Re: Uninstalling a package and its entourage

[Mike Castle]

Even shorter:

apt autopurge


Apropos to my recent message regarding system configuration, I keep a
personal metapackage around that lists the packages I really want.
About once a quarter I do the following (as root):

# apt-mark showmanual | grep -v mrc-mars | xargs apt-mark auto
# apt autopurge

(Where mrc-mars is my top level metapackage.  For *me*, that is the
only manual package I ever want on a machine.)

If I see something that it wants to purge but I do actually want, I
add it to my personal metapackage (build and push) and then do the
above again.

mrc

Re: Current best practices for system configuration management?

[Mike Castle]

On Sat, Apr 20, 2024 at 4:40 PM Mike Castle  wrote:
> Thanks for all of the commentary so far.
>
> Once I get something working, I will *try* to remember to follow up
> here with what I've managed to cobble together.

I have done quite a bit of research and experimentation and finally
settled on a solution that seems like it will work for me:

Plain old debian/* control files along with config-package-dev.

Effectively, I've abandoned `equivs` and now just using plain old
`debhelper` with a small wrapper script.

One thing I've learned along the way is that debian/control files can
build multiple .deb files.  But the "control" files that `equivs` uses
are not really the same as regular control files, so need one per
package.  Since I created a hierarchy of packages, I needed several
configs for equivs.  Now, I can use just one, and it is MUCH faster
(about 8 seconds total).

I'm still working on replacing my old equivs stuff, but, I do have
some things I'm installing now.  Essentially my setup looks like this:
debian/control:
Maintainer: Mike Castle 
Source: mrc
Build-Depends: debhelper-compat (= 12), config-package-dev
Rules-Requires-Root: no
Standards-Version: 4.5.1
Section: metapackages
Priority: optional

Package: mrc-base
Architecture: all
Depends: ${misc:Depends},
 bc,
 cron,
 ...
 zip,

Package: mrc-mozilla
Architecture: all
Description: MRC's Mozilla apt configuration
 Lots of files under /etc/apt.
Depends: ${misc:Depends},

Package: mrc-desktop
Architecture: all
Description: MRC's desktop installation
 Graphical stuff.
Depends: ${misc:Depends},
 mrc-base,
 mrc-games,
 mrc-python,
 mrc-mozilla,

... more meta packages ...

Package: mrc-mars
Architecture: all
Description: MRC's host mars
 Host specific deps.
Provides: ${diverted-files}
Conflicts: ${diverted-files}
Depends: ${misc:Depends},
 mrc-desktop,
 mrc-development,
 mrc-virtual-machines,

... most host packages ...

A few other debian control files:
$ find debian/ -type f
debian/source/format
debian/mrc-mars.install
debian/mrc-mars.displace
debian/mrc-mozilla.install
debian/copyright
debian/changelog
debian/rules
debian/control

$ cat debian/rules
#!/usr/bin/make -f

%:
dh $@ --with=config-package

The files that match .* are the magic.  For example:
$ cat debian/mrc-mozilla.install
files/mozilla/* /

$ find files/mozilla -type f
files/mozilla/usr/share/lintian/overrides/mrc-mozilla
files/mozilla/etc/apt/sources.list.d/mozilla.list
files/mozilla/etc/apt/preferences.d/mozilla
files/mozilla/etc/apt/keyrings/packages.mozilla.org.asc

The etc/apt files come from
https://support.mozilla.org/en-US/kb/install-firefox-linux#w_install-firefox-deb-package-for-debian-based-distributions
and the lintian overrides is:
$ cat files/mozilla/usr/share/lintian/overrides/mrc-mozilla
mrc-mozilla: package-installs-apt-sources
mrc-mozilla: package-installs-apt-preferences

For the host packages, it is slightly more complicated, but not much
once I figured it out.  Again, a basic "copy in everything under this
directory" install file:
$ cat debian/mrc-mars.install
files/mars/* /

which is currently this:
$ find files/mars -type f
files/mars/etc/hostname.mrc

Then the interesting bit:
$ cat debian/mrc-mars.displace
/etc/hostname.mrc

This "dispace" file is something that the `config-package` addon takes
care of.  It sets up backing up the existing file name (/etc/hostname)
and symlinks in my version:
$ ls -l /etc/hostname*
lrwxrwxrwx 1 root root 12 May 27 08:57 /etc/hostname -> hostname.mrc
-rw-r--r-- 1 root root  5 May 27 06:52 /etc/hostname.mrc
-rw-r--r-- 1 root root  6 May 27 07:37 /etc/hostname.mrc-orig

If I purge mrc-mars, then /etc/hostname.mrc-orig is moved back to /etc/hostname.

(The way config-package-dev works is, it takes the first portion of
the package names, in this case "mrc", to identify what file names to
use.)

To drive it all, I currently use a simple wrapper script:
$ cat doit.sh
#!/bin/bash

set -e

work_dir=$(mktemp -d)
find -depth | cpio -pdm ${work_dir}/work
cd ${work_dir}/work
debuild --no-conf --no-sign --lintian-opts --info

OUTPUT=/srv/deb/packages
rm -rf $OUTPUT
mkdir -p $OUTPUT
cp ${work_dir}/*.deb $OUTPUT
cd $OUTPUT

dpkg-scanpackages . > Packages

echo "work_dir was ${work_dir}"

Not the fanciest, but "works for me" (so far).  Essentially it builds
things in /tmp because the debian packaging tool chain always drops
the files into the parent directory and leaves things that *I*
currently do not care about littering my disk.  Also, it drops
intermediate files into the debian/* directory, making it more
difficult to know what needs to go under SCM.  With this approach,
"git status" will be easier to read.

The only drawback so far, over my previous approach of one `equivs`
config per package, is that now, all built debs share the same version
number.  With the individual approach, I could add one package to a
metapackage, and when I do &qu

Re: Current best practices for system configuration management?

[Mike Castle]

Hah!

https://lists.debian.org/debian-user/2013/08/msg00042.html

Re: Current best practices for system configuration management?

[Mike Castle]

Thanks for all the suggestions so far.

Like Alex, one of my physical machines is a laptop that is not always
on the home network.  Though I'm usually connected to *something*.
I'm still debating whether to bother with a VPN or trying something
like a tailnet.

Heck, before I adopted Debian and ran my own
Linux-from-Scratch-started-before-LFS-or-Debian-existed
"distribution", I used to run $HOME over NFS over WiFi.  Sharing
configs that way would be acceptable for me.  I also used to keep my
configs in CVS and pushed them using rsync-over-ssh-as-root, hence
"current best practices".  :->

Linux-Fan's MDVL stuff looks interesting.  Everything can always be
solved by one more layer of redirection.  :->  I just have to suss out
the Java and Ant stuff to understand the basics.  Personally, I'm a
big fan of NIH-Syndrome (otherwise I wouldn't have run a LFS type
system for ~15 years).  So it is likely I would reinvent what is
happening there.

One thing Linux-Fan mentioned was `config-package-dev`.  In my OP, I
commented about ``slightly old to really old tools'', and that was one
I was thinking of.  It looks like it hasn't been touched in seven
years, and I wasn't sure if it still worked.  But that drive by
comments lends some hope.  Using it would help address Alex's concern
about modifying existing config files.  That debhelper extension is
designed precisely for that situation.  But, its age is pretty much
what inspired me to start this thread.

Interestingly enough, I sent the OP just before I went to a BayLISA
talk about Apple's PKL.  Since I tend to only use whatever comes out
of the Debian repos (my only exceptions being Firefox and an emacs
package the maintainers won't fix for bookworm), I'm unlikely to do
more than look at it,  But, between it and MDVL, I think I'm
definitely going to try to make sure I don't configure myself into a
corner.  :->

Thanks for all of the commentary so far.

Once I get something working, I will *try* to remember to follow up
here with what I've managed to cobble together.

Cheers!
mrc

PS: Actually, I used to share $HOME (and /usr) over PLIP, so, it is
probably obvious that FS speed is not always a concern for me.

Current best practices for system configuration management?

[Mike Castle]

For a while now, I've been using `equivs-build` for maintaining a
hierarchy of metapackages to control what is installed on my various
machines.  Generally, I can do `apt install mrc-$(hostname -s)` and
I'm golden.

Now, I would like to expand that into also setting up various config
files that I currently do manually, for example, the `/etc/apt/*`
configs I need to make the above work.  For a single set of files,
manual isn't bad, but as I want to get into setting up LDAP, autofs,
and so on, it is time to explore solutions.  I only have four systems
at the moment (two physical and two virtual), so I don't think I need
something too fancy.

My first thought was to simply add a `Files:` section to *.control
files I use for my metapackages.  After all, for configs going into
*.d directories, they are usually easy to just drop in and remove, no
editing in place required.  But, that is when I discovered that all
files under `/etc` are treated specially.

I've found a lot of documentation out there, but, of course, much of
it is out of date.  https://wiki.debian.org/ConfigPackages , for
example, seems to recommend slightly old to really old tools.  Tools
like `ansible`, `puppet` and so on seem, at first blush, aimed at
larger installations than mine.  But maybe other's experience with
them will show they scale down fine?

Anyway, suggestions based upon actually experience would be appreciated.

Cheers,
mrc

Re: Inclusive terminology (instead of master/slave) for network bonding/LACP

[Mike Castle]

On Fri, Mar 15, 2024 at 1:49 AM Alain D D Williams  wrote:
> We seem to be told that this must be done by those who will not be doing the
> work.

Was that explicitly stated anywhere?  Or is the lack of any type of
explicit "I'm willing to help drive this" statements leading to that
conclusion?

mrc

Re: Inclusive terminology (instead of master/slave) for network bonding/LACP

[Mike Castle]

On Fri, Feb 23, 2024 at 2:07 AM Alain D D Williams  wrote:
> It is "fixing" an issue for today's English speakers. Should we scour our
> systems looking for similar issues in other languages ? Then in, say, 20 years
> time when different words will then be considered offensive, by some, do this
> all again ?

Yes.

Re: top bar the way I want it

[Mike Castle]

On Sat, Jan 20, 2024 at 9:12 AM Jeremy Nicoll
 wrote:
> And, of course, write notes to yourself for EVERY change like this, so
> you can remember how you did it.

I actually have a quarterly reminder for myself to review my various
systems and take notes on changes.  Installed packages, make sure
config files are captured in source control, was I running any A/B
experiments to see if I like a new font better than the old one, etc.
While most of this I do as I go along, I find having a regular true-up
is useful to make sure I didn't miss anything.

This applies to my computers, phones, car gadgets, kitchen layout, etc.

mrc

Re: disable auto-linking of /bin -> /usr/bin/

[Mike Castle]

On Wed, Jan 10, 2024 at 10:57 AM  wrote:
> Yes, the main reason for the separation of /usr has more or less
> disappeared with the arrival of initramfs, but still... why.

To some extent, it will make it easier for packaging.

Look at any package built using autoconf, for instance, you run:
./configure --prefix=/usr

Well, except for those you want installed in /, in which case you use
'--prefix=/'

But, what if you don't want all of those in /, you want some in / and
some in /usr?  Requires more manual work on part of the package
maintainer to make sure that things work properly, that files are
split across the destinations properly, reference config files
appropriately, and so on.

With usrmerge, that particular class of problems goes away.

Back when I used my own home grown distribution (I was doing Linux
>From Scratch before LFS was even a thing), that was one of the issues
I'd run into every once in a while.

mrc

Re: disable auto-linking of /bin -> /usr/bin/

[Mike Castle]

On Wed, Jan 10, 2024 at 9:53 AM Andy Smith  wrote:
> น There was another use-case which is "sharing a read-only /usr
>   between systems by NFS, etc." but at the time this was widely
>   regarded a lost cause as so many other things violated the
>   premise.

I did that for years.

Then again, when I started doing that, I was using PLIP over a
null-printer cable.  But even after I could afford larger harddrives
(so I had room to install /usr), and Ethernet cards (and later a hub),
I still ran /usr over NFS.

Personally, I'm rather saddened by usrmerge.  But, such is life.

mrc

Re: Content of /etc/ethers

[Mike Castle]

On Wed, Jan 3, 2024 at 6:58 PM Greg Wooledge  wrote:
> What's not really stated anywhere is *why* these library functions
> exist.  I don't see many practical application for a library function
> that reads a text file full of MAC addresses and hostnames, looks up
> one of them, and spits out the other half of the line it's on.  You'd
> get more usefulness just doing "grep somename /etc/ethers".
>
> The Debian ethers(5) page references arp(8), so one might conclude
> that this file is intended to augment/prettify the output of "arp" so
> that it contains hostnames in addition to (or instead of) MAC addresses.
> But the main use of the "arp" command has always been to find out the MAC
> address of a host whose IP address you already know (or can get from DNS),
> but whose MAC address is not currently known.  So, if you've already got
> a text file full of these MAC addresses, why would you even need to run
> the arp command in the first place?

IPv4 and IPv6 are not the only network protocols that can be used on
Ethernet.  Nor were they the first, and definitely not the only ones
on unix-like systems.

One can do raw ethernet packets.  Or IPX, NetBIOS (used by SMB,
Lantastic, and others), AppleTalk, Banyan's VINES, PPPoE, DECnet, ,
probably XNS.

Probably many I missed, and perhaps some future enet based protocols
that are not inet based.

Reading https://en.wikipedia.org/wiki/Xerox_Network_Systems:
> XNS also helped to validate the design of the 4.2BSD network subsystem by 
> providing a second protocol suite, one which was significantly different from 
> the Internet protocols; by implementing both stacks in the same kernel, 
> Berkeley researchers demonstrated that the design was suitable for more than 
> just IP.

The getent(1) support ethers, so you can probably use that to test any
information you put in there.  As mentioned in the getent man page,
ethers is one of the databases supported by NSS.

mrc

Re: Recommended simple PDF viewer to replace Evince

[Mike Castle]

I use Evince probably once a week or so from the command line.  I do
not see that error, though I think I have in the past.  I suspect that
if you are seeing those issues with the current bookworm release, it
is likely a problem local to you.

You could be missing a package that evince expects to be there, but
there is a missing dependency (likely, making it a Debian problem).
You could have configured something some time ago that no longer makes
sense, either Gnome based as a whole, something local to Evince, etc,
or a cache corruption (which makes it a local issue which could be a
challenge to track down).  Or a good old fashioned bug somewhere along
the line (perhaps making it a Gnome/Evince bug).

One URL from the man page, http://www.gnome.org/projects/evince/,
seems to go through a series of JavaScript based redirects to end up
at https://apps.gnome.org.  That does give a GitLab based 404 error.
I'm not sure if that is the "throws an error" situation you were
referring to or not.  Possibly related, https://circle.gnome.org gives
the same 404.  Both of those links are listed in the footer of
https:///www.gnome.org/.

However, there is another site listed in the man page:
https://gitlab.gnome.org/GNOME/evince/issues , and that does appear to
work.  That might be another venue you can try if you want to resolve
the issue instead of abandoning Evince.

As far as alternatives, I think both current versions of Firefox and
Chrome support PDFs natively.  Also, before I started using evince, I
used to use gv (based on ghostview) quite a bit.  The following seems
to list most of the various programs discussed in this thread, plus a
couple of others:

apt-cache search pdf-viewer

mrc

Re: Performance of my computer

[Mike Castle]

On Mon, Oct 30, 2023 at 5:14 PM Van Snyder  wrote:
> On Mon, 2023-10-30 at 19:40 +, piorunz wrote:
> On 30/10/2023 18:56, Van Snyder wrote:
> Firefox, in every version I've used so far, appears to have memory
>
> leaks. If I kill it, not by clicking its little "X" or Alt-F4, but with
>
> "kill -9", so that it reopens everything when I restart it, my memory
>
> usage immediately drops by 75%. Then it creeps back up.
>
>
> Firefox doesn't have any memory leaks. It actively uses buffers, cache,
>
> filling available memory. I have Firefox running for days, sometimes
>
> weeks. On slow laptop, and fast workstation PC. Same result, no crashes,
>
> no memory leaks.
>
>
> Then why does it use 1/3 as much memory to display the same pages and tabs 
> when I kill it and restart it? That's a symptom of memory leakage.

Do you happen to keep the Web Developer Tools open?  Or console logs
set to persistent?  Those will keep lots of things in memory just in
case you need them as a developer trying to debug an issue.
Generally, if I notice my memory usage going up, I will be sure to
close the Dev Tools and reload the tab a couple of times.  Then,
eventually, the JavaScript garbage collector will kick in and start
releasing the items the Dev Tools have stopped tracking.

FF also has built in performance tools that can be used to determine
what may be using resources.  Menu -> More tools -> Task Manager (aka,
shift-esc or about:processes).

Nothing Debian specific here.
https://support.mozilla.org/en-US/kb/task-manager-tabs-or-extensions-are-slowing-firefox#w_task-manager
has more details.

mrc

Re: A file synchronization tool that respects hardlinks

[Mike Castle]

rsync supports hardlinks.

--hard-links, -H preserve hard links

Though, in general, the purpose of something like darcs is to
*provide* the syncing.

mrc

Re: Single-page application [was: debian.org - broken Download link.]

[Mike Castle]

On Mon, Oct 9, 2023 at 6:11 AM  wrote:

> Gah, no. As a user I hate those with all my guts. Page "state" is
> distributed in some intransparent way across client and server and
> there is no way to refer to "something" via an URL.

Many modern SPAs track state via URL, so they can be referenced.  And
not just as a ?query or #fragment either.

mrc

Re: Git for backup storage

[Mike Castle]

Something I played with recently was
https://packages.debian.org/stable/vcs/git-filter-repo

But you definitely want to run tests on real data before you decide
that deleting old data saves your anything, particularly with respect
to time.

If git is so efficient at storing this kind of data, then what do you
expect to gain by deleting old stuff, outside of a smaller log to go
through?

mrc

Re: How can I find packages manually installed using "dpkg -i"?

[Mike Castle]

Oops.  The 'grep -v -F' should be 'grep -v -f'.  Well, 'grep -v -F -f'
would probably be appropriate as well.

mrc

On Tue, Oct 3, 2023 at 7:58 PM Mike Castle  wrote:
>
> Some tools I've been using lately are apt-mark and "dpkg-query --show".
>
> The following UNTESTED commands (ran as a normal user):
>
> (apt-mark showauto ; apt-mark showmanual) > apt-thinks-you-installed.txt
> dpkg-query --show --showformat='${Package}\n' | grep -v -F
> apt-thinks-you-installed.txt > rest.txt
>
> The file "rest.txt" should have a list of packages installed that were
> NOT installed via apt.  With any luck, it is small enough to examine
> manually.
>
> You could do something like "apt list" to get a list of all packages
> known by apt and see if you'd prefer to use just use the Debian
> instead of Mint versions.  And anything not in that list *probably*
> came from other manual sources and you can do what you will with that
> information.
>
> You could poke around in /var/lib/apt/lists/ and see if the files from
> the mint repos you used in the past are still there (I don't know if
> they get cleaned up or not, might get lucky).
>
>
> Regarding the comment in the thread about packages that the installer
> added that show up as manual, you can do something like the following
> to at least make apt think they were auto:
>
> dpkg-query --show --showformat='${Package} ${Priority}\n' | awk '$2 ==
> required {print $1}' > required.txt
> sudo apt-mark auto $(apt-mark showmanual | grep -F required.txt)  #
> apt-mark will prompt, so you don't want to use xargs
>
> Again, the above is untested, so verify first!
>
> You might do the same for other priorities, like  standard or
> important.  If for no other reason than breaking the list of packages
> into smaller, digestible chunks that you can focus on.  For example,
> on my machine:
> $ dpkg-query --show --showformat='${Priority}\n' | sort | uniq -c | sort -n
>   5 extra
>  29 important
>  29 standard
>  33 required
>1472 optional
>
> I could probably handle going through those smaller collections to
> identify where they came from fairly easily.  But that big optional
> collection, not so much.  For something like that, I might add
> ${Section} to the --showformat option, and divide them up that way.
>
> Also, as a future project, you might consider creating metapackages to
> help organize your installation.  Again, for my machine:
> $ apt-mark showmanual | wc -l
> 1
> $ apt-mark showauto | wc -l
> 1563
>
> I have a handful of debian control files that I use (base, desktop,
> dev, serviceX, serviceY, machine1, machine2,...).  The machine ones
> depends on the services they host (NFS, LDAP, VMs), and whether they
> need a GUI (desktop), whether I build on them (dev), or play games,
> etc.  Then each machine, after a base install I do something like:
>
> apt-mark auto $(apt-mark showmanual)
> apt install machineN
> apt autoremove --purge
>
> Of course, I monitor that autoremove to make sure it doesn't do
> anything silly, and if it tries to remove a package I missed, I go add
> it to the appropriate control file.  My simple little way of doing
> this is:
>
> $ cat doit.sh
> #!/bin/bash
>
> for v in *.control; do
>   equivs-build $v > $v.log &
> done
>
> echo 'Waiting'
> wait
> echo 'Done waiting'
>
> OUTPUT=/srv/deb/packages
> rm -rf $OUTPUT
> mkdir -p $OUTPUT
> cp *.deb $OUTPUT
> cd $OUTPUT
>
> dpkg-scanpackages . > Packages
> $ cat /etc/apt/sources.list.d/mrc-home.list
> deb [trusted=yes] file:/srv/deb/packages ./
>
> And yes, I should do better than the [trusted=yes].
>
> Good luck on your upgrade!
> mrc

Re: How can I find packages manually installed using "dpkg -i"?

[Mike Castle]

Some tools I've been using lately are apt-mark and "dpkg-query --show".

The following UNTESTED commands (ran as a normal user):

(apt-mark showauto ; apt-mark showmanual) > apt-thinks-you-installed.txt
dpkg-query --show --showformat='${Package}\n' | grep -v -F
apt-thinks-you-installed.txt > rest.txt

The file "rest.txt" should have a list of packages installed that were
NOT installed via apt.  With any luck, it is small enough to examine
manually.

You could do something like "apt list" to get a list of all packages
known by apt and see if you'd prefer to use just use the Debian
instead of Mint versions.  And anything not in that list *probably*
came from other manual sources and you can do what you will with that
information.

You could poke around in /var/lib/apt/lists/ and see if the files from
the mint repos you used in the past are still there (I don't know if
they get cleaned up or not, might get lucky).


Regarding the comment in the thread about packages that the installer
added that show up as manual, you can do something like the following
to at least make apt think they were auto:

dpkg-query --show --showformat='${Package} ${Priority}\n' | awk '$2 ==
required {print $1}' > required.txt
sudo apt-mark auto $(apt-mark showmanual | grep -F required.txt)  #
apt-mark will prompt, so you don't want to use xargs

Again, the above is untested, so verify first!

You might do the same for other priorities, like  standard or
important.  If for no other reason than breaking the list of packages
into smaller, digestible chunks that you can focus on.  For example,
on my machine:
$ dpkg-query --show --showformat='${Priority}\n' | sort | uniq -c | sort -n
  5 extra
 29 important
 29 standard
 33 required
   1472 optional

I could probably handle going through those smaller collections to
identify where they came from fairly easily.  But that big optional
collection, not so much.  For something like that, I might add
${Section} to the --showformat option, and divide them up that way.

Also, as a future project, you might consider creating metapackages to
help organize your installation.  Again, for my machine:
$ apt-mark showmanual | wc -l
1
$ apt-mark showauto | wc -l
1563

I have a handful of debian control files that I use (base, desktop,
dev, serviceX, serviceY, machine1, machine2,...).  The machine ones
depends on the services they host (NFS, LDAP, VMs), and whether they
need a GUI (desktop), whether I build on them (dev), or play games,
etc.  Then each machine, after a base install I do something like:

apt-mark auto $(apt-mark showmanual)
apt install machineN
apt autoremove --purge

Of course, I monitor that autoremove to make sure it doesn't do
anything silly, and if it tries to remove a package I missed, I go add
it to the appropriate control file.  My simple little way of doing
this is:

$ cat doit.sh
#!/bin/bash

for v in *.control; do
  equivs-build $v > $v.log &
done

echo 'Waiting'
wait
echo 'Done waiting'

OUTPUT=/srv/deb/packages
rm -rf $OUTPUT
mkdir -p $OUTPUT
cp *.deb $OUTPUT
cd $OUTPUT

dpkg-scanpackages . > Packages
$ cat /etc/apt/sources.list.d/mrc-home.list
deb [trusted=yes] file:/srv/deb/packages ./

And yes, I should do better than the [trusted=yes].

Good luck on your upgrade!
mrc

Re: XFCE4 without panels

[Mike Castle]

I just tried this in a VM and it seemed to work.

>From a command line:
xfce4-panel -q
find ~/.config | grep panel

Remove the xfce4-panel.xml   (I also removed the empty directory just
named panel.)

The lack of panels seems to have survived a reboot.

I don't know if it is sufficient for every variation involving saving
session state or what not.  But might get you started in the right
direction.

If worst comes to pass, you might be able to put the xfce4-panel -q
command in a shell script that automatically launches when you log in.

Good luck!
mrc

Re: is it unusual that 12.1 is released so soon after 12?

[Mike Castle]

I think it is kind of like buying a new ANYTHING.

Some folks will buy a new model as soon as it comes out.

Some will wait a few months to see if anyone else is having problems with it.

Whether it is a vehicle, electronic device, refrigerator, MS-Windows,
new online service, etc.

As more folks use the ANYTHING, more issues will be found.
Fortunately with something like operating systems, changes/fixes can
usually be made fairly quickly.

mrc

Re: is it unusual that 12.1 is released so soon after 12?

[Mike Castle]

7: 2013-05-04  https://www.debian.org/News/2013/20130504
7.1: 2013-06-15  https://www.debian.org/News/2013/20130615
42 days

8: 2015-04-25  https://www.debian.org/News/2015/20150426
8.1: 2015-06-06  https://www.debian.org/News/2015/20150606
42 days

9: 2017-06-17  https://www.debian.org/News/2017/20170617
9.1: 2017-07-22  https://www.debian.org/News/2017/20170722
35 days

10: 2019-07-06  https://www.debian.org/News/2019/20190706
10.1: 2019-09-07  https://www.debian.org/News/2019/20190907
63 days

11: 2021-08-14  https://www.debian.org/News/2021/20210814
11.1: 2021-10-09  https://www.debian.org/News/2021/20211009
56 days

12: 2023-06-10  https://www.debian.org/News/2023/20230610
12.1: 2023-07-22  https://www.debian.org/News/2023/20230722
42 days

Re: Wireless temperature & humidity measurement

[Mike Castle]

I was just researching this myself a couple of days ago, and spent
several hours going down a rabbit hole.

It seems that many folks are going the way of using an open source
solution, Home Assistant (aka, HA), (https://www.home-assistant.io/).
Even to the point where I found that folks that used to have
standalone code that could maybe read a sensor was migrated over to HA
and then only supported there.  HA really wants to be the only thing
installed on a machine, to the point where if you do install it on
Debian, if you add any other packages, they won't support you.  But,
installing their HA Operating System (HAOS), which appears to be
Debian based, can be with any number of VMs, some stuff via OCI
compatible stuff (like Docker, and I suppose, Linux containers, though
I'm less sure of this).

All of that is overkill for what you (and I) are looking for:  simple
scraping abilities.

But, it does have a large list of supported hardware and protocols.
So, you could go through the hardware list, find something you like,
and then extract the appropriate software bits.  Or at least identify
what software packages are necessary.

Good luck!  And please, if you find something you like and get up and
running, follow up.

mrc

Re: unzip files bigger than 4 GB

[Mike Castle]

Nvm, confused 2G with 4G.

Sorry for the noise.

On Wed, Jun 14, 2023 at 12:21 PM Mike Castle  wrote:
>
> It seems like it should.  I haven't upgraded my system yet:
>
> $ unzip -v | grep -e 6 -e LARGE
> UnZip 6.00 of 20 April 2009, by Debian. Original by Info-ZIP.
> USE_DEFLATE64 (PKZIP 4.x Deflate64(tm) supported)
> LARGE_FILE_SUPPORT (large files over 2 GiB supported)
> ZIP64_SUPPORT (archives using Zip64 for large files supported)
> USE_BZIP2 (PKZIP 4.6+, using bzip2 lib version 1.0.8, 13-Jul-2019)
>
> On Wed, Jun 14, 2023 at 11:19 AM Van Snyder  wrote:
> >
> > unzip v 6.0 (the version delivered with Debian 10) doesn't work with files 
> > bigger than 2^32 bytes.
> >
> > Is there an alternative program to do it?
> >

Re: unzip files bigger than 4 GB

[Mike Castle]

It seems like it should.  I haven't upgraded my system yet:

$ unzip -v | grep -e 6 -e LARGE
UnZip 6.00 of 20 April 2009, by Debian. Original by Info-ZIP.
USE_DEFLATE64 (PKZIP 4.x Deflate64(tm) supported)
LARGE_FILE_SUPPORT (large files over 2 GiB supported)
ZIP64_SUPPORT (archives using Zip64 for large files supported)
USE_BZIP2 (PKZIP 4.6+, using bzip2 lib version 1.0.8, 13-Jul-2019)

On Wed, Jun 14, 2023 at 11:19 AM Van Snyder  wrote:
>
> unzip v 6.0 (the version delivered with Debian 10) doesn't work with files 
> bigger than 2^32 bytes.
>
> Is there an alternative program to do it?
>

Re: Running Debian without initramfs?

[Mike Castle]

On Thu, Jun 8, 2023 at 10:50 AM Greg Wooledge  wrote:
> Merged-usr is officially mandated for bookworm, and upgrades to bookworm
> will do the merge, if it hasn't already happened.

End of an era.  My first Linux system (predating the existence of
Debian), mounted /usr over NFS over PLIP.

I couldn't afford a large enough harddrive for the second system, nor
ethernet cards (and a local shop was going to charge me $50 to make a
crossover cable if I went that route!).



Anyway, I think it is also pretty common to install merged-usr on LVM
as well, which I think is another reason to need initramfs.

mrc

Re: Multi Desktop Windows?

[Mike Castle]

Depends on your desktop/window manager, most likely.

For me, with XFCE, it is ctrl-alt- by default.  And they appear
to be configurable in the Settings -> Window Manager -> Keyboard section.

mrc

Re: Most maintainable way to install perl modules on Debian sysetms

[Mike Castle]

I would not be surprised if the version number indicated the module in not
Pure Perl, but rather includes some C source code.  Which would then need
to be compiled specifically for the version of Perl installed.

mrc

Re: /home as a symlink?

[Mike Castle]

You could run into issues where the value of 'pwd' does not equal the value
of 'readlink -f .'.

For myself, I use autofs with autohome.  It's been a while since I've
looked at the details, but I believe it simply does with bind mount
described elsewhere in this thread.  My main machines happen to be down at
the moment, so I can't provide a working example.

Outside of that, I would update /etc/passwd to point to the new location
instead.

mrc

Re: Beowulf gone?

[Mike Castle]

I took rendering video to be an immediate example, but not necessarily
the only thing of interest.

Re: Beowulf gone?

[Mike Castle]

Down to it's basic, rendering videos is nothing more than a simple
map-reduce, partioning a workload in a bunch of identical bits of
processing.  That could be done with N machines and a few simple shell
scripts.  Not really any need for anything fancy.  What the fancier
software gives you is stuff like automatic retries, fault tolerance,
fancy ways of tracking the process of a job, and so on.  But they
likely require more management to maintain.

Other reasons for a cluster might be web serving.  You want to be able
to handle some amount of queries per second with a reasonable latency
for most of those queries.  That may require you to scale in different
ways.  You need more machines to handle serving up the web content, a
bit more to handle the various types of processing that needs to be
done, some sort of scaling solution for your data store.  Still
parallelism, but each thing happening in parallel is likely a
different type of task.  A simple thing like Puppet or Chef could help
you keep your machines in order.

Maybe you don't need the computational power of a bunch of CPUs, but
you need all of the memory for something?  Maybe you need a bunch of
machines with 128G of ram to look like a single machine with 1P of
ram.  So you need to have something specialized for moving bits around
efficiently (either code to the data, or data to the code).

Maybe you want to have different types of pipelines that do completely
different things, and you want to make efficient use of the machinery
you have.  Say, you're doing both of the above, and during some hours
of the day, you need more processing power to go towards web serving,
you take away resources from you map-reduce.  During off hours, you
can let the MR have more processing power.  Now you're dynamically
trading off resources, and this is where some sort of clustering
management might become useful.

There is not likely to be any one solution for everyone.  So no longer
having anything exactly like Beowulf or some sort of direct
replacement is not suprising.  Likely stuff was learned from Beowulf
and friends.  Some things worked well, some things no, some things
were never used, some things were needed.  Folks take that experience
and build new tools and systems.  Old ones languish.  And no direct
replacement exists.

Likely you need to break down what you really want to do, then look to
see what solutions might work best for you, and experiment with the
various ones, and see which one you like best.

mrc

Re: MTP device mount gnome 3

[Mike Castle]

On Sat, Jan 21, 2017 at 11:47 AM, Tony Baldwin  wrote:
> My experience has been that this whole "MTP" thing, instead of just mounting
> phones like they used to, as a storage device, has been a real horror show,

It's less of a horror show than having two operating systems trying to
write to the same block device at the same time.  Or having to disable
access to the block device from one OS, which can really mess up
applications trying to access it.  What happens on your computer if
you are running an application where the binary itself or the data it
needs is on a USB device and you pull the device out?  Now imagine
that describing everything you're running.

> I've given up on getting it to work consistently and have installed dropbox
> on my phone and desktop to move stuff back and forth, which also has its
> headaches and limitations, but is consistent, and works..

That's essentially what MTP does (only without using an third location
to sync against).  Well, probably closer to using FTP to files to/from
the device, only without using TCP/IP.  (Oh, just refreshing and
apparently one can run MTP over TCP/IP, though I suspect that
implementations being talked about here are still direct USB
connections.)

Any fancy GUI for navigating file systems should function as well for
an MTP attached device as it would for say, an FTP service.

>From what I've heard, MS had 3rd tier engineers develop MTP, and even
they are a little ashamed of it.  But, it's still the primary protocol
out there.  So, the idea of MTP is sound, but the particular design,
and the various implementations, well... there is probably room for
improvement.  But it's unlikely to happen with anything less than full
industry support.

mrc

Re: Some Flash news

[Mike Castle]

On Mon, May 16, 2016 at 2:26 PM, Sven Arvidsson  wrote:
> That's Shumway from Mozilla.

Google's Swiffy fits into this domain as well.

mrc

Re: Iceweasel: permissions.default.image=3 (that is, prevent third-party images from loading) is ignored for some sites?

[Mike Castle]

I believe that cached images will still load.

mrc

Re: turning off Google Chrome's warning

[Mike Castle]

On Thu, Jan 21, 2016 at 6:59 PM, Frank McCormick
 wrote:
> So I just
> might remove google-chrome and live with chromium for now. An install of
> 64-bit Debian is not in the cards for now.


At some point, there may be 64-bit only code introduced into Chrome
that could cause subtle bugs on 32-bit systems (due to lack of
thorough testing).  Just something to keep an eye on if you're stuck
with the 32-bit world for a couple more years.

Hopefully fewer issues than the 8/16-bit to 32-bit migrations that
happened in decades past, due to better compiler supports for warnings
about non-portable code.  But could still happen.

mrc

Re: turning off Google Chrome's warning

[Mike Castle]

Besides switching to 64-bit or chromium or keeping the browser open?

(Actually, does chromium issue the same warning?)

mrc

Re: OT: reply styles, family matters

[Mike Castle]

On Mon, Nov 30, 2015 at 6:27 PM, Neal P. Murphy
 wrote:
> When you reply to and critique an essay, you would likely reply in top-post 
> form and leave the essay at the bottom so that readers, whom you may safely 
> assume have already read it, may conveniently reference it.


I don't think you can ever safely assume that anyone as read it.
That is why top-posting is always frowned upon.

mrc

Re: Mouse blanker?

[Mike Castle]

Well, -idle 0 will hide it right away.  But it'll get lots of false
positives about thinking you've stopped moving the mouse.

And unclutter has been around for just over 23 years now.

mrc

Re: Mouse blanker?

[Mike Castle]

Installed by default, meh.

But I'm pretty sure it is enabled by default.

cat /etc/X11/Xsession.d/90unclutter
# /etc/X11/Xsession.d/90unclutter
# This file is sourced by Xsession(5), not executed.

if [ -e /etc/default/unclutter ]
then
. /etc/default/unclutter
fi

if [ -x /usr/bin/unclutter ] && [ "${START_UNCLUTTER}" = "true" ]
then
/usr/bin/unclutter ${EXTRA_OPTS} &
fi

Re: ask.debian.net

[Mike Castle]

On Fri, Aug 28, 2015 at 5:34 AM, Lev Lazinskiy l...@levlaz.org wrote:
 1. It is very approachable to anyone since a lot of people have already
 used Stack Overflow.

 2. It has better search tools.

 3. Actual Answers float to the top (instead of having to read through en
 endless stream of threads or forum pages). This is great and provides an
 added value of creating a sort of mini knowledge base.

Personally, I find all three of these points to be false (at least
when it comes to SO style interfaces).

1. I hate having to have new userids and passwords just to use another
damned site.
2. I have more luck using a general search engine (in my case, Google)
than SO's search.
3. In SO I almost always have to read the entire thread anyway to get
additional context about the top voted answer, and often it's not he
best answer.

Maybe this has more to do with the fact that I tend to hunt for harder
questions.  Maybe that style works for nice simple questions, useful
for those just starting out.  But once you need something meatier,
just not as useful.

mrc

Re: Virtual noobie

[Mike Castle]

With VirtualBox, one has the option to install a bunch of guess
additions that help the guest and host work better together.

Is such needed/useful with KVM and friends?

FWIW, I use vbox as it comes with the installation, mostly because I'm
too lazy to download the upstream version.   Seems to work acceptably,
though I'm not stressing it any.

mrc

Re: OT: *FREE* VPN! Why?

[Mike Castle]

On Thu, Aug 6, 2015 at 9:36 AM, Jape Person jap...@comcast.net wrote:
 Brings up another point. I've always wondered how the sticky fingers crowd
 could manage all the key-presses necessary for arranging proper security.


One handed Dvorak keyboard mappings.

mrc


-- 
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Archive: 
https://lists.debian.org/ca+t9imz1ijxehegv4fpepmuyndvihkse5rqlcxnxbtgoec6...@mail.gmail.com

Re: What pulls in the tray of my /dev/sr1 ?

[Mike Castle]

On Sat, Jul 25, 2015 at 4:44 AM, Thomas Schmitt scdbac...@gmx.net wrote:

 LG Germany answered quickly and stated that the drive is
 not known to show this behavior under MS-Windows.
 (Linux is not on their compatibility list, they say.)


Has the drive displayed this behavior since you turned on the machine,
or just you just start to notice it after a while?

Maybe it only starts to happen after it's been on for a while, and
snarkWindow machines don't stay booted long enough/snark.

Ok, really maybe it only starts to happen after some specific event,
which could be any thing you'd done while working on libburn, or some
internal timer kicked off, or something like that.

It wasn't clear if your stability testing revolved around libburn or
not.  If it is, then maybe something tweaked it that wouldn't have
happened just a user box (vs a developer who might be more likely to
do something unusual).

mrc


-- 
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Archive: 
https://lists.debian.org/ca+t9imxkw0_ue_4zrd-ybdyrbsvvoul3utl-tcz2mpunmp9...@mail.gmail.com

Re: shutting off screen blanker forever?

[Mike Castle]

For xfce, you might try this:

Settings Manager  Session and Startup  Application Autostart
Scroll down and uncheck Screensaver.

There may be additional things you need to do to make sure session stuff
isn't loading screensavers through some other mechanism (i.e, squirreled
away in a saved session or something).  For that you might need to
explicitly exit the screensaver (either through Settings Manager 
Screensaver or a command line like xscreensaver-command -exit)

And you likely need the xset stuff you already did on top of that.

Of course, all of that is assuming you are using xscreensaver.

ps -ef | grep screensave

might be useful if the above doesn't work out for you.

mrc

PS: I dropped emc-users.  I think the netiquette these days is to NOT post
to multiple email groups, as it is likely a respondent isn't subscribed to
all of them.

Re: Another less woe

[Mike Castle]

I'm using i386.


-- 
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Archive: 
http://lists.debian.org/ca+t9imx+mhum0+pgmucg9j3qbsb013t2keeczwiwqhmdajf...@mail.gmail.com

Re: Another less woe

[Mike Castle]

Darn.

It happened to me on three different machines.  Though they're all the
same arch.

mrc


-- 
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Archive: 
http://lists.debian.org/ca+t9imxysjq7p6woef2sx736q38ubofwaymiodoq_t8rxds...@mail.gmail.com

Another less woe

[Mike Castle]

I just filed a bug on this, but I'm wanting a sanity check on this:

If I do something like:

less /usr/share/dict/words

then do this search:

(a|b)(c|d)

it crashes with a double-free error.


I'm not doing anything terribly funky there, am I?

mrc


-- 
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Archive: 
http://lists.debian.org/CA+t9iMx-o4nJBw-LRT+T3Wi99_VXTxrAdtB_RoXxU4nGW=o...@mail.gmail.com

Re: OT: Copying a URL from a text browser (was: Re: text browsers)

[Mike Castle]

On Sat, Mar 3, 2012 at 1:56 PM, Randy Kramer rhkra...@gmail.com wrote:

 Maybe it can be done readily in lynx and I just haven't spotted how to do it?

I have ~/.lynx/external to which I just added this line:
EXTERNAL:http:echo %s | xsel -i:TRUE


Then I can navigate to a link and hit the `.' key.  If there are more
than one commands registered for a particular protocol (http in this
case, which covers https), lynx will pop up a selection menu,
otherwise it'll just run it.

Though, in this case you could just as easily register something like:
EXTERNAL:http:firefox -new-tab %s:TRUE

Actually, looks like out of the box, lynx comes configured for running
x-www-browser, so whatever you have for your default browser should
kick in (but may not be what you want for new-tab, new-window,
profile, etc.

mrc


-- 
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Archive: 
http://lists.debian.org/CA+t9iMxoOSZEQVpMUz389KhiFYK9Kbdim=8r1tn3rh3kpvi...@mail.gmail.com

XFCE and automatic monitor detection

[Mike Castle]

This is mostly a laptop question, but probably general enough that I
want to post it here instead.

So one thing that I think Gnome2 had over XFCE  is better multiple
monitor support.  I could plug in a new monitor and the right thing
would just happen.  More importantly, I could unplug the monitor, and
everything would automatically collapse down onto the laptop.
Particularly useful if your screen is now locked.

XFCE seems to have a different philosophy on this matter:  use xrandr
(or similar).

Fine...OK... so rather than having to fire up arandr every time or
a script (tough to do if your screen is locked and the password
display is on the disconnected monitor), I'd like to have it happen
automatically.

So what kind of options do I have?

Some searching seems to indicate that it's a matter of polling.
Something akin to running xrandr -q every so often, and when a state
changes, do something appropriate.  There seems to be a hint that some
sort of even could be fired and caught instead (via acpi, dbus, other
assorted buzzwords), but that maybe it doesn't really work.


Does there exist any software that I can just run at login time that I
can configure such that ``when you see these particular monitors
attached, switch into this mode.'' However it does that, I don't care
as long as it doesn't bog down the machine.  And ideally work if a
monitor is unplugged when the machine is locked.

If not, what would be my approach for writing something for myself?

Thanks,
mrc


-- 
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Archive: 
http://lists.debian.org/CA+t9iMzvqEKXLZy1_BXcg5-VCz_9=npV=wmylymogsfpns4...@mail.gmail.com

/etc/network/interfaces whacked

[Mike Castle]

In addition to the Gnome 3 stuff, I just experienced another issue
with upgrading my laptop on the testing track.

Something whacked my /etc/network/interfaces.


Fortunately I happen to have a backup of the / partition, so I'll be
able to walk through multiple installs to try to identify the culprit
(though slow going as takes a while to rewrite the partition).


The first time through /etc/network/interfaces was simply no where to
be seen.  The second time, it was present, but missing things like my
wpa2 key.


Fortunately one of my desktops that I've updated has not been bitten
by this (or I didn't have any local customizations that were
affected).

mrc


-- 
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Archive: 
http://lists.debian.org/ca+t9imxkcgmekysnjjax-ncx3zb13w+xgjb4ktxsemcx31o...@mail.gmail.com

Re: /etc/network/interfaces whacked

[Mike Castle]

On Sun, Nov 13, 2011 at 10:11 AM, Mike Castle dalgoda+deb...@gmail.com wrote:
 In addition to the Gnome 3 stuff, I just experienced another issue
 with upgrading my laptop on the testing track.

 Something whacked my /etc/network/interfaces.


 Fortunately I happen to have a backup of the / partition, so I'll be
 able to walk through multiple installs to try to identify the culprit
 (though slow going as takes a while to rewrite the partition).


 The first time through /etc/network/interfaces was simply no where to
 be seen.  The second time, it was present, but missing things like my
 wpa2 key.


 Fortunately one of my desktops that I've updated has not been bitten
 by this (or I didn't have any local customizations that were
 affected).

Just to follow up on my own post, seems like 3rd times a charm.

This time I updated package by package (took output of apt-get upgrade
-s and looped over the packages in the same order using apt-get
install).

I did run apt-get update in there, so maybe my first couple of
attempts hit a bad package that was quickly fixed.

mrc


--
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Archive: 
http://lists.debian.org/ca+t9imwa-bu85sdxptcmebmen0xjyw7rvsaijt5bpkbh2sf...@mail.gmail.com

Re: DO NOT BUY Western Digital Green Drives (also present in WD Elements external USB cases)

[Mike Castle]

On Sat, Sep 3, 2011 at 1:14 PM, Luke Kenneth Casson Leighton
luke.leigh...@gmail.com wrote:
 just a word of warning: on absolutely no account, not for any reason,
 should you buy WD Green drives.

 i've just spent a hair-raising 6 weeks discovering that these drives,
 when pushed above a mere 40 Centigrade, become so unstable that they
 can actually become completely unresponsive, shut down, and leave the
 linux kernel in a completely unstable state, especially if they are
 part of a RAID1 mirror.

For what it's worth, WD _does_ say not to use Green drives in RAID and such.

http://wdc.custhelp.com/app/answers/detail/a_id/1397

mrc


-- 
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Archive: 
http://lists.debian.org/CA+t9iMz0MDct83sEEwXn4mZh2q=vbsypm52nhk5ju+awxpy...@mail.gmail.com

Re: is there no sane, minimal, graphical RSS feed reader in existance?

[Mike Castle]

On Fri, Jul 15, 2011 at 1:20 AM, Johannes Schauer j.scha...@email.de wrote:

 What I'm using now is liferea which is okay but could be more minimal
 and mainly, is way too slow to enjoy using it (search for the fsync
 issue).

Sounds like you may need to tune your filesystem.  If fsync() is
causing a problem, it probably means that it's forcing all pending
data writes for that filesystem to disk, not just that file.
Unfortunately this seems to be the default configuration because so
many broken programs seem to depend on that.  Sad really.  A file
system gets tweaked to make bad programs work well, then end up
slowing down correct programs.  So people break the correct programs
to get around the performance problems and end up depending on the new
feature of the filesystem.

Personally, I tune all of my filesystems with journal_data_writeback
enabled, and if any system apps break, I'll file bugs against them.
But meanwhile, apps that do correctly use fsync() don't mess up
performance for the whole system.

As a compromise, you might try making your home directory (or whatever
directory liferea keeps its data on) is configured with that mount
option and see if it improves performance.

Normally I'd point someone to this article:
http://thunk.org/tytso/blog/2009/03/15/dont-fear-the-fsync

but the site seems to be down at the moment.

Fortunately, I ended up going into my RSS reader of choice, Google
Reader, and finding the article and reshared it via my Buzz feed.  I
think you'd now be able to read it at:
https://plus.google.com/117219624378904478730/posts/Mwn9a2woZYv



Our of curiosity, why choose a local app over a central service like
Reader or any of the others out there?

I used to be a big fan of such local apps, but since I could be on any
number of machines (2 home desktops, a couple of laptops, few machines
at work), I've found a web app a lot more convenient.

mrc


-- 
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Archive: 
http://lists.debian.org/CA+t9iMzJEttGK4+oq_tHBtOZkT=wmy5exw20d9mbtm27pd0...@mail.gmail.com

Re: New automount (NFS) permission issues

[Mike Castle]

Did some more testing.  All of the problems seems to be client side.

Dropping back to 2.6.32, both automounts and the flock $0 script work over NFS.

But did discover something interesting.

After a fresh boot, the follow both work with 2.6.32:
$ ls /share/images
$ ls /share/images/

With 2.6.39, the first always fails, and the second succeeds (that is
the second succeeds if it's the first thing done after a reboot).


mrc


-- 
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Archive: 
http://lists.debian.org/CA+t9iMzH+YhaSFiZ9NVCB7L7RbSB5=qsjqf3xcqolmu6hha...@mail.gmail.com

New automount (NFS) permission issues

[Mike Castle]

I am running debian/testing across a number of machines, all mostly up
to date (usually any given machine is no more than a week behind).

Some time ago, maybe a couple of months, I started noticing some
problems with my automounted NFS mounts, and wondering if anyone else
has noticed something similar to the problems I'm about to describe.

It's possible that I have something misconfigured that has worked for
years, but now something is more strict.  But I've been unsuccessful
in finding anything in searches.  And I'm not sure if this might be an
autofs, kernel, or some other issue.  Any suggestions on what kind of
debugging to turn on would be helpful as well.


The effect I'm seeing is:  Immediately after boot, a normal user can
access any NFS directory causing it to automount.  After some amount
of time though, after the mount is automatically unmounted, normal
user can no longer do that, and instead root has to be used to access
the directory (causing automount to remount).

My configuration has not changed for a couple of years and has worked
fine until recently.

Each machine has a number of directories in /export that are exported
via NFS.  Most of these are configured for mounting into /share,
though /home is as well.  A typical exportfs entry looks like this:

/export/images
192.168.1.0/255.255.255.0(rw,async,no_subtree_check,root_squash)


The automount configuration is served up via ldap, and my configs
there look like:
dn: ou=auto.share,ou=automount,dc=mrc,dc=home
ou: auto.share
objectClass: top
objectClass: automountMap

dn: cn=/share,ou=auto.master,ou=automount,dc=mrc,dc=home
objectClass: top
objectClass: automount
automountInformation: ldap:ou=auto.share,ou=automount,dc=mrc,dc=home
--timeout=600
cn: /share

dn: cn=images,ou=auto.share,ou=automount,dc=mrc,dc=home
objectClass: automount
automountInformation: thune:/export/images
cn: images




This seems to happen on both local (direct mounts) and remote (NFS mounts).


Neither [autofs reload] nor [autofs restart] seems to help the
situation once is gets into this state.



One thing I've been doing as a temporary work around has been opening
extra shells and dropping them into the directories I want to keep
mounted.  Sort of defeats the whole purpose of automounting, but such
is the life of hacks.  Also, I've recently been seeing problems with
util-linux's flock(1) in my home directory.  I use it as a
serialization tool:

$ cat bin/ser
#!/bin/bash
flock $0 $@

And that has suddenly started failing with:
flock: /home/nexus/bin/ser: Input/output error

Which was quite scary at first (crap, /home is dying!) but is turning
out to actually just be a constant NFS glitch.  Not sure if related to
the automount issue, but would not be surprising.


I've scoured all of the stuff in /var/log with no luck.


Does anyone see anything I have misconfigured above?  Or extra configs
I need to share?

Any  ``Yes, you need to read X as it explains a recent change?''

Any recommend flags to turn on for additional debugging?

Anyone else seeing similar issues?

Thanks,
mrc


-- 
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Archive: 
http://lists.debian.org/ca+t9imyshmqwu2we79c3arwizsepfboww4dx13+l_lxzghd...@mail.gmail.com

Re: disk problems: which ATA?

[Mike Castle]

On Sun, Jul 3, 2011 at 11:25 AM, Ross Boylan
rossboy...@stanfordalumni.org wrote:
 How can I tell which ata device is which hard drive?  It's come up
 several times for me, most recently with
 ata2.00: exception Emask 0x0 SAct 0x0 SErr 0x0 action 0x6 frozen

Depending on how long since boot, you can often explore the output of
dmesg to figure out which drive is which.

Sometimes what I do is something like this pseudocode:

for disk in /dev/sd?; do
  echo $disk
  smartctl -i $disk | grep -e Model -e Serial
done

And write down each working drive's model and serial number.

Reboot. and do it again.

In many cases, I've been lucky that the failing drive would work for a
little while (say, until a write).  So I could compare the two lists,
and figure out which model/serial is failing, and pull it.

Failing that, I have a list of known good disks, and can go through
all of the disks in the machine until I find the failing one.  And at
that point, since I have the case open, reseat all cables and cards,
just in case that's the problem.

mrc


--
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Archive: 
http://lists.debian.org/CA+t9iMwkN-Atum5kdYUdQmjA1Tmat+bz=6ka-zzaf4nttgs...@mail.gmail.com

Re: [OT] Google search default lang.

[Mike Castle]

On Sun, Jun 19, 2011 at 8:55 AM, Hugo Vanwoerkom hvw59...@care2.com wrote:
 Hi,

 Using Google search always returns my results in Spanish because Google has
 figured out that my ISP is in a Spanish speaking country. But I want the
 results in US Eglish and always have to do an extra mouse click on
 'Google.com in English'

 Apparently Google does not record that I always click on that and adjusts
 the default language.

 Anybody know how to set the default language for search?

There has been a recent discussion of this over on Hacker News:
http://news.ycombinator.com/item?id=2673898

Well, OK, looks like the discussion is mostly about personalization of
search results, but geo-location and guessing languages does fit into
that.

Searching for that page for the word English seems to indicate that it
might be a bug.

Also, it looks like the reason Google (and other websites) do this is
because there seems to be a large number of browsers out there with
Accept-Language set incorrectly, so anything that says English is just
tossed out and whatever left is used (and if there is nothing left,
geo-locate).  I thought at one point en-gb would work, but some recent
testing I tried failed (but it was only about 10seconds worth of
testing).

mrc


-- 
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Archive: 
http://lists.debian.org/BANLkTim4m2jf28mzMuqSGPus6eriHY=e...@mail.gmail.com

ssh suddenly prompting for passphrase

[Mike Castle]

Just sharing something that happened to me.

After a recent upgrade with debian/testing, I noticed that ssh would
pop up a window asking for my password, and this would be AFTER
running ssh-add.

Turns out that I needed to read this bit in
/usr/share/doc/gnome-keyring/README.Debian:


The GNOME keyring includes the functionality of the SSH and GPG agents,
and it can break some setups, especially if ssh-agent and/or gpg-agent
is started by hand.

You can disable a specific component by removing the gnome-keyring-gpg
and gnome-keyring-ssh elements from the startup applications. The
interface depends on your session manager; for GNOME you can use
gnome-session-properties. You can also simply edit
/etc/xdg/autostart/gnome-keyring-*.desktop.


Now for me, surprise GUI prompts are annoying for a couple of reasons.
 First, it's new.  I've never seen that before this recent upgrade.
Second, I usually run screen, and it's possible the that first time I
run ssh out of this machine would have been after I first ssh-ed into
it, did screen -x, then ssh back out.  So there'd be some random X app
prompting for my password on a machine several buildings away.

Anyway, from my GNOME desktop, it was simply
System/Preferences/Startup Applications and uncheck GNOME Keyring SSH
Agent

Cheers,
mrc


-- 
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Archive: 
http://lists.debian.org/BANLkTikpaBSntxKFmtBp=hspxitou35...@mail.gmail.com

Re: Conversion to Ext4 in LVM

[Mike Castle]

On Thu, Nov 11, 2010 at 10:47 PM, Didar Hossain didar.hoss...@gmail.com wrote:
 Personally, I would not recommend converting, but, rather creating a
 separate partition
 for ext4 to test it out.

For my use case, in order to get the benefits for using ext4 over
ext3, it worked better to create a new filesystem with ext4 and move
all of the data onto it and them removing the old filesystem.

mrc


-- 
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Archive: 
http://lists.debian.org/aanlktimvf72zq68pw_t=edcverdt0tcypxyesq5pf...@mail.gmail.com

Re: The NAME environment variable

[Mike Castle]

On Fri, Aug 27, 2010 at 5:42 PM, T o n g mlist4sunt...@yahoo.com wrote:
 This is the first time that I found the NAME variable missing from the
 environment. How common is this?

Not present on my Debian/testing system.
I have USER, USERNAME and LOGNAME, but no NAME.

mrc


-- 
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Archive: 
http://lists.debian.org/aanlktik32l=8xqjshbv-v6_uy1e2t9erjlxsohnck...@mail.gmail.com

Re: xorg.conf -- nvidia to ati

[Mike Castle]

On Thu, Jun 24, 2010 at 7:38 PM, Alan Ianson agian...@gmail.com wrote:
 Any ideas on what I need to change?

I just switched to not using any xorg.conf at all, which I think is
the ``new'' recommendation.  I put new in quotes because I think I saw
someone at work the other day say something like ``You've not needed
one for five years,'' but I may not have had the context right.

So far so good, with two machines with different NVidia cards using
nouveau and one using radeon.

mrc


-- 
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Archive: 
http://lists.debian.org/aanlktim4c6s16zltcebqxugp_-3sf8cwq_m9llznl...@mail.gmail.com

Re: Filesystem recommendations

[Mike Castle]

On Sat, Apr 24, 2010 at 10:53 AM, B. Alexander stor...@gmail.com wrote:
 Does anyone have suggestions and practical experience with the pros and cons
 of the various filesystems?

Google is switching (has switched by now?) all of it's servers over to
ext4.  A web search will turn up more details on the subject.  But
they are mostly lots of big files.

mrc


-- 
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Archive: 
http://lists.debian.org/j2v537f90651004250829j1b0cb458y681b1732c2c2d...@mail.gmail.com

Re: Why does installing gnome packages versioned 2.28+6 insist on installing gnash?

[Mike Castle]

On Wed, Mar 17, 2010 at 1:05 AM, Mark Allums m...@allums.com wrote:
 Gnash is a noble effort.  Gnash sucks.  I want choice, and my choice is
 Adobe Flash.  Installing Gnash screws up Flash.  Right now, I can refuse to
 update GNOME on Squeeze any further, but the time will come when that will
 not be a viable option.  Why does GNOME require Gnash?  And what can I do to
 put a stop to it?

I think this is why we have the whole testing process.

Not all that long ago, one of my almost-daily upgrades pulled in a
whole bunch of dependencies that I really didn't want (details don't
matter).  But, I thought, eh, oh well.  Such is life.

A few days later, all of those packages that I didn't want were marked
as no longer being depended upon, so I was able to remove them.  Yay!
Someone fixed an errant dependency.

I hope that, as the fall out from this thread, something similar will
work out here.

Something along the lines, of I guess, that the Adobe based flash
package will be modified to say 'Provides: flash-plugin', as will
Gnash.  Then appropriate packages will say they depend on flash-plugin
rather than gnash explicitly.

The whole conflicts between gnash/adobe flash issue that someone else
brought up could be worked out later.

Would it have been better if things went along the Provides route
first?  Probably, but maybe there were other issues that prevented
that from happening, and these suggestions are already planned, just
not yet implemented.

Such is the life of living on the edge sometimes.

mrc


--
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Archive: 
http://lists.debian.org/537f90651003171208v1b79b154t48e258cab80ea...@mail.gmail.com

Re: w32codecs

[Mike Castle]

On Mon, Mar 1, 2010 at 7:30 AM, Celejar cele...@gmail.com wrote:
  My problem was that I hadn't realized that d-m had a
 non-free section at all.

I get the feeling the non-free section is new.

mrc


--
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Archive: 
http://lists.debian.org/537f90651003011311x671942e8n4c9a1fcd59686...@mail.gmail.com

autofs stopped working on testing

[Mike Castle]

Has anyone else noticed that autofs has stopped working on testing?
I'm really just digging into the debugging process, so may not have
read all of the necessary docs quite yet.

I've had autofs working for /home and a /share hierarchy for quite
some time now, and haven't had too many problems with it.  But the
initial set up was long enough ago that I don't exactly remember the
details of setting it up.

I've had one machine that's been giving me grief and had to reboot
daily for the last several days, so I'm pretty sure that the cause is
the last day or two.  I'm not seeing any packages, however, that look
like they might have caused the problem.  Nothing like slapd, glibc,
autofs, nss or the like.

Oh ... I just remembered... / on the ldap server was full, and I ended
up nuking a lot of stuff on that partition.  I wonder if I got overly
zealous and deleted something important.  I hope not.

Anyway, the symptom is that no autmount maps are being seen:

/etc/init.d/autofs start
no automount maps defined.


Everything is saved in ldap, nothing in /etc

ldapsearch -x

still shows all of the pertinent information



Still, any suggestions on how to proceed on this would be appreciated.

mrc


-- 
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Re: autofs stopped working on testing

[Mike Castle]

Solved!

On Sun, Jan 10, 2010 at 9:38 AM, Mike Castle dalgoda+deb...@gmail.com wrote:
 Oh ... I just remembered... / on the ldap server was full, and I ended
 up nuking a lot of stuff on that partition.  I wonder if I got overly
 zealous and deleted something important.  I hope not.

Not sure if I deleted too much, or full partition caused problem, but
I ended up restoring the slapd database from a backup, and that now
everything appears to work again.

I didn't suspect ldap at first, since auth was still working, and I
didn't see anything obvious in any log files about DB corruption.
Odd.

Still, all seems to be working now.

mrc


--
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

messed up error messages from gcc

[Mike Castle]

I typically keep my environment pretty stripped down, and so it may
turn out that I'm missing some package that causes the following
problem.  But I've not yet been able to figure it out.  I'm hoping the
masses out here will immediately recognize the problem as ``Oh yes,
you need ... ''



Running testing.


gcc -v shows:
gcc version 4.3.4 (Debian 4.3.4-6)


$ cat t.c
void foo(void) {
  bar();
}
$ gcc -Wall -Werror t.c
cc1: warnings being treated as errors
t.c: In function â:
t.c:2: error: implicit declaration of function â



I expect something closer to:
cc1: warnings being treated as errors
t.c: In function 'foo':
t.c:2: warning: implicit declaration of function 'bar'




Any thoughts?

Thanks,
mrc


--
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Re: messed up error messages from gcc

[Mike Castle]

On Wed, Dec 2, 2009 at 10:41 AM, Kumar Appaiah
a.ku...@alumni.iitm.ac.in wrote:
 Could you please try running LC_ALL=C gcc -Wall -Werror t.c and let us
 know if that solves the issue?

Yup.  That did it.  Thanks for the quick analysis.

LANG= gcc ...

had the same effect.

That's what I get for letting it set the darned thing to LANG=en_US.UTF-8


So, what's the proper solution to this?  Do I need to install
something?  Or rebuild a locale database somewhere? (if the latter, I
would have thought that it would have been done automatically upon
appropriate installs along the way.)  Just go back to C?

mrc


-- 
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Re: messed up error messages from gcc

[Mike Castle]

On Wed, Dec 2, 2009 at 11:14 AM, Mike Castle dalgoda+deb...@gmail.com wrote:
 So, what's the proper solution to this?  Do I need to install
 something?  Or rebuild a locale database somewhere? (if the latter, I
 would have thought that it would have been done automatically upon
 appropriate installs along the way.)  Just go back to C?


Just capturing more information.

Apparently every gcc-4.* has this same issue.  gcc-3.4 does not.

I always thought that part of the joy of the way GNU did translations
was that, if it wasn't available, it would always fall back to the
strings written into the source code (typically English, though not
necessary).  Though, admittedly, the last time I actually read the
docs on any of this stuff was probably over a decade ago, so things
have probably changed.


I guess this boils down to:  is this a bug in gcc or a bug in my set up?

mrc


--
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Re: messed up error messages from gcc

[Mike Castle]

On Wed, Dec 2, 2009 at 12:44 PM, Sven Joachim svenj...@gmx.de wrote:
 Maybe your terminal is not in Unicode mode?

Good possibility, but, I thought that would only matter when non-ascii
characters came into play.



Oh... ok.. I just found the UTF 8 item on xterm and there actually is
a minor difference:
 $ LANG=en_US.utf8 gcc -Werror -Wall t.c 21 | od -c
000   c   c   1   :   w   a   r   n   i   n   g   s   b   e
020   i   n   g   t   r   e   a   t   e   d   a   s   e
040   r   r   o   r   s  \n   t   .   c   :   I   n   f   u
060   n   c   t   i   o   n   �200 230   f   o   o   �200 231
100   :  \n   t   .   c   :   2   :   w   a   r   n   i   n   g
120   :   i   m   p   l   i   c   i   t   d   e   c   l   a
140   r   a   t   i   o   n   o   f   f   u   n   c   t   i
160   o   n   �200 230   b   a   r   �200 231  \n

$ LANG=en_US gcc -Werror -Wall t.c 21 | od -c
000   c   c   1   :   w   a   r   n   i   n   g   s   b   e
020   i   n   g   t   r   e   a   t   e   d   a   s   e
040   r   r   o   r   s  \n   t   .   c   :   I   n   f   u
060   n   c   t   i   o   n   '   f   o   o   '   :  \n   t   .
100   c   :   2   :   w   a   r   n   i   n   g   :   i   m
120   p   l   i   c   i   t   d   e   c   l   a   r   a   t   i
140   o   n   o   f   f   u   n   c   t   i   o   n   '
160   b   a   r   '  \n


Note that the single quotes wrapping foo are different.  I guess gcc-3
used back-tick and single-tick, and gcc-4 uses the more modern quote
characters, that my environment couldn't handle.

So, now to read up on which resource I need to set up for xterm to
make this the default.


Many thanks all for pointing me in the right direction.

mrc


--
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Re: what is the proper grub2 update procedure when changing disks

[Mike Castle]

On Mon, Nov 30, 2009 at 2:10 AM, David Goodenough
david.goodeno...@btconnect.com wrote:
 In the old Grub1 days if I had a bootable disk die and I copied its contents
 across to a new disk and wanted to make it bootable I followed a procedure
 that ran grub, looked for /boot/grub/stage1, set root to that hd, and then
 setup that hd, and I was done.

 But now Grub2 seems to rely on UUIDs rather than device names and so
 this does not work, or rather needs an extra step.

 So can someone point me to either an existing Howto which documents
 the new procedure (I looked using Google but got nothing but the old procedure
 above) or can someone outline the new procedure.  This is of course a
 request for a Debian procedure, I am running a nearly up to date sid.

I don't have access to the shell history, but I think the incantation
that I used was something like:

mount /dev/sdb1 /mnt
# copy stuff into /mnt
grub-install --root-directory=/mnt /dev/sdb
umount /mnt


The UUID stuff is mostly regenerated when you run grub-install.  Any
local overrides that you might have in place would need updating
first, of course.  Same goes for any UUID references in /etc/fstab and
similar.

mrc


--
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Re: Grep on dictionary words

[Mike Castle]

On Sat, Nov 28, 2009 at 7:13 AM, Dotan Cohen dotanco...@gmail.com wrote:
 I have a long binary file (about 12 MB) that I need to extract the
 text from via strings. Naturally, there are a lot of junk lines such
 as these:
 pDuf
 #k0H}g)
 GoV5
 rLeY1
 TMlq,*

 Is there a way to grep the output of strings in order to only show
 lines that contain words found in the aspell dictionary? Thanks in
 advance.


I typically use something like:

strings -n 5

And that removes most of the noise.  Typically the strings I'm looking
for tend to be longer, so if I miss the occasional short text, I'm OK
with that.


Another option might be to do something like:

look '' | grep ..  dict.txt   # get rid of all single letter words
strings logfile | fgrep -f dict.txt


mrc


-- 
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Re: clive doesn't extract youtube file

[Mike Castle]

On Tue, Nov 24, 2009 at 3:35 PM, Rodolfo Medina
rodolfo.med...@gmail.com wrote:
 I installed clive in my Lenny partition with: `aptitude install clive', but it
 seems that it does not manage to download the video I installed it for:

I believe that youtube has retired some support for a variety of
formats, so maybe the software simply needs to be updated for what is
now supported.

A recent web search for [youtube downloader] turned up a number of
discussions about this, and that is what I'm basing my reply on.


-- 
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Re: floppy mounting

[Mike Castle]

On Fri, Oct 30, 2009 at 10:39 AM, Jochen Schulz m...@well-adjusted.de wrote:
 dr.hugo.z.hackenbush:

 Hi, I am having trouble mounting the floppy in lenny .Can mount as root
 but wont let me mount as user?  Tried   #adduser (name) floppy  in
 terminal   but still wont let me in? any clues please?

 You need to login again after adding your user to the floppy group. Did
 you do that?

I would also recommend setting up pam to handle this for you using the
pam_group stuff.Whoever logs in on the console (via tty or gdm)
should probably get access to all sorts of physical devices, like
floppy, audio devices, cd/dvd drives, and so on.

That way, if you ssh into your machine, you don't accidentally start
affecting local devices remotely (you'd have to explicitly go through
root to do so).

mrc


--
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Re: Samba ls date weirdness

[Mike Castle]

You might also consider find -printf and stat as other options.


-- 
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Re: Further to fast booting for a debian system - changing getty to rungetty

[Mike Castle]

On Thu, Oct 8, 2009 at 8:21 AM, Lisi lisi.re...@gmail.com wrote:
 *So - how do I change my getty to rungetty?*

rungetty takes a different set of command line options than getty.

From reading the man page, it looks like you only need one argument:
the tty.  This doesn't seem too surprising since it looks like
rungetty isn't designed to run on serial lines, so it doesn't need to
know about linespeed.


So you need to make your entries look like:

1:2345:respawn:/sbin/rungetty tty1



Also, when I ran just plain rungetty from the command line, it dropped
a usage line into /var/log/auth.log


I'm not sure if you checked all of your log files, but I tend to do:
cd /var/log
ls -latr

just to see what was last modified, as messages get logged to a
variety of places.

mrc


-- 
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Re: df/du shows big difference of used space

[Mike Castle]

On Wed, Sep 9, 2009 at 9:56 AM, niclaswniclaswm...@gmail.com wrote:
 I have a 1500G hard drive, encrypted.
 Different commands shows different usage:

 As root, from root directory:
 df -m
 Filesystem   1M-blocks  Used Available Use% Mounted
 on
 /dev/mapper/d   1390840   128452535665  98% /

 That is 1284G used of 1391G which would be like 92%!
 Not 98%
 Why does it say 98%

Traditionally, 5% of every file system is reserved for root to do with
as it pleases.  In some filesystem (particularly older ones), all
sorts of things could go wonky whenever the system got 100% full.

Depending on the filesystem you're using, you can view or change that
reservation, though you may not really want to.

 du -sh *
 adds up to about 933G

 So the question is why the big difference of used space between df and du

 du shows 933G used
 df shows 1284G used

 Can anyone explain the big difference, 350G, of used disk space between du
 and df commands?

Two things.

First, your du -sh * misses . files/directories.  Depending on what's
going on, that could be significant.  I always use [du -sh .] myself.
Well, strictly speaking, the . isn't necessary (at least with GNU df,
not sure about others any more), but my fingers have been using that
for years and it's ingrained.

Second, on unix-like systems, it's possible to have a file open and
deleted.  It's common to do that with temp files, for instance.   It
could be that some process is running that still has a large file
open, but there's no corresponding file name to find for du to find
and account for, but df knows is still taking up space.

Tools like lsof may help in cases like that.  It has options to show
deleted files.  Though, personally, I tend to do something like:
ls -l /proc/*/fd/* | grep deleted

Mostly, because I learned that before lsof was written and I've never
really bothered to learn the ins and outs of lsof.

mrc


-- 
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Re: how to start a delayed shutdown over ssh and disconnect immediately?

[Mike Castle]

On Wed, Aug 12, 2009 at 9:52 AM, David
Christensendpchr...@holgerdanske.com wrote:

 I don't see a time delay option for poweroff.  I need a time delay to
 solve chicken and egg problems with shared folders, name resolution,
 etc. -- e.g. I need to tell all the machines to shutdown while they're
 all still running.  If the wrong machine shutdowns down prematurely, the
 overall shutdown script fails and I end up with some machines up and
 some machines down.

It seems to me that THIS is the problem that you should be solving,
not how to time your shutdown to be just right.

mrc


--
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Re: What is the best setup to compute in the burning hot sun?

[Mike Castle]

On Fri, Jul 31, 2009 at 11:39 PM, Tyler MacDonaldty...@yi.org wrote:
 My yard extends farther than I can piss - but not farther than I can run an
 extension cord.  It has a nice fir tree near that back.  Seems like a nice
 place to hang out and program, if I can see the screen clearly and not melt
 my interface to the digial universe.  I'll bring a jug of ice cold water and
 always keep another chilling in the fridge to keep my personal epidermal
 cooling system operational.

Don't forget, with the extension cord, you can put a fridge out there
by you as well.


You might look at some anti glare screens.  May make it harder to pair
program if you have someone come over.

If you haven't already, be sure to install some software to monitor
the internal temps, just so you can get a feel for what range they run
out during the day and such.  And keep the vents on the machine clear;
don't set it on towels and depending on how dusty, blowing out with
air every so often.  (Then again, with two cats, our laptops have more
problems inside.)

mrc


--
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Re: can I use ext4 now?

[Mike Castle]

On Wed, Jul 29, 2009 at 1:15 PM, Rick Pasottor...@niof.net wrote:

 mount -t ext4 -o nodelalloc /dev/sdc1 /s3

Leave off the -t ext4 and it should mount, though as ext4dev.  Or use
-t ext4dev.


There are some known bugs with the kernel you're using, hence all of
the recommendations for newer kernels (that also switch from ext4dev
to ext4).

However, if you just want to play with it, you could do that.

Personally, I have about 4TB in ext4dev, but it's all stuff that, if I
lost it, I'd be annoyed, but not loose anything important.

mrc


-- 
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Re: can I use ext4 now?

[Mike Castle]

On Thu, Jul 30, 2009 at 10:19 AM, Tim Tebbittteb...@gmail.com wrote:
 you should mount ext4dev filesystems using -o nodelalloc and only use
 freshly created filesystems using mke2fs -t ext4dev

Fortunately, the OP was doing exactly this.

mrc


-- 
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Re: a tool that allows to continue copying between HDDs

[Mike Castle]

On Tue, Jul 14, 2009 at 5:25 AM, Tzafrir Cohentzaf...@cohens.org.il wrote:
 So why not just use cp -a ?

Probably because, when I first learned to do this stuff, the system I
used did have a -a option to cp, but did have rsync installed.  And
now it's more muscle memory than anything.  (I'm almost to the point
where I can tell kids to get off my lawn.)

Also, one can do:
find . -depth | grep -v filter | cpio

And filter out some troublesome files.

mrc


-- 
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Re: lvm2 - question about pvmove

[Mike Castle]

Don't try to do two things at once.  If something goes wrong, you
won't know which is the cause.

Just put in the new drive and partition it into swap + lvm

swapoff /dev/sda1
vi /etc/fstab   # remove swap
pvcreate
vgextend
pvmove -v /dev/sdb2
vgreduce /dev/sdb2

shutdown and remove the bad drive

At that point, everything should be good and you can continue on.
Moreover, you'll still have nearly 400G on the new drive that you can
play with.

Now you can lvcreate a new volume and move stuff over and set up mount
points at your leisure.  You can then reduce the existing partition
and eventually get to the form you're looking for.

mrc


-- 
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Re: lvm2 - question about pvmove

[Mike Castle]

On Mon, Jul 13, 2009 at 10:41 AM, Ron Johnsonron.l.john...@cox.net wrote:


 What if I want 4 small partitions instead of one monster 1TB partition?
  I've read that you need a target at least as large as the source.

 (I've got this aching feeling that 1TB partitions are just not a good idea,
 and that granularity is always a good idea...)

It makes no sense to have multiple LVM partitions on the same disk,
just to put them back together again as one big volume group.  I mean,
what's the purpose of using LVM in the first place then?

No.  Partition the disk into one swap + one LVM.

Then use the LVM stuff to parcel out the space to various file systems.

The only case that I could see for partitioning the drive into
multiple LVM chunks would be if you want to experiment with using the
LVM commands like pvmove and you only have one disk.

For example, I have one machine with four disks, two 500GB and two
1TB.  The partition schemes on those are:
/dev/sda1   1  17  136521   82  Linux swap / Solaris
/dev/sda2  18  121601   976623480   8e  Linux LVM
/dev/sdb1   1  17  136521   82  Linux swap / Solaris
/dev/sdb2  18  121601   976623480   8e  Linux LVM
/dev/sdc1   1  17  136521   82  Linux swap / Solaris
/dev/sdc2  18   60801   488247480   8e  Linux LVM
/dev/sdd1   1  17  136521   82  Linux swap / Solaris
/dev/sdd2  18   60801   488247480   8e  Linux LVM

That gives me a VG of 2.73TB, which I have partitioned out into into a
number of LVs of all sorts of different sizes:
thune:~# lvdisplay | grep LV.Size
  LV Size300.00 GB
  LV Size128.00 GB
  LV Size70.00 GB
  LV Size5.00 GB
  LV Size512.00 MB
  LV Size16.00 GB
  LV Size1.00 GB
  LV Size14.00 GB
  LV Size1.94 TB
  LV Size20.00 GB
  LV Size25.00 GB

Hmmm.. . I wonder what that 1GB one is

One my other machine, it's similar, only with less swap across spindles:
/dev/sda1   1  182401  1465136001   8e  Linux LVM
/dev/sdb1   1  17  136521   82  Linux swap / Solaris
/dev/sdb2  18  121601   976623480   8e  Linux LVM

For another 2.27TB with:
  LV Size8.00 GB
  LV Size15.00 GB
  LV Size800.00 GB
  LV Size200.00 GB
  LV Size200.00 GB
  LV Size200.00 GB
  LV Size900.00 GB


Nothing wrong with big disk partitions (fdisk); just break it up into
smaller LVs

mrc


-- 
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Re: a tool that allows to continue copying between HDDs

[Mike Castle]

On Mon, Jul 13, 2009 at 7:45 AM, Sthu Deussthu.d...@gmail.com wrote:
 Is there a tool with which I can continue copying from HDD to another after
 some interrupt?

Depending on how stable you need the destination file system, but I often do:

find . -depth | cpio -pdmv /path/to/dest

Followed up by an rsync.   The cpio will leave partial files in place,
so the next time through it will skip that, but the rsync will fix it
up.

mrc


-- 
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Re: a tool that allows to continue copying between HDDs

[Mike Castle]

On Mon, Jul 13, 2009 at 10:58 AM, Neal Hogannealho...@gmail.com wrote:

 I'm curious b/c I am mildly interested in the OP's question and I
 briefly attempted to decipher the above response. There is no man page
 for 'mc' and google tends to lean towards midnight connection.

$ apt-cache search mc | grep -w mc
xnc - X Northern Captain nc/mc-like filemanager for X
mc - midnight commander - a powerful file manager
mc-dbg - midnight commander - a powerful file manager - debug package


-- 
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Re: lvm2 - question about pvmove

[Mike Castle]

On Mon, Jul 13, 2009 at 12:24 PM, Boyd Stephen Smith
Jr.b...@iguanasuicide.net wrote:
 pvcreate /dev/sdc1
 pvcreate /dev/sdc2
 pvcreate /dev/sdc3
 pvcreate /dev/sdc4
 vgextend $vg /dev/sdc1
 vgextend $vg /dev/sdc2
 vgextend $vg /dev/sdc3
 vgextend $vg /dev/sdc4
 pvmove /dev/sda2
 pvmove /dev/sdb
 vgreduce $vg /dev/sda2
 vgreduce $vg /dev/sdb
 pvremove /dev/sda2
 pvremove /dev/sdb


Wouldn't you want to move the first pvremove up after the first
pvmove?  Otherwise the second pvmove might choose to move onto the
device you just cleared out.

mrc


-- 
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Re: a tool that allows to continue copying between HDDs

[Mike Castle]

On Mon, Jul 13, 2009 at 12:20 PM, Celejarcele...@gmail.com wrote:
 I'm no expert in this stuff, so I'm curious - what is gained by this
 over a straight rsync?


In my experience,  find | cpio is faster than rsync for moving raw
data around.  Not sure why, but it feels that way.  It's been a long
time since I've done any speed tests.

If I remember correctly, rsync will still use one process for reading
and another for writing, so you end up reading gigs from disk, shoving
gigs over pipe, writing gigs to disk.  I'm not sure if that's still
the case or not.  The common tar cf - | tar xf - solution has the same
issue.  find | cpio just shoves the list of file names across the
pipe, so there's nearly a third less data being moved around.  Of
course, all of that may be immaterial on modern machines.

It may also be that find | cpio has less of the fragmentation issues
that rsync does (see discussions in the week or two on this list).
But that's pure guess work on my part.

mrc


-- 
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Re: Iceweasel's rendering faster than Firefox?

[Mike Castle]

On Mon, Jul 13, 2009 at 12:16 PM, Celejarcele...@gmail.com wrote:
 How, then, can there be a significant
 performance gap?

Maybe it's merely build options?  static vs dynamic libraries?  Maybe
FF has extra debugging turned on, or some feature that you'll find out
down the line that might be missed, but IW has turned off because it
slows things down?

mrc


-- 
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Re: Is a Universal Desktop Experience possible?

[Mike Castle]

On Sat, Jul 11, 2009 at 5:13 AM, John Haslerjhas...@debian.org wrote:

 And many apps keep files open while running, leading to lockouts or
 races.

 Elucidate.

Firefox.

You can only have a profile open on one machine at a time.

Very annoying.

There is no need to a profile to be tied to exactly one running
instance of the application.  Much of the data is stored in sqlite,
which I believe supports this mode of operation.  It should be
possible to do this.

Instead of launching FF directly, I launch a wrapper with picks a
different profile based on the machine name (and whether it's under
VNC or not).  I keep one profile as the master, and have a script that
syncs the rest to the master profile.  I keep any info I want shared
in online systems (i..e, bookmarks, history).

mrc


-- 
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Re: ext4 stable enough to entrust it with data?

[Mike Castle]

On Tue, Jun 30, 2009 at 11:40 AM, leel...@yun.yagibdah.de wrote:
 It seems you can convert ext3 to ext4 later, so I'm thinking
 of using ext3 for now and maybe converting later.

If you go this route, be sure you use a later kernel, .28+.  .26 has
known issues with mixing extent/non-extent files on the same system (I
think?, verify to be sure).  And I can attest to growing ext4
filesystems online is deadly.  :-/

On the other hand, I've decided to just grow offline (it's only me) as
I've been too lazy to figure out how to pull in newer kernels, so even
the bad stuff is pretty good.

mrc


-- 
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Re: OT: launching jobs in a combined serial parallel way

[Mike Castle]

On Mon, Jun 22, 2009 at 5:17 PM, Kamaraju S
Kusumanchiraju.mailingli...@gmail.com wrote:
 proga, progb are completely independent. They take couple of hours to
 finish. The time to complete proga, progb are not same.

 progc should to be launched only after both proga, progb are finished. progc
 takes another couple of hours to finish.

I've taken to using flock for such things if I'm launching them from
other scripts.  I forget which package and I can't look right now (my
machine died this morning).

I think it's like this:
flock /path/to/lockfile programtorun --options --to --programtorun

I would do something like:
flock /path/to/proga /path/to/proga --options_for_a
flock /path/to/progb /path/to/progb --options_for_b

flock /path/to/proga flock /path/to/progb /path/to/progc --options_for_c

That is, use an flock on the program itself as the serialization point.

Well, this works when the file I'm locking is a shell script in my
~/bin/ directory.  Not so sure if it's a /usr/bin/ type of file
without write access.  In that case you could always create files in
/tmp/

And there may be better ways to lock on multiple files than chaining
flock like I did above.  Read the man page.

mrc


-- 
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Re: OT: launching jobs in a combined serial parallel way

[Mike Castle]

On Wed, Jun 24, 2009 at 4:46 PM, Mike Castledalgoda+deb...@gmail.com wrote:
 I've taken to using flock for such things if I'm launching them from
 other scripts.  I forget which package and I can't look right now (my
 machine died this morning).

To clarify, I meant to say:

I've taken to using flock for such things if I'm launching them from
different windows.

mrc


-- 
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Re: convert standard time to unix time bash

[Mike Castle]

I think cross posting to so many lists, particularly across domains is
considered rude.

Meanwhile...

On Thu, Jun 18, 2009 at 11:30 AM, Tony Asnicarasnica...@gmail.com wrote:
 but how can I convert standard time to unix time? :D

date +%s

You can mix it with -d for fun things, if you need the time for specific dates:

date +%s -d '4 days ago'
date +%s -d '1970-01-01UTC'


-- 
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Re: OT: how to send a key code to a running process?

[Mike Castle]

On Wed, Jun 10, 2009 at 9:45 AM, Sthu Deussthu.d...@gmail.com wrote:
 Good day.

 How I can send a space key code to a running process by running a simple
 program? - I need this for mplayer to pause/continue. For now I have to refer
 to the console it is running in, but I want to use hot-keys - that will run a
 simple command that will send the signal to mplayer.

If you're not wedded to mplayer, you might consider other players as well.

I used to use xmms all of the time (and since, I've not found one
that's worked as well).

These days I use totem and I have FVWM bind keystrokes to:
Key P   A   CM  Exec totem --play-pause 
Key Right   A   CM  Exec totem --next 
Key LeftA   CM  Exec totem --previous 
Key Right   A   SM  Exec totem --seek-fwd 
Key LeftA   SM  Exec totem --seek-bwd 

Many other GUI players offer something similar.

With the links that David posted, you have to remember to launch
mplayer with the right flags.  Maybe that can be configured to do that
all of the time?  (I've not actually tried... will have to now
actually).

mrc


-- 
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Re: Program for quoting text like in email?

[Mike Castle]

On Mon, Jun 8, 2009 at 8:02 AM, Kumar Appaiaha.ku...@alumni.iitm.ac.in wrote:
 How about passing the text through fmt -w 80|sed 's/^/ //'?

Or use the -p option to fmt


-- 
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Re: mass conversion from ogg to mp3

[Mike Castle]

On Mon, May 25, 2009 at 1:54 PM, marc gm...@auxbuss.com wrote:
 ogg is a container, so unless you used a lossless codec (i.e. FLAC) then
 the mp3s are going to sound horrible, especially as mp3 also has sound
 shaping and, usually, produces variable bit rate files.

I thought most ogg's were typically vorbis.  And vorbis has all of
those same qualities that you ascribe to mp3: shaping and VBR.

 If you really have to do this, then I'd use the best codec you can find
 and stick to CBR files.

Unless one is dealing with a broken player that can't handle VBR,
you'd want to choose VBR over CBR.  The codec is able to assign more
bits to portions that need it, and fewer bits where it isn't as
necessary.  CBR might be useful for broadcast or real time streaming.
But for storage/playback, you'll want VBR.

mrc


-- 
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Re: Why is the kernel in testing so far behind what's current?

[Mike Castle]

On Fri, May 15, 2009 at 2:30 PM, Paul Johnson ba...@ursamundi.org wrote:

 Anything that doesn't have a showstopping-type bug filed against it in
 sid moves to testing after a week, as I understand it.  You might check
 bugs.d.o for information.  Is there something in particular you need
 from 27, 28, or 29?

ext4 that doesn't cause the system to lock up after resizing?

mrc


-- 
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Re: Why is the kernel in testing so far behind what's current?

[Mike Castle]

On Wed, May 13, 2009 at 4:41 AM, Nate Bargmann n...@n0nb.us wrote:

 In my experience, there is no point in bringing in a more recent kernel
 package until 2.6.30 is released which includes drm and video driver
 fixes required by the latest Xorg packages although the latest 2.6.2902
 package enabled 3D rendering again in the latest Xorg packages in Sid,
 at least for the Radeon chipset.

Personally, I'd like to see more stable ext4.  The currently packaged
kernel definitely has issues with ext4dev.

mrc


-- 
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Re: why must emacs depend on sound packages?

[Mike Castle]

On Sat, May 2, 2009 at 3:31 PM, Paul Scott waterho...@ultrasw.com wrote:
 GNU is an OS,  Linux is a kernel.

 Unfortunately popular usage has led to Linux incorrectly meaning GNU/Linux
 and even more.

How much GNU software is required before it has to have the GNU moniker?

If my machine uses the Linux kernel is mostly busybox instead of
coreutils/textutils/shutils do I have to keep using GNU/Linux?

If I use a BSD kernel with mostly GNU software, do I have to call it
GNU/BSD?  (Something I'd find very amusing, by the way.)

mrc


-- 
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org

Re: why must emacs depend on sound packages?

[Mike Castle]

On Sat, May 2, 2009 at 5:58 PM, Mike Castle dalgoda+deb...@gmail.com wrote:

 If I use a BSD kernel with mostly GNU software, do I have to call it
 GNU/BSD?  (Something I'd find very amusing, by the way.)

Oddly enough, in a completely different context, I did just come
across a reference to GNU/kFreeBSD.  So I guess folks DO use that
nomenclature.

Still think it's a bit odd, mind you.

mrc


-- 
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org