at Castle Square Apartments, also in the South End, 
> would be another possibility:
> http://cstoboston.org/communityprograms/squaretech/
> I've volunteered with those organizations and others connected 
> with low-income housing developments in Boston and can generally 
> find a good home for a working computer, but not much else in the 
> way of hardware.
>- Stephen Ronan
>
>
> On Fri, 1 Apr 2016, Joseph Guarino wrote:
>
>>
>> Hi everyone,
>>
>> I recall there was a local non-profit that accepted hardware 
>> donations and did free FOSS training.  I tried to look it up to 
>> no avail.  Anyone know of a local non-profit that does such 
>> things?  I've a bunch o' hardware I'd like to donate.  Enjoy 
>> the weekend!
>>
>> Thanks,
>> Joseph
>> ___
>> Discuss mailing list
>> Discuss@blu.org
>> http://lists.blu.org/mailman/listinfo/discuss
>>
> ___
> Discuss mailing list
> Discuss@blu.org
> http://lists.blu.org/mailman/listinfo/discuss
>
___
Discuss mailing list
Discuss@blu.org
http://lists.blu.org/mailman/listinfo/discuss

Re: [Discuss] issue with mysql in a VM on Ubuntu 14.04.4 LTS

[Mike Small]

Jerry Feldman  writes:

> My friend got back with some information.
>
> This is a dedicated web server VM (virtual machine) running Ubuntu 14.04
> LTS.  This is only running a single site, yet CPU usage reaches 100%
> with only a few simultaneous users viewing the site.  Looking deeper it
> appears that MySQL is using the majority of the CPU resources.  When
> this happens, the hosted website is extremely slow or unavailable. 
> Restarting the Apache and MySQL services temporarily helps, but a reboot
> seems to have a longer effect. 
>
> In troubleshooting the issue we have made several changes to various
> conf files to increase buffers, caches, memory limitations, etc.

What about DB level stuff?  Can you narrow it to a particular query?
Have indexes changed?  Does MySQL have some concept of statistics that
help the query optimizer and can get stale?

___
Discuss mailing list
Discuss@blu.org
http://lists.blu.org/mailman/listinfo/discuss

Re: [Discuss] Encrypt Everything? Good Luck With That

[Mike Small]

Bill Horne <b...@horne.net> writes:

> contribution from Apple. It's also possible that Apple's execs wanted 
> some free ink and to boost the iPhone's reputation for security, and 
> that everyone inside the beltway knew how this would play out months ago.

Well that blew up in their faces. Oh, maybe that will be in the next
iThing. If you try to open it up it blows up in a wisp of stylist white
and silver dust, perhaps briefly coalescing into a sad mac image before
dissipating.

Off topic to this thread, but rms is speaking in Kendall at 2:30
today. Wish I could go but I have a meeting I feel I shouldn't blow off
for some reason.

https://www.fsf.org/events/rms-20160329-cambridge

-- 
Mike Small
sma...@sdf.org
___
Discuss mailing list
Discuss@blu.org
http://lists.blu.org/mailman/listinfo/discuss

Re: [Discuss] systemd explanations

[Mike Small]

Bill Ricker <bill.n1...@gmail.com> writes:
> ​Some of Perl's harshest critics are its most ardent supporters.​
> And unlike some communities, feedback works.

And those I would have no problem with. It's rather people advocating
competing languages in the same family (including the author of one
popular programming text) who genuinely dislike Perl who cause me
trouble, specifically in that their wisecracks hurt my motivation to
learn the languages they advocate.  Despite consciously knowing it's
stupid, I struggle not to mentally take up this advocacy battle to the
point that in my spare time I find myself wanting to learn any language
but Python, a language that does seem to have some advantages, if
nothing else for having a lot of cachet right now.

In fact, thinking about it, it's only when it comes from Python
programmers that Perl bashing messes with my head. For instance, there
was a well known Common Lisp programmer, Erik Naggum, who quite
viciously attacked both Perl and C++, but that somehow never damaged my
interest in Lisp.  I probably wouldn't be too bothered if a Ruby
programmer said something either, that language also seeming (at least
to me) to be going approximately nowhere in terms of its popularity. And
if a Tcl programmer ... well I might give that soul a hug.

I guess this is something of a personal problem. Maybe I should try
attending some of the Python events to cure myself of it.

-- 
Mike Small
sma...@sdf.org
___
Discuss mailing list
Discuss@blu.org
http://lists.blu.org/mailman/listinfo/discuss

Re: [Discuss] systemd explanations

[Mike Small]

Bill Ricker <bill.n1...@gmail.com> writes:
> But yes, the social history of the creator of SystemD and the CF resulting
> from of some of his prior code are two reasons (in addition to 'Unix
> Philosophy' if one subscribes to that) to take at best a
> wait-and-see/show-me attitude to this SystemD 'glorious revolution'.

So can we call Slackware users Jacobites then? It does have a nicer ring
than Slacker.

-- 
Mike Small
sma...@sdf.org
___
Discuss mailing list
Discuss@blu.org
http://lists.blu.org/mailman/listinfo/discuss

Re: [Discuss] systemd explanations

[Mike Small]

IngeGNUe <ingeg...@riseup.net> writes:
> For everything else, GuixSD isn't running systemd so maybe that's worth
> a look :)

I've been keen to try it out again since David's talk but still haven't
gotten around to it. Hopefully this weekend I'll finally get a
Slackware/Guix system going, i.e. start using Guix on my existing
Slackware. But then Slackware's coming out with a new release soon,
which will bring its package selection back into relative modernity.
Heck, they even have pulseaudio now. So the timing for trying Guix may
not be right just yet, at least in terms of using it as a tool to get
a fresher selection of packages.

Still, I'm quite excited about Guix and GuixSD, not because the latter
doesn't run systemd (whatever that is), or because it's more Unixy
(quite the opposite), but because of the Scheme. After years of liking
OpenBSD best, for some strange reason -- maybe an allergic reaction to
all the anti-GNU, anti-GPL, and anti-rms sentiment on BSD mailing lists
and podcasts -- I'd like to try a system with a really strong GNU
flavour to it (beyond emacs).  To have that with guile/scheme as the
primary scripting language, and in a defacto way rather than only as a
statement in a coding standard, that would be very interesting.  Just so
long as this Scheme love doesn't translate into Perl hate. I can't abide
another community rife with Perl bashing.

-- 
Mike Small
sma...@sdf.org
___
Discuss mailing list
Discuss@blu.org
http://lists.blu.org/mailman/listinfo/discuss

Re: [Discuss] Converting "rich" (MIME) email to plain text

[Mike Small]

Michael Tiernan  writes:

> I'm sure that I'm not the first who tried to find an easy way to filter 
> a piece of email so that only the plain text comes out.
>
> I can find lots of things about going plain to HTML but I've not seen 
> anything that allows you to just extract the "Content-Type: text/plain" 
> section of an email.
>
> Any pointers available? I don't want to try and reinvent the reinvented 
> wheel.

Are you willing to use a different agent than Thunderbird? I recall Gnus
having settings you could use to prioritize which alternative should be
displayed. There were even hooks I was able to use to have different
preferences in response to header patterns. mutt has a setting called
alternative_order for this. I've just started using RMAIL (in emacs) and
it seems to get it right without customization, favouring text/plain
over text/html. Now I just need to teach it to automatically delete
the emails that lack text/plain.

Maybe Thunderbird has something like this, perhaps via a plugin, but my
(limited) Thunderbird experience has been uninspiring.

___
Discuss mailing list
Discuss@blu.org
http://lists.blu.org/mailman/listinfo/discuss

Re: [Discuss] Duplicate entries in Gnu PG

[Mike Small]

On Tue, Feb 02, 2016 at 02:40:59PM -0600, Derek Martin wrote:
> On Sat, Jan 30, 2016 at 10:51:18AM -0500, Jerry Feldman wrote:
> > My primary key appears to be duplicated in my GnuPG keyring. I can
> > manually encrypt and decrypt keys.
> > [gaf@gaf ~]$ gpg --list-keys | grep B7F14F2F
> > pub   4096R/B7F14F2F 2014-09-15 [expires: 2017-09-14]
> > pub   4096R/B7F14F2F 2014-09-15 [expires: 2017-09-14]
> 
> Seems like a bug...  However, the trouble with grep is that it filters

Perhaps this one?
https://bugs.gnupg.org/gnupg/issue2193

Unfortunately there's no mention of workarounds or repair procedures.

-- 
Mike Small
sma...@sdf.org
___
Discuss mailing list
Discuss@blu.org
http://lists.blu.org/mailman/listinfo/discuss

Re: [Discuss] bluetooth

[Mike Small]

On Thu, Jan 28, 2016 at 12:07:35PM -0500, dan moylan wrote:
> 
> i have recently been given an ivation boomering bluetooth
> speaker and am fumbling about trying to make use of it.
> i started by looking at:
> 
> https://wiki.archlinux.org/index.php/bluetooth
> 
> but seem to be doing something wrong.  e.g.
> 
> moylan ~[1042] sudo bluetoothctl
> [sudo] password for moylan:
> [NEW] Controller B8:EE:65:34:F2:B4 aldeberon [default]
> [bluetooth]# power on
> Failed to set power on: org.bluez.Error.Blocked
> [bluetooth]# scan on
> Failed to start discovery: org.bluez.Error.NotReady
> [bluetooth]# quit
> [DEL] Controller B8:EE:65:34:F2:B4 aldeberon [default]
> 
> the wiki says to turn the power on with bluetoothctl, but
> that seems not to work.
> 
> what have i missed?

Since no one who knows about this stuff has answered yet I'll ask
if there's any further information in your log files. Those errors
to me sound like bluez is waiting for something else to start up
before it can properly get going.

Unfortunately, the only thing I know about bluez is that it's what
forced Pat Volkerding to finally include PulseAudio in the pending
release of Slackware. Apparently bluez recently made PA a hard
requirement. Say, do you have PulseAudio installed and running
properly?

-- 
Mike Small
sma...@sdf.org
___
Discuss mailing list
Discuss@blu.org
http://lists.blu.org/mailman/listinfo/discuss

Re: [Discuss] bluetooth

[Mike Small]

On Thu, Jan 28, 2016 at 06:13:32PM -0500, dan moylan wrote:
> ran bluetoothctl power on and found this using journalctl -f
> 
> Jan 28 17:59:13 aldeberon bluetoothd[734]: Failed to set mode: Blocked 
> through rfkill (0x12)
> 
> ps shows:
>  2113  1000 S 
> any thoughts?

journalctl must be a systemd thing, eh? Everything I know about systemd
I will have learned in April when Christoph gives his next talk.

This rfkill command / subsystem seems a decent thing to duckduckgo on:
http://manpages.ubuntu.com/manpages/lucid/man1/rfkill.1.html
http://www.mjmwired.net/kernel/Documentation/rfkill.txt

These are old pages, but can you run rfkill, first with the list
argument and then with unblock and a reasonable index or type from
the list output? 

Is it possible that this is a laptop with one of those little
buttons above the keyboard with the button in the state where the
bluetooth radio is off?  This is what _rfkill list_ output looks
like on my machine, first with the wifi button along the top in
the off state and then with it turned on:

4098r0:~$ /usr/sbin/rfkill list
0: hp-wifi: Wireless LAN
Soft blocked: no
Hard blocked: yes
1: phy0: Wireless LAN
Soft blocked: no
Hard blocked: yes
2: hp-bluetooth: Bluetooth
Soft blocked: no
Hard blocked: yes
4099r0:~$ /usr/sbin/rfkill list
0: hp-wifi: Wireless LAN
Soft blocked: no
Hard blocked: no
1: phy0: Wireless LAN
Soft blocked: no
Hard blocked: no
2: hp-bluetooth: Bluetooth
Soft blocked: no
Hard blocked: no
3: hci0: Bluetooth
Soft blocked: no
    Hard blocked: no

-- 
Mike Small
sma...@sdf.org

___
Discuss mailing list
Discuss@blu.org
http://lists.blu.org/mailman/listinfo/discuss

Re: [Discuss] Local ISP Recommendations?

[Mike Small]

On Fri, Jan 22, 2016 at 12:36:15PM +, Edward Ned Harvey (blu) wrote:
> The risk of being hacked is *not* so much the risk of someone
> accessing your mail. It's the risk of someone doing illegal shit
> on your system, and you getting the blame for it. Try 10 years in
> prison, and being permanently registered as a sex offender, probably
> getting divorced, because someone thought that was *your* kiddie
> porn. You find yourself in the position of being presumed guilty,
> having to prove your innocence, because illegal material was indeed
> found in your system, or in your account.
> 

Has this scenario, that someone uses another's system or ip to
download child porn or violate copyrights, ever happened to anyone
in a real legal case where the innocent party wasn't able to
establish his or her innocence? It's something I've wondered about
in considering whether to open a wireless access point for neighbours
and passers by.  Now I don't think either of my laptops' wireless
chips can function as access points but I'm curious about it still:
perhaps in the future I'll have something to offer.  As I recall,
the EFF tended to be quite dismissive of this possibility when they
were promoting the idea that people share their internet connections.

-- 
Mike Small
sma...@sdf.org
___
Discuss mailing list
Discuss@blu.org
http://lists.blu.org/mailman/listinfo/discuss

Re: [Discuss] 4K (or 5K) resolution for Linux desktop

[Mike Small]

On Tue, Jan 05, 2016 at 11:29:05PM -0500, Rich Pieri wrote:
> On 1/5/2016 9:51 PM, Robert Krawitz wrote:
> > I don't agree.  Even if you can't easily resolve the individual
> > pixels, the higher density improves the definition and may allow you
> > to e. g. use a smaller font or smaller icons and be able to read them
> > as easily.
> 
> A 10 point typeface on a 15" display panel at 1080p will be a given
> size. The same 10 point typeface on a 15" display panel at UHD will be
> 1/4 the size, appearing as the equivalent of a 2.5 point typeface at
> 1080p. All the "improved definition" marketing hype in the world won't
> make legible a 2.5 point-looking face at reasonable viewing distances.

Eh? I thought point meant point and that this only happened when
you make the mistake of specifying your font size in pixels, e.g.
pixelsize instead of size in fontconfig language. Though that
seems a common mistake. e.g. I was looking at st (simple terminal
from the suckless project) and its default font is specified in
pixel size giving poor results for me even on an old monitor at
1680x1024.

-- 
Mike Small
sma...@sdf.org
___
Discuss mailing list
Discuss@blu.org
http://lists.blu.org/mailman/listinfo/discuss

Re: [Discuss] 4K (or 5K) resolution for Linux desktop

[Mike Small]

On Wed, Jan 06, 2016 at 11:56:27AM -0500, Rich Pieri wrote:
> On 1/6/2016 10:05 AM, Mike Small wrote:
> > Eh? I thought point meant point and that this only happened when
> > you make the mistake of specifying your font size in pixels, e.g.
> > pixelsize instead of size in fontconfig language. Though that
> 
> What point means is how large a typeface is in print. Points and picas
> are absolute measurements like inches. An inch is an inch. A point is a
> point.
> 
> When translated to computer screens the pixel size and density (ppi)
> does matter because 100 pixels on one screen is not the same absolute
> dimensions as 100 pixels on a different screen. Most contemporary
> desktop environments have mechanisms for scaling so that a 1 inch line
> is 1 inch regardless of the display's ppi -- but most also default to
> 96ppi which means you get the behavior I described.

I'd think one would want to change the default ppi. I remember it
being 75x75 (or was it 72x72) in the past and over time went up as
new monitors came out with higher resolutions. This should happen
again, yeah? But if these things are short and wide would the new
ppi have to be 100x200 dpi or something not symmetric? I bet there
are bugs to be found from that.

> Which brings me back to the point I made about screen size: you have to
> upscale everything on a UHD screen in order to make everything appear to
> be the same size as it would appear on a 1080p screen with the same
> dimensions. Which is to say, the 4 times greater resolution of UHD is
> wasted if you need to make everything 4 times larger in order to achieve
> consistency.

By "upscale everything" you mean setting the ppi to an accurate
value?  Well yeah I'm also doubtful that fonts at the higher
resolution (and ppi setting) on a "normal size" monitor would
look better enough to justify the present cost of these things...

> 
> Or you can use a physically larger display. A UHD display needs to be
> about 4 times larger (twice as wide, twice as tall) as a 1080p display
> to achieve identical (or close enough) absolute sizes of displayed
> objects without scaling. If you are comfortable with a 17" display at
> 1080p then you will need a 35" UHD display to achieve a similar level of
> comfort without scaling.

... but yes on a very big monitor it sounds possibly worthwhile.
I haven't seen X on a really large monitor, but I'd guess fonts
look less good.

It's a nice thought having a huge monitor, what, now that I use
Rob Pike's acme as a text editor at times. The screen gets a little
crowded once you get beyond seven or eight panes. On the other
hand, that's the number beyond which most humans have trouble
keeping track of things so maybe not so useful. It would be like
having a larger apartment.  I'd spend more time looking around this
vast screen trying to find where I (or acme or a window manager)
put things.

-- 
Mike Small
sma...@sdf.org
___
Discuss mailing list
Discuss@blu.org
http://lists.blu.org/mailman/listinfo/discuss

Re: [Discuss] Linux on laptops

[Mike Small]

On Thu, Nov 12, 2015 at 02:11:19PM +, Edward Ned Harvey (blu) wrote:
> So far, what I'm inclined to do, is go to a local store such as Microcenter 
> or BestBuy, ask them what their return policy is, create a "dd" image of the 
> internal hard drive before first power-on, and then simply blow it away with 
> a linux installer. See what happens. 

Do any stores let you boot from a live cd in the store?  Some must.
I seem to recall an OpenBSD user, can't remember where, accumulating
a pile of dmesges of different laptops by going to certain stores
and trying out what they have.

-- 
Mike Small
sma...@sdf.org
___
Discuss mailing list
Discuss@blu.org
http://lists.blu.org/mailman/listinfo/discuss