Re: Debian hardware compatibility

[Ben Caradoc-Davies]

On 20/04/17 12:33, Ben Caradoc-Davies wrote:

Asus UEFI BIOS (H110I-PLUS BIOS 3202) defaults to incorrect turbo boost
multipliers but I was able to manually set the correct "Per CPU" values
in the BIOS. For a 7700K, these should be 45/44/44/44 if you are not
overclocking (base clock 100MHz):
https://en.wikipedia.org/wiki/Kaby_Lake


For the record, upgrading to H110I-PLUS BIOS 3402 fixed the turbo boost 
multipliers.


Kind regards,

--
Ben Caradoc-Davies 
Director
Transient Software Limited 
New Zealand

Re: Console fonts

[David Wright]

On Tue 02 May 2017 at 12:20:01 (-0700), Larry Dighera wrote:

> Yes.  It's a new single-board computer platform that began shipping ~April
> 14, 2017.  I can personally confirm that Tails Linux X11 runs fine on this
> platform, and the manufacturer (Udoo) claims to have successfully installed
> Debian.  
[...]
> Given the Udoo team claims to have installed Debian on their hardware, and
> Tails Linux runs on it, I'd prefer to sort out the issues, and see if I
> (we?) can effect a useable system.

I'm sorry if everyone knows which Debian (jessie, stretch, sid)
and kernel version that the Udoo team installed. My deduction
from the lines above was that the OP ran Tails¹, not that the
Udoo team ran Tails.

Hence my thinking that there might be a reference to what the
Udoo team installed that I (and perhaps others) hadn't seen.
Sorry to mystify anyone.

¹
http://distrowatch.com/table.php?distribution=tails

Cheers,
David.

Re: Console fonts

[David Wright]

On Tue 02 May 2017 at 12:20:01 (-0700), Larry Dighera wrote:
> On Sun, 30 Apr 2017 21:34:14 -0400, Felix Miata  wrote:
> >
> >Even if OP can get the ttys working to his liking, I still think it's very
> >likely a lost cause trying to use Jessie on his hardware. Stretch is very 
> >near
> >ready to release, and probably OP's better next move.
> >
> 
> In this message thread:
> http://www.udoo.org/forum/threads/debian-jessie-linux-os-installation.6819/
> others have also suggested Stretch.  I looked at the existing bugs:
> https://bugs.debian.org/release-critical/other/testing.html
> https://bugs.debian.org/release-critical/
> 
> Release-critical bugs status
> 
> Tue May 2 17:00:00 UTC 2017
> 
> Total number of release-critical bugs: 1649
> Number that have a patch: 271
> Number that have a fix prepared and waiting to upload: 38
> Number that are being ignored: 78
> Number concerning the current stable release: 699
> Number concerning the next release: 149
> 
> 
> The reason for installing Debian was because I have been impressed with its
> stability and few update issues compared to other Linux flavors I've used,
> so I was/am reluctant to overwrite Jessie with Stretch.

Scaling up the words of that ridiculous advert:
699 (stable/jessie) is greater than 149 (testing/stretch).

But, seriously, those figures need a lot of interpreting.

> Given the Udoo team claims to have installed Debian on their hardware, and
> Tails Linux runs on it, I'd prefer to sort out the issues, and see if I
> (we?) can effect a useable system.

Reference? It's worth posting exactly what they installed;
distribution, kernel version, etc.

Cheers,
David.

Re: IDE to USB Converters

[David Christensen]

On 05/02/2017 06:02 AM, Martin McCormick wrote:

If one buys an IDE to USB converter for an older chassis
that currently can only boot from the IDE interface, will a
bootable usb thumb drive work with such a setup?

I actually have been using an IDE-SATA converter on these
systems but 2 of 4 SATA drives have completely died after 7 or 8
years of service and thumb drives are cheaper and can be
written to and read from on anything from a raspberry Pi to a
P.C.

The SATA drives that died simply don't even act like they are
plugged in. Oh, they get slightly warm but they don't show up as
a device.



On 05/02/2017 08:20 AM, Martin McCormick wrote:
> I know that the picture I have in my
> head is a board with an IDE connector on one side and a usb port
> on the other where one installs the thumb drive. It looks like an
> IDE drive to the mother board.
>
>The systems which will use the adaptors can't boot off of
> their usb2.0 ports and even if they did, they would be so slow
> that one wouldn't probably be too happy with the results.
>
>When using the IDE-SATA adaptors, these systems actually
> respond to shell commands and key presses much faster than they
> did using an electro mechanical IDE drive because they are always
> running at full speed so a seek to anywhere is governed by the
> bus speed and not how fast the heads physically move.
>
>Anyway, the reason I didn't just order one is to make
> sure that we aren't talking about the other direction. I have one
> of those and it works fairly well when one wants to format a SATA
> drive or work with an old IDE drive.

On 05/02/2017 09:41 AM, Martin McCormick wrote:
> Thanks for responding. I failed to mention that those SATA disks
> were SSD disks so it probably makes more sense to just buy some
> replacement drives as the IDE-SATA adaptors are still good.

I use SanDisk Ultra Fit USB 3.0 16 GB flash drives (~$10 each) as Debian 
7 system drives in P4 and newer systems.  They come close to saturating 
the USB 2.0 ports on my older machines (~25 MB/s read and write).  They 
work fine as a poor man's SSD for headless servers, but can get choppy 
for interactive use (notably, concurrent write loads).



IDE to SATA adapters and SATA SSD's are definitely going to be at or 
near the pinnacle of performance for your machines.



Other possibilities include:

1.  Get a USB 3.0 flash drive and install Debian onto it using a capable 
computer.  Then for each and every USB port on the target computer, plug 
in the USB flash drive with Debian and work through various combinations 
of BIOS settings to see if you can make it boot.  STFW for the same 
make/model motherboard to see if anyone has figured it out.


2.  Install a USB 3.0/PCI adapter that supports booting from USB flash 
drives and repeat #1 for its ports.


3.  Buy IDE flash modules or SSD's.

4.  Install a SATA/PCI host bus adapter (HBA) and buy SATA SSD's.


Understand that a given motherboard might have one or more PCI buses, 
and each bus might have one or more connectors.  For maximum 
performance, you want only one adapter card on each PCI bus.  (For PCIe, 
each connector usually has a dedicated bus.)



David

Re: Console fonts

[Felix Miata]

Larry Dighera composed on 2017-05-02 12:20 (UTC-0700):
...
>>Possibly Grub could be reconfigured to make a
>>lesser mode like 1440x900 or less explicit. 

> I'm willing to edit grub, if you're willing to provide specific
> instructions.  Unfortunately, my Unix experience predates grub (AT Unix
> SVR3 ~1994).

In general, this is enough howto:
https://askubuntu.com/questions/19486/how-do-i-add-a-kernel-boot-parameter

When I want 1440x900 as the mode on my Jessie ttys, I append as instructed 
there:

video=1440x900@60

The result is that

# fbset; inxi -G

reports mode 1440x900 using 180 columns and 56 rows. That won't produce a tty
result you can appreciate before your missing /dev/fb0 problem is solved, but is
the manner in which troubleshooting parameters can be supplied to the kernel in
attempting to identify solutions to video (and other) problems.

>>Lack of /dev/dri/card0 explains why
>>X doesn't work, the kernel found no supported gfx device.

> It sounds like you've discovered the root cause of the issue.  I failed to
> find /dev/dri let alone card0.  I have no idea what this means.
Technically, I don't either, but the gist as I understand it is that you have
video hardware that pure Jessie does not support.

DRI is Direct Rendering Infrastructure, just one of a multitude of software bits
that comprise working video on a Linux PC.

>>When I boot an Intel video Jessie PC with no video parameters on cmdline,
>>root=LABEL=deb8jessie plymouth.enable=0 noresume

> ???

That shows the cmdline from that boot included no video parameters other than
one to disable Plymouth functionality.

To find out what video parameters were on your cmdline on current boot, do:

# cat /proc/cmdline

>>I've seen nothing in thread explicitly explaining why he has no /dev/fb0, but
>>unless and until /dev/fb0 exists, ttys are stuck in 80x25. Maybe that is
>>something installing a working Plymouth can fix. I don't know, as I never use
>>Plymouth, and suspect it would also be victim of unsupported hardware.

> I did install Plymouth: apt-get install Plymouth.  It didn't seem to make
> any noticeable change.  (There are some Plymouth entries in daemon.log and
> syslog.)  Subsequently running lightdm still yields a black display screen.

Since it didn't help, and unless you have reason to think you might want or need
Plymouth in the future, I recommend keeping your Jessie simpler by reversing the
process:

# apt-get purge plymouth

> Given the Udoo team claims to have installed Debian on their hardware, and
> Tails Linux runs on it, I'd prefer to sort out the issues, and see if I
> (we?) can effect a useable system.
I seriously doubt Udoo's team installed pure Debian Jessie on the hardware you
have. Tails 2.12 uses the same Xorg version as Jessie (1.16.4), but with
approximately the same kernel as Debian Stretch (4.9 vs 4.9.13). This suggests
Jessie might work for you if you enable backports and install a linux-image much
newer than Jessie's 3.16. To enable backports, see:

https://wiki.debian.org/Backports

Once enabled you can check availability before choosing whether or which to 
install:

# apt-cache search linux-image
-- 
"The wise are known for their understanding, and pleasant
words are persuasive." Proverbs 16:21 (New Living Translation)

 Team OS/2 ** Reg. Linux User #211409 ** a11y rocks!

Felix Miata  ***  http://fm.no-ip.com/

Re: VPS

[Erick Ocrospoma]

2017-05-02 12:13 GMT-05:00 Adolfo Maltez :

> Saludos.
>
>
> Yo tengo un debian con Digital Ocean, sin ningún problema.
> Puedes usar este enlace para el programa de referidos:
>
> https://m.do.co/c/2a67235360d3
>
> Las tarifas son buenas, no es muy caro.
>
> Att.
> Adolfo Maltez
>
> El 2 de mayo de 2017, 10:25, Santi Moreno 
> escribió:
>
>> Hola gente,
>> me gustaría saber si alguien tiene experiencias con servidores virtuales
>> contratados y me podríais recomendar algún buen proveedor que monte
>> Debian y con buena relación calidad/precio.
>> Conozco 1and1 y arsys pero antes de decantarme por uno u otro me
>> gustaría conocer otras opiniones.
>>
>> Gracias de antemano,
>> @santimoreno
>>
>>
>
​Hola,

Por precios, estos VPS SSD son los baratos, en ese orden:

OVH, Vultr, Linode.​

​
​Va depender de tus necesidades (Regiones disponibles, Storage adicional
IPs publicas adicionales, Backup, Soporte, etc).
Personalmente tambien tengo un VPS en DigitalOcean sobre Debian Jessie
(minimal OS, con un poco de tuning es una bala),
 sin embargo estoy planeando pasarme a Vultr, por precio y algunos features
adicionales. Espero vaya igual de bien :)


-- 


Erick.

Re: converting my local site to be https only access

[Gene Heskett]

On Tuesday 02 May 2017 17:18:13 Jonathan Dowland wrote:

> On Tue, May 02, 2017 at 04:42:04PM -0400, Gene Heskett wrote:
> > Humm, is this sounding like I should open up a fwd to port 443 on
> > this machine in dd-wrt? As it is, I am only NATing port 6309 to it
> > via the NAT menu.
>
> You will need to open another port up for HTTPS, yes, and 443 is the
> default. If you open up 443, then https:// will work for
> browsers, otherwise you'd need an explicit port e.g. https:// URI>:1234, exactly the same as you have for your current site over
> HTTP, not using port 80.
>
Thanks for that, and I have the ideal number in mind, the 6309 is a clone 
of the 6809, until you flip a couple bits in an un-acknowledged control 
register.  Then it pipelines the instruction fetch, saveing a cycle, and 
grows a few more registers and instructions, some of which support 32 
bit data. Net result being that the os, rewritten to take advantage, 
Nitros9, which used to be os9, is nearly 2x faster, at the same old 
clock speed.

> So long as you intend for attempts to connect to your non-SSL HTTP
> site to be redirected to the HTTPS one, you will need two ports, as
> the HTTP site is still listening and serving requests, even if they
> are all redirects.
>
> (It might actually be technically possible to run HTTP and HTTPS on
> the same port using some kind of clever detection scheme to see which
> the client was using on connection, in the same way it is possible to
> multiplex HTTPS and SSH on the same port; but it's a sufficiently
> niché trick that I don't recommend trying it)

I've enough examples of Murphy around here already. :(

Thanks for clarifying that, Jonathan.

Cheers, Gene Heskett
-- 
"There are four boxes to be used in defense of liberty:
 soap, ballot, jury, and ammo. Please use in that order."
-Ed Howdershelt (Author)
Genes Web page 

Re: converting my local site to be https only access

[Gene Heskett]

On Tuesday 02 May 2017 17:13:44 Jonathan Dowland wrote:

> On Tue, May 02, 2017 at 01:24:42PM +0200, to...@tuxteam.de wrote:
> > I don't understand that: do you find Apache's config worse, or
> > lighttpd's or nginx's?
>
> Sorry I was unclear: I meant I find Apache's config the worst. It was
> the first HTTPD I used, and I spent many years supporting it
> professionally. I only switched to lighttpd for personal stuff because
> it was much easier to get FastCGI working. But then the scales fell
> from my eyes; and I wondered why I hadn't considered alternatives
> sooner. The logic in professional web hosting circles was that Apache
> HTTPD was the only serious HTTPD to use for "real" web pages (at least
> back then); but the configuration language was always a nightmare, I
> just didn't know better.
>
> > At work, Apache (they want it badly and it's not mine anyway). At
> > home, lighttpd (it's mine, after all).
>
> I still use lighttpd for my main web server but I've been
> investigating nginx for my home NAS web server (which does much more
> proxying to web apps inside containers and suchlike, rather than
> serving content itself). I found some limitations with lighttpd's
> reverse-proxying and rewriting things.

I've been looking at nginx myself, and wondering if it was any easier to 
setup. Apache2 can be a mutant bear with 6 sore paws.

Cheers, Gene Heskett
-- 
"There are four boxes to be used in defense of liberty:
 soap, ballot, jury, and ammo. Please use in that order."
-Ed Howdershelt (Author)
Genes Web page 

Re: converting my local site to be https only access

[Jonathan Dowland]

On Tue, May 02, 2017 at 04:42:04PM -0400, Gene Heskett wrote:
> Humm, is this sounding like I should open up a fwd to port 443 on this 
> machine in dd-wrt? As it is, I am only NATing port 6309 to it via the 
> NAT menu.

You will need to open another port up for HTTPS, yes, and 443 is the default.
If you open up 443, then https:// will work for browsers, otherwise
you'd need an explicit port e.g. https://:1234, exactly the same as
you have for your current site over HTTP, not using port 80.

So long as you intend for attempts to connect to your non-SSL HTTP site to
be redirected to the HTTPS one, you will need two ports, as the HTTP site is
still listening and serving requests, even if they are all redirects.

(It might actually be technically possible to run HTTP and HTTPS on the same
port using some kind of clever detection scheme to see which the client was
using on connection, in the same way it is possible to multiplex HTTPS and
SSH on the same port; but it's a sufficiently niché trick that I don't
recommend trying it)

-- 
Jonathan Dowland

Re: converting my local site to be https only access

[Jonathan Dowland]

On Tue, May 02, 2017 at 01:24:42PM +0200, to...@tuxteam.de wrote:
> I don't understand that: do you find Apache's config worse, or
> lighttpd's or nginx's?

Sorry I was unclear: I meant I find Apache's config the worst. It was the first
HTTPD I used, and I spent many years supporting it professionally. I only 
switched
to lighttpd for personal stuff because it was much easier to get FastCGI 
working.
But then the scales fell from my eyes; and I wondered why I hadn't considered
alternatives sooner. The logic in professional web hosting circles was that 
Apache
HTTPD was the only serious HTTPD to use for "real" web pages (at least back 
then);
but the configuration language was always a nightmare, I just didn't know 
better.

> At work, Apache (they want it badly and it's not mine anyway). At
> home, lighttpd (it's mine, after all).

I still use lighttpd for my main web server but I've been investigating
nginx for my home NAS web server (which does much more proxying to web
apps inside containers and suchlike, rather than serving content itself).
I found some limitations with lighttpd's reverse-proxying and rewriting
things.

-- 
⢀⣴⠾⠻⢶⣦⠀ 
⣾⠁⢠⠒⠀⣿⡁ Jonathan Dowland
⢿⡄⠘⠷⠚⠋⠀ https://jmtd.net
⠈⠳⣄ Please do not CC me, I am subscribed to the list.

Re: underscore in xterm sometimes invisible

[David Griffith]

On Mon, 1 May 2017, Sven Joachim wrote:


On 2017-05-01 08:58 +, David Griffith wrote:


When I completely log out and log back in, the underscore character
will be invisible in xterm.  If I do "xrdb .Xresources", subsequent
xterms created will show underscores.

Investigating further, I tried starting with no .Xresources file.
This gave me a much-too-small xterm window.  When I do "xrdb -merge
.Xresources-foobar", which is my normal .Xresources file, I get my
favored xterm settings EXCEPT that underscores are invisible.  If I
don't use the "-merge" flag, then that will cause subsequent xterms to
show underscores. What's going on here?  How can I fix it?


This has been reported in bug #858142[1], but I have not been able to
reproduce it on my systems and I don't know what's going on.

Cheers,
  Sven

1. https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=858142


Thanks.  I've been able to add some hopefully helpful information to that 
bug report, to wit, I think the problem may be with xrdb(1).



--
David Griffith
d...@661.org

A: Because it fouls the order in which people normally read text.
Q: Why is top-posting such a bad thing?
A: Top-posting.
Q: What is the most annoying thing in e-mail?

Re: converting my local site to be https only access

[Gene Heskett]

On Tuesday 02 May 2017 06:35:19 Jonathan Dowland wrote:

> On Fri, Apr 28, 2017 at 11:35:06PM -0400, Gene Heskett wrote:
> > Greetings all;
> >
> > My web site (see the sig) is local, on this machine, in a pretty
> > tight sandbox, but not running https.
> >
> > Where can I find a tut that is a complete instruction set to have it
> > do an auto-redirect to itself, but using the "s" stuff regardless of
> > the accessing client as long as the client can handle the https
> > stuff this conversion will return to the client?
>
> Inferring that you are using apache2, a few notes from my observations
> of the thread as it currently stands.
>
> one of the instruction sets you were following was suggesting to use
> mod_rewrite. Personally, I think that's overkill. Achieving what you
> want is possible using a simple Redirect, which is provided by
> mod_rewrite, which is very likely already enabled (but if not, you can
> enable it in the same way, via symlink, which you can create via
> a2enmod, or by hand).
>
> We don't know how you have your site set up already, in particular
> whether or not you are using VirtualHosts. My advice would be to do so
> if you are not, so a "pre-step" would be migrating to them.
>
> Once using VirtualHosts, you can configure one to bind to port 80, and
> another to 443. From memory, it would look something like this
>
>  
>RedirectMatch permanent ^(.*)$ https://$1
>  
>  
># configuration for your website (now) lives here
>  
>
> Substituting  appropriately.
>
Humm, is this sounding like I should open up a fwd to port 443 on this 
machine in dd-wrt? As it is, I am only NATing port 6309 to it via the 
NAT menu. I have dd-wrt locked down pretty tightly.  No one has come 
thru it in at least a decade except a friend, and I had to give him the 
username & pw, twice, once for the router and once to get on into this 
machine.

Right?

Thanks Jonathan.

> This is all from memory as I haven't used apache2 myself for many
> years (and looking back, having since used things like lighttpd and
> more recently nginx, the configuration language is much worse;
> stockholm syndrome whilst I was a user perhaps?)


Cheers, Gene Heskett
-- 
"There are four boxes to be used in defense of liberty:
 soap, ballot, jury, and ammo. Please use in that order."
-Ed Howdershelt (Author)
Genes Web page 

Re: Console fonts

[Larry Dighera]

Hello Felix,

Thank you for your informative response to my issue.

My comments in-line below:


On Sun, 30 Apr 2017 21:34:14 -0400, Felix Miata 
wrote:

>Larry Dighera composed on 2017-04-30 16:40 (UTC-0700):
>[...]
>Previously, in OP https://lists.debian.org/debian-user/2017/04/msg00534.html :
>http://www.udoo.org/
>***
>   Without anything there indicating date of release, that URI strongly 
> suggests
>to me nevertheless that his hardware is newer than Jessie can be expected to
>support.
>

Yes.  It's a new single-board computer platform that began shipping ~April
14, 2017.  I can personally confirm that Tails Linux X11 runs fine on this
platform, and the manufacturer (Udoo) claims to have successfully installed
Debian.  

Specs are here: http://www.udoo.org/new-resources-udoo-x86/
Intel® Celeron® N3160, Quad Core @1.6GHz (Turbo Boost 2.24GHz), 2MB
Cache, 6W TDP.
Integrated Intel® HD Graphics controller
Three independent display support
HW decoding of HEVC(H.265), H.264, MPEG2, MVC, VC-1, VP8, WMV9,
JPEG/MJPEG formats
HW encoding of H.264, MVC and JPEG/MPEG formats
Video Interfaces
HDMI connector
2 x miniDP++ connectors
Video Resolution
Up to 3840 x 2160 24bpp @ 30Hz, 2560 x 1600 24bpp @60Hz
CIR (Consumer InfraRed) Sensor
Arduino 101 compatible shield
Integrated 6-axis combo sensor with accelerometer and gyroscope


Here is data from Debian Jessie on the Udoo X86 platform:

--- System Information
Kernel name:Linux 
Network node Hostname:  UdooX86Debian 
Kernel release: 3.16.0-4-amd64 
Kernel version: #1 SMP Debian 3.16.39-1+deb8u2 (2017-03-07) x86_64 GNU/Linux
Machine hardware name:   
Operating system:   

--- OS Release ---
PRETTY_NAME="Debian GNU/Linux 8 (jessie)"
NAME="Debian GNU/Linux"
VERSION_ID="8"
VERSION="8 (jessie)"
ID=debian
HOME_URL="http://www.debian.org/;
SUPPORT_URL="http://www.debian.org/support;
BUG_REPORT_URL="https://bugs.debian.org/;

  CPU Information -
Architecture:  x86_64
CPU op-mode(s):32-bit, 64-bit
Byte Order:Little Endian
CPU(s):4
On-line CPU(s) list:   0-3
Thread(s) per core:1
Core(s) per socket:4
Socket(s): 1
NUMA node(s):  1
Vendor ID: GenuineIntel
CPU family:6
Model: 76
Model name:Intel(R) Celeron(R) CPU  N3160  @ 1.60GHz
Stepping:  4
CPU MHz:   499.800
CPU max MHz:   2332.3999
CPU min MHz:   499.8000
BogoMIPS:  3199.86
Virtualization:VT-x
L1d cache: 24K
L1i cache: 32K
L2 cache:  1024K
NUMA node0 CPU(s): 0-3
processor   : 0
vendor_id   : GenuineIntel
cpu family  : 6
model   : 76
model name  : Intel(R) Celeron(R) CPU  N3160  @ 1.60GHz
stepping: 4
microcode   : 0x40a
cpu MHz : 499.800
cache size  : 1024 KB
physical id : 0
siblings: 4
core id : 0
cpu cores   : 4
apicid  : 0
initial apicid  : 0
fpu : yes
fpu_exception   : yes
cpuid level : 11
wp  : yes
flags   : fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca
cmov pat pse36 clflush dts acpi mmx fxsr sse sse2 ss ht tm pbe syscall nx
rdtscp lm constant_tsc arch_perfmon pebs bts rep_good nopl xtopology
nonstop_tsc aperfmperf pni pclmulqdq dtes64 monitor ds_cpl vmx est tm2 ssse3
cx16 xtpr pdcm sse4_1 sse4_2 movbe popcnt tsc_deadline_timer aes rdrand
lahf_lm 3dnowprefetch ida arat epb dtherm tpr_shadow vnmi flexpriority ept
vpid tsc_adjust smep erms
bogomips: 3199.86
clflush size: 64
cache_alignment : 64
address sizes   : 36 bits physical, 48 bits virtual
power management:

processor   : 1
vendor_id   : GenuineIntel
cpu family  : 6
model   : 76
model name  : Intel(R) Celeron(R) CPU  N3160  @ 1.60GHz
stepping: 4
microcode   : 0x40a
cpu MHz : 499.800
cache size  : 1024 KB
physical id : 0
siblings: 4
core id : 1
cpu cores   : 4
apicid  : 2
initial apicid  : 2
fpu : yes
fpu_exception   : yes
cpuid level : 11
wp  : yes
flags   : fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca
cmov pat pse36 clflush dts acpi mmx fxsr sse sse2 ss ht tm pbe syscall nx
rdtscp lm constant_tsc arch_perfmon pebs bts rep_good nopl xtopology
nonstop_tsc aperfmperf pni pclmulqdq dtes64 monitor ds_cpl vmx est tm2 ssse3
cx16 xtpr pdcm sse4_1 sse4_2 movbe popcnt tsc_deadline_timer aes rdrand
lahf_lm 3dnowprefetch ida arat epb dtherm tpr_shadow vnmi flexpriority ept
vpid tsc_adjust smep erms
bogomips: 3199.86
clflush size: 64
cache_alignment : 64
address sizes   : 36 bits physical, 48 bits virtual
power management:

processor   : 2
vendor_id   : GenuineIntel
cpu family  : 6
model   : 76
model name  : Intel(R) 

Re: Alternative for printer-driver-cups-pdf?

[Jason]

On Tue, 02 May 2017 20:04:55 +0200
Felix Natter  wrote:

>hello Brian,
>
>Brian  writes:
>> On Tue 02 May 2017 at 18:12:18 +0200, Felix Natter wrote:
>>
>>> does anybody know an alternative for PDF printing (from any application
>>> that can print) to printer-driver-cups-pdf?
>>> 
>>> The output of printer-driver-cups-pdf (in jessie and sid)
>>> is really bad, see:
>>> http://www2.inf.fh-brs.de/~fnatte2s/freeplane.pdf
>>> (might not be viewable by in-browser pdf-viewer, save and open with
>>> evince/okular to see the broken content: jagged edges, useless font
>>> rendering)
>>> 
>>> The application I am using (Freeplane 1.6) cannot produce PDFs on its
>>> own (due to incompatibility with the new SVG rendering).
>>> 
>>> There was a bug #847462, but it's resolved and didn't seem to help.
>>> 
>>> Any hints?
>>
>> brian@desktop:~$ file freeplane.pdf 
>> freeplane.pdf: PostScript document text conforming DSC level 3.0
>>
>> Do you want to start again? :)
>
>I tried renaming the file to freeplane.ps, and get the same result when
>viewing with evince ;-)
>
>> What application produced this PostScript file?
>
>Freeplane, a java mind map editor. I strongly assume that this is the
>built-in java printing functionality. A screenshot of the printing
>dialog is attached.
>
>Thanks and Best Regards,


According to the screenshot you included, the 'Print to File' option is
checked. This is my guess as to why a postscript was produced. That option
should not be checked when printing to cups-pdf.

Jason

Re: Boot Drives for older Systems

[Pascal Hambourg]

Le 02/05/2017 à 19:13, Martin McCormick a écrit :

This drive will be booting Debian so my question is not totally
off-topic. Some older P.C. BIOS programs will not boot any drive
over 32GB. Can you get a larger drive, partition it with a 32 GB
boot partition and expect it to work that way with the remainder
of the drive available for more space such as /home?


If I remember correctly, some old BIOS completely hang when detecting a 
drive bigger than 32 GiB. So such a disk cannot be used for booting with 
such a BIOS, unless it has a 32-GiB clip jumper or its HPA (host 
protected area) is set to less than 32 GiB.

Re: Alternative for printer-driver-cups-pdf?

[Brian]

On Tue 02 May 2017 at 20:04:55 +0200, Felix Natter wrote:

> hello Brian,

Hello Felix.
 
> Brian  writes:
> > On Tue 02 May 2017 at 18:12:18 +0200, Felix Natter wrote:
> >
> >> does anybody know an alternative for PDF printing (from any application
> >> that can print) to printer-driver-cups-pdf?
> >> 
> >> The output of printer-driver-cups-pdf (in jessie and sid)
> >> is really bad, see:
> >> http://www2.inf.fh-brs.de/~fnatte2s/freeplane.pdf
> >> (might not be viewable by in-browser pdf-viewer, save and open with
> >> evince/okular to see the broken content: jagged edges, useless font 
> >> rendering)
> >> 
> >> The application I am using (Freeplane 1.6) cannot produce PDFs on its
> >> own (due to incompatibility with the new SVG rendering).
> >> 
> >> There was a bug #847462, but it's resolved and didn't seem to help.
> >> 
> >> Any hints?
> >
> > brian@desktop:~$ file freeplane.pdf 
> > freeplane.pdf: PostScript document text conforming DSC level 3.0
> >
> > Do you want to start again? :)
> 
> I tried renaming the file to freeplane.ps, and get the same result when
> viewing with evince ;-)

Which you would. Renaming the file does not alter its contents. The
point is that freeplane file has not been produced by cups-pdf. It is
not (as you assert) the output of printer-driver-cups-pdf.
 
> > What application produced this PostScript file?
> 
> Freeplane, a java mind map editor. I strongly assume that this is the
> built-in java printing functionality. A screenshot of the printing
> dialog is attached.

If freeplane produced the PostScript file it is highly unlikely any
post-processing of it would produce any better output.

The print dialog is not quite gtkprint (see the one for Firefox or
Evince). Whatever it is doing, it does not appear to be producing a PDF
with the PDF printer it displays. (Producing a PDF is the raison d'être
for cups-pdf).

-- 
Brian.

Re: Alternative for printer-driver-cups-pdf?

[Lisi Reisz]

On Tuesday 02 May 2017 19:04:55 Felix Natter wrote:
> I tried renaming the file to freeplane.ps, and get the same result when
> viewing with evince ;-)

Renaming is generally a waste of time in Linux, which looks at the file not 
the file extension.  You are thinking of Another OS. ;-)

Lisi

Re: Alternative for printer-driver-cups-pdf?

[Felix Natter]

hello Brian,

Brian  writes:
> On Tue 02 May 2017 at 18:12:18 +0200, Felix Natter wrote:
>
>> does anybody know an alternative for PDF printing (from any application
>> that can print) to printer-driver-cups-pdf?
>> 
>> The output of printer-driver-cups-pdf (in jessie and sid)
>> is really bad, see:
>> http://www2.inf.fh-brs.de/~fnatte2s/freeplane.pdf
>> (might not be viewable by in-browser pdf-viewer, save and open with
>> evince/okular to see the broken content: jagged edges, useless font 
>> rendering)
>> 
>> The application I am using (Freeplane 1.6) cannot produce PDFs on its
>> own (due to incompatibility with the new SVG rendering).
>> 
>> There was a bug #847462, but it's resolved and didn't seem to help.
>> 
>> Any hints?
>
> brian@desktop:~$ file freeplane.pdf 
> freeplane.pdf: PostScript document text conforming DSC level 3.0
>
> Do you want to start again? :)

I tried renaming the file to freeplane.ps, and get the same result when
viewing with evince ;-)

> What application produced this PostScript file?

Freeplane, a java mind map editor. I strongly assume that this is the
built-in java printing functionality. A screenshot of the printing
dialog is attached.

Thanks and Best Regards,
-- 
Felix Natter

Re: systemd: com inhibir un servei?

[Ernest Adrogué]

2017-05- 2, 18:32 (+0200); Alex Muntada escriu:
> A més a més, comprova si el servei està configurat amb suport per
> a sysv-rc, perquè aleshores potser també et calgui desactivar-lo:
> 
> $ ls -l /etc/rc?.d/*lirc*
> $ sudo update-rc.d -f {nom-del-servei-a-sysv-rc} remove

En aquest cas no ha calgut.

Merci.

Re: systemd: com inhibir un servei?

[Ernest Adrogué]

2017-05- 2, 18:29 (+0200); Marcos escriu:
> 
> En principi hauria de ser suficient això:
> 
> # systemctl stop lircd-uinput.service
> # systemctl disable lircd-uinput.service

Era això.

Gràcies.

Re: Boot Drives for older Systems

[Dan Ritter]

On Tue, May 02, 2017 at 12:13:37PM -0500, Martin McCormick wrote:
> This drive will be booting Debian so my question is not totally
> off-topic. Some older P.C. BIOS programs will not boot any drive
> over 32GB. Can you get a larger drive, partition it with a 32 GB
> boot partition and expect it to work that way with the remainder
> of the drive available for more space such as /home?

Usually yes. The exception is that some BIOSes have problems
even recognizing the existence of a disk larger than X TB, where
X=2 or 3.

>   Maybe a better form of this question is- Are there any
> unexpected gotchas likely in using SDA1 for the boot drive and
> SDA2-x  for the rest? It will probably be SDA2 for most of the
> rest with SDA3 being extended and SDA5 swap.

No, this is perfectly normal.

> My 300 GB drive still works but it is around 8 years old and is a
> spinning  electro mechanical type and it will fail some day about
> a millisecond after it was working perfectly.

That is the way these things go.

General reminders that you probably don't need, but someone will: 

RAID is for uptime, backups are for recovery.

If you don't have an automatic backup, you probably don't have 
a current backup.

If you don't store your backup in a format that you can read
without special hardware or software, you probably won't have
that hardware or software when you need it the most.

-dsr-

Boot Drives for older Systems

[Martin McCormick]

This drive will be booting Debian so my question is not totally
off-topic. Some older P.C. BIOS programs will not boot any drive
over 32GB. Can you get a larger drive, partition it with a 32 GB
boot partition and expect it to work that way with the remainder
of the drive available for more space such as /home?

This sounds like and probably is a stupid question since
I presently do this very thing but the boot drive is /dev/sda1
and the big drive is /dev/sdb1 which has about 300 GB on it and
runs off the secondary slot on the same controller with no
known issues.

Maybe a better form of this question is- Are there any
unexpected gotchas likely in using SDA1 for the boot drive and
SDA2-x  for the rest? It will probably be SDA2 for most of the
rest with SDA3 being extended and SDA5 swap.

My 300 GB drive still works but it is around 8 years old and is a
spinning  electro mechanical type and it will fail some day about
a millisecond after it was working perfectly.

Re: VPS

[Adolfo Maltez]

Saludos.


Yo tengo un debian con Digital Ocean, sin ningún problema.
Puedes usar este enlace para el programa de referidos:

https://m.do.co/c/2a67235360d3

Las tarifas son buenas, no es muy caro.

Att.
Adolfo Maltez

El 2 de mayo de 2017, 10:25, Santi Moreno  escribió:

> Hola gente,
> me gustaría saber si alguien tiene experiencias con servidores virtuales
> contratados y me podríais recomendar algún buen proveedor que monte
> Debian y con buena relación calidad/precio.
> Conozco 1and1 y arsys pero antes de decantarme por uno u otro me
> gustaría conocer otras opiniones.
>
> Gracias de antemano,
> @santimoreno
>
>

Re: VPS

[john vera]

Hola, hoy dia estoy trabajando para unas vps con un servidor de dedicado
con la gente de netrouting, hasta ahora el soporte (en ingles) es bueno,
soporte
vía chat y en buen tiempo de respuesta

tengo instalado en ser servidor proxmox y mediante el trabajo con maquinas
virtuales
con KVM y alli distintos sistemas operativos según mi necesidad.

los precios puedes consultarlos en la web de netrouting  y ver si van
acordes a tu
disponibilidad.

Saludos!

El 2 de mayo de 2017, 11:55, Santi Moreno  escribió:

> Hola gente,
> me gustaría saber si alguien tiene experiencias con servidores virtuales
> contratados y me podríais recomendar algún buen proveedor que monte
> Debian y con buena relación calidad/precio.
> Conozco 1and1 y arsys pero antes de decantarme por uno u otro me
> gustaría conocer otras opiniones.
>
> Gracias de antemano,
> @santimoreno
>
>


-- 
John M. A. Vera F.
Debian user & VaSLibre Valencia-Venezuela
Linux Counter # 467192
Huella digital de la clave: 6402 0164 2FE9 AB99 555E EAAE BAA0 AC94 1CA7
7A86
y la verdad OS hará Libres!!

Re: Alternative for printer-driver-cups-pdf?

[Brian]

On Tue 02 May 2017 at 18:12:18 +0200, Felix Natter wrote:

> does anybody know an alternative for PDF printing (from any application
> that can print) to printer-driver-cups-pdf?
> 
> The output of printer-driver-cups-pdf (in jessie and sid)
> is really bad, see:
> http://www2.inf.fh-brs.de/~fnatte2s/freeplane.pdf
> (might not be viewable by in-browser pdf-viewer, save and open with
> evince/okular to see the broken content: jagged edges, useless font rendering)
> 
> The application I am using (Freeplane 1.6) cannot produce PDFs on its
> own (due to incompatibility with the new SVG rendering).
> 
> There was a bug #847462, but it's resolved and didn't seem to help.
> 
> Any hints?

brian@desktop:~$ file freeplane.pdf 
freeplane.pdf: PostScript document text conforming DSC level 3.0

Do you want to start again? :)

What application produced this PostScript file?

-- 
Brian.

Re: Gestion de très gros FS

[Thierry Bugier Pineau]

Bonjour

Ce n'est pas tordre le système de snapshot que de les garder longtemps ?

Pour lvm les snapshots sont en copy on write. Apparemment btrfs ferait pareil 
vu la description du comportement.

Tout ce que j'ai lu dit que les snapshot sont des instantanés qui servent le 
temps d'un backup. Autrement dit : 
- on fige le FS sur le disque 
- on met à part les écritures (une sorte de tampon, et lvm ou le FS le gère 
selon le cas)
-jusqu'à ce que le FS figé soit pleinement exploité pour une tâche de 
sauvegarde.
- la tâche terminée, on détruit le snapshot en y incorporant les écritures 
précédemment mises à part (là encore lvm ou le FS gère cela)

On peut tout aussi bien historiser les sauvegardes et laisser le backup les 
gérer (backuppc par exemple mais guère pour un usage pro sur grosse quantités 
de données)


Le 2 mai 2017 17:46:40 GMT+02:00, Daniel Caillibaud  a 
écrit :
>Le 21/03/17 à 11:42, Pierre Malard  a écrit :
>PM> J’ai lu que BtrFS semblait se présenter comme le « successeur » de
>ext4 et proposait un
>PM> redimensionnement à chaud en complément du gestionnaire de volumes
>logiques de Linux. Il
>PM> permettrait également l’agrégat de préifériques et la gestion de «
>snapshots
>PM> » (https://fr.wikipedia.org/wiki/Btrfs). Avez-vous une expérience
>dans ce domaine et est-ce
>PM> que cela répondrait à notre besoin de gros volumes extensibles ?
>
>J'arrive longtemps après la question, au cas où ça serve à d'autres…
>
>Je n'ai pas d'expérience de btrfs sur de tels volumes, mais sur 3~4To
>de datas avec bcp de
>snapshots, il faut faire attention à l'ordre des snaphots pour garder
>une "filiation la plus
>linéaire possible".
>
>C'était pour du backup, je faisais
>- rsync de pleins de vm dans last (un subvolume)
>- delete Monday && snapshot de last sur Monday le lundi
>- … idem les autres jours, avec en plus le dimanche un
>- delete week_XX && snapshot de last sur week_XX
>
>mais de temps en temps, et de plus en plus souvent avec l'augmentation
>du nb de snapshots, le
>delete faisait complètement exploser le système, à retardement (lorsque
>btrfs nettoie ses
>metadatas, un peu plus tard, si le rsync démarre avant que tout soit
>nettoyé).
>
>L'explosion se traduisait par un système qui fige, avec ou sans oomkill
>tous azimuts. 
>
>Un expert btrfs m'a confirmé avoir déjà vu la RAM exploser dans ce
>genre de cas, sans vraiment
>savoir pourquoi… (que le load explose parce que le fs devient très lent
>ça peut s'expliquer,
>mais pas qu'il consomme énormément de RAM).
>
>C'est visiblement lié au fait du nb de snapshots qui dépendaient du
>volume dans lequel
>j'écrivais, chaque écriture sur un fichier déclenchant une cascade
>d'opérations pour que tous
>les subvolumes retrouvent leurs petits (tous ses snapshots doivent se
>mettre à jour sur
>l'ancienne version du fichier, trouver lequel la détient, etc.)
>
>En modifiant la rotation pour faire
>mv Monday avirer
>mv last Monday
>snapshot Monday last
>rsync vers last
>
>ça va bcp mieux (chaque écriture ne déclenche qu'un seul copy on write
>sur le dernier snapshot
>sans que les autres n'aient à faire qqchose, la suppression d'un
>subvolume n'entraînant de
>modif que chez son unique "fils").
>
>Tout ça pour dire que btrfs reste chatouilleux et peut partir en vrille
>(machine HS mais pas
>perdu d'octet), même si la gestion des snapshots reste un avantage très
>appréciable.
>
>Et je n'ai pas encore osé passer au btrfs send/receive pour
>synchroniser deux volumes, mais
>chez d'autres ça marche vraiment très bien (10~100 × plus rapide que
>rsync suivant le nb de
>fichiers, le volume et la BP dispo).
>
>-- 
>Daniel
>
>On devrait construire les villes a la campagne
>car l'air y est plus pur !
>Alphonse Allais

-- 
Envoyé de mon appareil Android avec Courriel K-9 Mail. Veuillez excuser ma 
brièveté.

Re: systemd: com inhibir un servei?

[Marcos]

Hola,

A 2017-05-02 17:24, Ernest Adrogué escrigué:


Com es fa per fer que no s'inici aquest servei?


En principi hauria de ser suficient això:

# systemctl stop lircd-uinput.service
# systemctl disable lircd-uinput.service

Salut,
--
Marcos

Re: systemd: com inhibir un servei?

[Alex Muntada]

Ernest Adrogué:

> /lib/systemd/system/lircd-uinput.service
> 
> ...
> 
> Com es fa per fer que no s'inici aquest servei?

$ sudo systemctl disable lircd-uinput.service

A més a més, comprova si el servei està configurat amb suport per
a sysv-rc, perquè aleshores potser també et calgui desactivar-lo:

$ ls -l /etc/rc?.d/*lirc*
$ sudo update-rc.d -f {nom-del-servei-a-sysv-rc} remove

Salut,
Alex

Syslog: rpc.imapd and nss_getpwnam “does not map into [ourdomain.tld]”

[Ron Leach]

List, good evening,

We have set up a new server intended only to serve archival and 
current files to other users on our network.  The primary applications 
are nfs and samba; sshd and exim (only for system mail to route to our 
on-site MTA on another machine) are also operating.


Syslog is showing continual errors saying:

May  2 06:25:07 Server6 rpc.idmapd[1799]: nss_getpwnam: name 
'ron@inet' does not map into domain 'ourdomain.tld'


This report occurs every 10 minutes.  Earlier in syslog, similar 
errors occur for:


May  2 06:25:04 Server6 rpc.idmapd[1799]: nss_getpwnam: name 
'sambauser1@localdomain' does not map into domain 'ourdomain.tld'


and

May  1 17:25:04 Server6 rpc.idmapd[1799]: nss_getpwnam: name 
'nobody@localdomain' does not map into domain 'ourdomain.tld'


Samba and nfs *appear* to be working, but I don't like having 
something set up that must be wrong in some aspect.  There are 2 
aspects of this report I do not understand.  (i) what are rpc.idmapd 
and nss_getpwnam normally trying to do?  (ii) why is the machine 
trying to use 'inet' and 'localdomain' when the machine - evidently, 
from the syslog report - already knows about our actual domain 
(anonymised to 'ourdomain.tld')?


If nss_getpwnam is to do with passwords, then the 3 applications I can 
think of, that are using usernames and passwords, are samba, exim, and 
sshd.


Grateful for any thoughts,

regards, Ron

Re: IDE to USB Converters

[Martin McCormick]

Dan Ritter  writes:
> Your SATA disks lasted 7-8 years? That's a reasonable lifetime
> for spinning disks. Replace them with more SATA disks, either
> spinning or SSD.

Thanks for responding. I failed to mention that those SATA disks
were SSD disks so it probably makes more sense to just buy some
replacement drives as the IDE-SATA adaptors are still good.

Martin

Re: Carte WiFi PCI Belkin F5D7000 : RÉSOLU

[andre_debian]

On Tuesday 02 May 2017 16:31:10 Frédéric MASSOT wrote:
> Le 02/05/2017 à 15:30, andre_deb...@numericable.fr a écrit :
> > Étonnant, pas de pilote pour Linux mais elle est bien détectée 
> > et fonctionnelle.

> Il y a nécessairement un pilote de chargé, tu peux le voir avec la
> commande : lspci -k :

Network controller: Ralink corp. RT2500 Wireless 802.11bg (rev 01)
Subsystem: Belkin F5D7000 v2000/v3000 Wireless G Desktop Card
Kernel driver in use: rt2500pci

André

VPS

[Santi Moreno]

Hola gente,
me gustaría saber si alguien tiene experiencias con servidores virtuales
contratados y me podríais recomendar algún buen proveedor que monte
Debian y con buena relación calidad/precio.
Conozco 1and1 y arsys pero antes de decantarme por uno u otro me
gustaría conocer otras opiniones.

Gracias de antemano,
@santimoreno

Alternative for printer-driver-cups-pdf?

[Felix Natter]

hello Debian users!

does anybody know an alternative for PDF printing (from any application
that can print) to printer-driver-cups-pdf?

The output of printer-driver-cups-pdf (in jessie and sid)
is really bad, see:
http://www2.inf.fh-brs.de/~fnatte2s/freeplane.pdf
(might not be viewable by in-browser pdf-viewer, save and open with
evince/okular to see the broken content: jagged edges, useless font rendering)

The application I am using (Freeplane 1.6) cannot produce PDFs on its
own (due to incompatibility with the new SVG rendering).

There was a bug #847462, but it's resolved and didn't seem to help.

Any hints?

Cheers and Best Regards,
-- 
Felix Natter

Re: Only root can write on USB disk

[Greg Wooledge]

On Tue, May 02, 2017 at 07:16:57PM +0430, Mostafa Shahverdy wrote:
> I have a very annoying problem. I can't write to my usb drives (fat32,
> ntfs, etc.) without root permissions. How can I fix this?

Mount the file system with "-o uid=youruser" to have the files presented
by the kernel as being "owned" by that user.

Re: comment faite-vous pour installer flashplayer?

[Daniel Caillibaud]

Le 04/04/17 à 19:35, Thierry Bugier Pineau  a écrit :
TBP> La question que je me pose sur ce sujet est : pourquoi avez vous besoin de 
Flash
TBP> aujourd'hui ? N'est il pas mieux de demander aux webmaster de se mettre à 
jour ?

C'est pas toujours possible…

L'association Sésamath a construit pendant 10 ans une base d'exercices 
interactifs de
mathématiques (libres, et construits collaborativement par plein de profs de 
math bénévoles),
essentiellement pour le collège. Elle est passé à js (+svg+d'autres technos) 
depuis plusieurs
années, mais il faudra encore un paquet avant d'avoir tout réécrit !
(+ de 4000 exercices, pour le refaire en js faut compter plusieurs heures par 
exo avec
développeur js/html/css/svg, plus long à former qu'un codeur flash)

Un aperçu sur http://mathenpoche.sesamath.net/

-- 
Daniel

Ils m'ont mal sous-estimé !
Georges W. Bush (6/11/2000)

Re: IDE to USB Converters

[Dan Ritter]

On Tue, May 02, 2017 at 08:02:32AM -0500, Martin McCormick wrote:
>   If one buys an IDE to USB converter for an older chassis
> that currently can only boot from the IDE interface, will a
> bootable usb thumb drive work with such a setup?
> 
>   I actually have been using an IDE-SATA converter on these
> systems but 2 of 4 SATA drives have completely died after 7 or 8
> years of service and thumb drives are cheaper and can be
> written to and read from on anything from a raspberry Pi to a
> P.C.
> 
> The SATA drives that died simply don't even act like they are
> plugged in. Oh, they get slightly warm but they don't show up as
> a device.

Your SATA disks lasted 7-8 years? That's a reasonable lifetime
for spinning disks. Replace them with more SATA disks, either
spinning or SSD.

I would expect a USB thumb drive to last considerably less than
a single year in a situation where they are written to at
significant rates.

-dsr-

Re: Gestion de très gros FS

[Daniel Caillibaud]

Le 21/03/17 à 11:42, Pierre Malard  a écrit :
PM> J’ai lu que BtrFS semblait se présenter comme le « successeur » de ext4 et 
proposait un
PM> redimensionnement à chaud en complément du gestionnaire de volumes logiques 
de Linux. Il
PM> permettrait également l’agrégat de préifériques et la gestion de « snapshots
PM> » (https://fr.wikipedia.org/wiki/Btrfs). Avez-vous une expérience dans ce 
domaine et est-ce
PM> que cela répondrait à notre besoin de gros volumes extensibles ?

J'arrive longtemps après la question, au cas où ça serve à d'autres…

Je n'ai pas d'expérience de btrfs sur de tels volumes, mais sur 3~4To de datas 
avec bcp de
snapshots, il faut faire attention à l'ordre des snaphots pour garder une 
"filiation la plus
linéaire possible".

C'était pour du backup, je faisais
- rsync de pleins de vm dans last (un subvolume)
- delete Monday && snapshot de last sur Monday le lundi
- … idem les autres jours, avec en plus le dimanche un
- delete week_XX && snapshot de last sur week_XX

mais de temps en temps, et de plus en plus souvent avec l'augmentation du nb de 
snapshots, le
delete faisait complètement exploser le système, à retardement (lorsque btrfs 
nettoie ses
metadatas, un peu plus tard, si le rsync démarre avant que tout soit nettoyé).

L'explosion se traduisait par un système qui fige, avec ou sans oomkill tous 
azimuts. 

Un expert btrfs m'a confirmé avoir déjà vu la RAM exploser dans ce genre de 
cas, sans vraiment
savoir pourquoi… (que le load explose parce que le fs devient très lent ça peut 
s'expliquer,
mais pas qu'il consomme énormément de RAM).

C'est visiblement lié au fait du nb de snapshots qui dépendaient du volume dans 
lequel
j'écrivais, chaque écriture sur un fichier déclenchant une cascade d'opérations 
pour que tous
les subvolumes retrouvent leurs petits (tous ses snapshots doivent se mettre à 
jour sur
l'ancienne version du fichier, trouver lequel la détient, etc.)

En modifiant la rotation pour faire
mv Monday avirer
mv last Monday
snapshot Monday last
rsync vers last

ça va bcp mieux (chaque écriture ne déclenche qu'un seul copy on write sur le 
dernier snapshot
sans que les autres n'aient à faire qqchose, la suppression d'un subvolume 
n'entraînant de
modif que chez son unique "fils").

Tout ça pour dire que btrfs reste chatouilleux et peut partir en vrille 
(machine HS mais pas
perdu d'octet), même si la gestion des snapshots reste un avantage très 
appréciable.

Et je n'ai pas encore osé passer au btrfs send/receive pour synchroniser deux 
volumes, mais
chez d'autres ça marche vraiment très bien (10~100 × plus rapide que rsync 
suivant le nb de
fichiers, le volume et la BP dispo).

-- 
Daniel

On devrait construire les villes a la campagne
car l'air y est plus pur !
Alphonse Allais

systemd: com inhibir un servei?

[Ernest Adrogué]

Hola,

El paquet lirc instal·la el fitxer

/lib/systemd/system/lircd-uinput.service

de manera que systemd inicia el servei "lircd-uinput" en arrencar.  El
problema és que aquest programa genera esdeveniments de tecla
fantasma.  Per exemple, en els logins tty m'apareix el caràcter "^@"
repetidament, com si algú estigués prement una tecla estranya.

Com es fa per fer que no s'inici aquest servei?

Salut.

Re: IDE to USB Converters

[Martin McCormick]

 writes:
> Does such an animal exist? I mean in the direction you are thinking
> of (I know the "other" direction exists).

That's what bothers me, too. I know that the picture I have in my
head is a board with an IDE connector on one side and a usb port
on the other where one installs the thumb drive. It looks like an
IDE drive to the mother board.

The systems which will use the adaptors can't boot off of
their usb2.0 ports and even if they did, they would be so slow
that one wouldn't probably be too happy with the results. 

When using the IDE-SATA adaptors, these systems actually
respond to shell commands and key presses much faster than they
did using an electro mechanical IDE drive because they are always
running at full speed so a seek to anywhere is governed by the
bus speed and not how fast the heads physically move.

Anyway, the reason I didn't just order one is to make
sure that we aren't talking about the other direction. I have one
of those and it works fairly well when one wants to format a SATA
drive or work with an old IDE drive.

Thank you .

Martin

Re: Carte WiFi PCI Belkin F5D7000 : RÉSOLU

[maderios]

On 05/02/2017 03:30 PM, andre_deb...@numericable.fr wrote:


Étonnant, pas de pilote pour Linux mais elle est bien détectée et
fonctionnelle.
La quasi totalité des pilotes sont disponibles avec le noyau Linux sous 
forme de modules qui sont chargés au démarrage seulement en cas de 
nécessité, ici une carte wifi. Ceci explique pourquoi posséder un 
matériel récent implique l'utilisation d'un noyau le plus récent possible.


--
Maderios

Only root can write on USB disk

[Mostafa Shahverdy]

Dear all

I have a very annoying problem. I can't write to my usb drives (fat32,
ntfs, etc.) without root permissions. How can I fix this?

-- 
Regards,
Mostafa Shahverdy 


signature.asc
Description: PGP signature

Re: Carte WiFi PCI Belkin F5D7000 : RÉSOLU

[Frédéric MASSOT]

Le 02/05/2017 à 15:30, andre_deb...@numericable.fr a écrit :
> On Tuesday 02 May 2017 14:24:40 err...@free.fr wrote:
>> On 05/02/2017 02:19 PM, andre_deb...@numericable.fr wrote:
>>> Parmi vous, qui a réussi à installer une carte WiFi PCI Belkin Wireless G 
>>> F5D7000 sous Linux (Jessie) ?
>>> Sur le site de Belkin. il n'existe pas de pilote pour Linux, que pour
>>> Windows. 
>>> C'est une  carte assez ancienne, si je peux l'installer tant mieux, 
>>> sinon tant pis.
> 
>> Pour beaucoup de prériphériques dont il n'y a pas de drivers libres, 
>> ou pas tout à fait libre, il faut installer le paquet firmware-nonfree :
> 
> Ce que j'ai fait,
> 
> la carte est bien détectée, la WiFi marche très bien (100%) avec "wicd".
> 
> Finies les câbles Ethernet qui se déroulent dans mon appartement :-)
> 
> Étonnant, pas de pilote pour Linux mais elle est bien détectée et 
> fonctionnelle.

Il y a nécessairement un pilote de chargé, tu peut le voir avec la
commande : lspci -k


-- 
==
|  FRÉDÉRIC MASSOT   |
| http://www.juliana-multimedia.com  |
|   mailto:frede...@juliana-multimedia.com   |
| +33.(0)2.97.54.77.94  +33.(0)6.67.19.95.69 |
===Debian=GNU/Linux===

Re: O/T question, but not completely

[tomas]

-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1

On Tue, May 02, 2017 at 09:53:49AM -0400, Greg Wooledge wrote:
> On Tue, May 02, 2017 at 09:38:28AM -0400, Steve Matzura wrote:
> > configure.ac:87: warning: macro 'AM_PATH_MAD' not found in library
> > configure.ac:89: warning: macro 'AM_PATH_LAME' not found in library
> 
> MAD and LAME are MP3 decoder/encoder libraries.  Perhaps these macros
> are supposed to come from those projects.  Ultimately you may have to
> ask the people who maintain the thing you're trying to compile.

Yeah, they rang some bell with me (lame pun, I know ;-)

As a possible lead, try this:

  https://sources.debian.net/advancedsearch/

under "Code Search" you can enter a regular expression, equivalent to
doing

  https://codesearch.debian.net/search?q=AM_PATH_MAD

The hits I see are for package avifile. Perhaps you might want to
install (some version of) libavifile-x.y-dev. Or provide some
suitable sources.

Regards
- -- tomás
-BEGIN PGP SIGNATURE-
Version: GnuPG v1.4.12 (GNU/Linux)

iEYEARECAAYFAlkIkvgACgkQBcgs9XrR2kbjFQCfe6WuBI/BOk2+pv2dWQt4SFRL
McgAnAxHBzPWgsQKheXZ+eIVs/zwwWnu
=TmhN
-END PGP SIGNATURE-

Re: O/T question, but not completely

[tomas]

-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1

On Tue, May 02, 2017 at 09:38:28AM -0400, Steve Matzura wrote:
> Please reply privately, unless you really think it's on-topic. Me, I'm
> not sure.
> 
> I trying to build  a package using automake (aclocal). I have all the
> tools installed that I need--autotools, autotools-dev, libtool, etc
> After patching the package's configure script, I attempted to
> regenerate the build files with:

I'd think that isn't that badly off-topic (that said, I do accept
corrections :)

> aclocal -I m4/autoconfautomake
> 
> and received the following two errors:
> 
> configure.ac:87: warning: macro 'AM_PATH_MAD' not found in library
> configure.ac:89: warning: macro 'AM_PATH_LAME' not found in library

Hm. This looks like some undefined variables. What are you trying
to build? Does it already exist in some form as a Debian package
(perhaps one could snarf hints from that)?

cheers
- -- t
-BEGIN PGP SIGNATURE-
Version: GnuPG v1.4.12 (GNU/Linux)

iEYEARECAAYFAlkIkCcACgkQBcgs9XrR2kYzSwCfbFcagC5q3JMIeGBeKEmqdv4+
kXEAn2tVXsfKLtfcxsFphhBT2Phy+rXg
=1yOc
-END PGP SIGNATURE-

Re: O/T question, but not completely

[Greg Wooledge]

On Tue, May 02, 2017 at 09:38:28AM -0400, Steve Matzura wrote:
> configure.ac:87: warning: macro 'AM_PATH_MAD' not found in library
> configure.ac:89: warning: macro 'AM_PATH_LAME' not found in library

MAD and LAME are MP3 decoder/encoder libraries.  Perhaps these macros
are supposed to come from those projects.  Ultimately you may have to
ask the people who maintain the thing you're trying to compile.

O/T question, but not completely

[Steve Matzura]

Please reply privately, unless you really think it's on-topic. Me, I'm
not sure.

I trying to build  a package using automake (aclocal). I have all the
tools installed that I need--autotools, autotools-dev, libtool, etc
After patching the package's configure script, I attempted to
regenerate the build files with:

aclocal -I m4/autoconfautomake

and received the following two errors:

configure.ac:87: warning: macro 'AM_PATH_MAD' not found in library
configure.ac:89: warning: macro 'AM_PATH_LAME' not found in library

I have spent over three hours scouring the net for information on
these two macros and come up with nothing except one reference in
aclocal.m4.

If you or someone you know knows how this all works, please reply
privately, with my thanks in advance to you, and to the list owner for
allowing this message.

Re: IDE to USB Converters

[tomas]

-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1

On Tue, May 02, 2017 at 08:02:32AM -0500, Martin McCormick wrote:
>   If one buys an IDE to USB converter for an older chassis
> that currently can only boot from the IDE interface, will a
> bootable usb thumb drive work with such a setup?

Does such an animal exist? I mean in the direction you are thinking
of (I know the "other" direction exists).

regards
- -- tomás
-BEGIN PGP SIGNATURE-
Version: GnuPG v1.4.12 (GNU/Linux)

iEYEARECAAYFAlkIjRcACgkQBcgs9XrR2kZv7ACcC99aERVWSaJbrwk0SCxDy+Q9
MsIAn1A5XVT2InCT0pCu03vUN3IxbGuX
=37Rm
-END PGP SIGNATURE-

Re: Problemas creando lanzador invocando gksudo desde el menú

[Julian Daich]

Hola,

Solucioné el problema poniendo la ruta de ejecución completa en lugar
de" ~". Parece que eso al menú no le gusta.

Saludos,

Julián

El día 2 de mayo de 2017, 8:37, Maikel Enrique Pernía Matos
 escribió:
> Julian:
>
> Particularmente cuando hago una aplicación que necesite ejecutarse con
> privilegios de «root» es en ella donde hago el chequeo y no en los
> lanzadores de la app o en un script en bash intermedio, ahora si lo que
> deseas es crear un lanzador con estas características quizás te sirva
> este ejemplo para lanzar una app como root:
>
> 
> [Desktop Entry]
> Name=MiApp
> Version=1.0
> Exec=su-to-root -X -c /usr/bin/miapp
> Comment=Comentario generico de la app
> Icon=/usr/share/icons/miapp.svg
> Type=Application
> Terminal=false
> StartupNotify=true
> Encoding=UTF-8
> Categories=System;Development;
> 
>
> Puedes copiar el anterior ejemplo modificar los valores correspondientes
> para que se ejecuta tu aplicación y guardarlo como: «nombre.desktop».
>
> También puedes utilizar en la variable «Exec» gksu como bien te han
> comentado otros colegas o si en el fichero «sudoers» has definido que no
> verifique la contraseña para tu usuario podrías utilizar sudo, esto
> ultimo (no chequear contraseña) no te lo recomiendo.
>
>
> Saludos,
> Maikel
>
>
> El sáb, 29-04-2017 a las 19:05 +0200, Julian Daich escribió:
>> Hola,
>>
>> Creé un lanzador en el menú de XFCE4 que invoca el siguiente comando
>> gksudo 'java -jar ~/bin/WIFIAuditor.jar'
>>
>> Después de pedir la contraseña no hace nada. El mismo resultado
>> creando un lanzador en el panel.
>>
>> Si uso la terminal o Alt+ F2 sí funciona¿ alguna idea?
>>
>> Saludos,
>>
>
>
>
> --
> Este mensaje le ha llegado mediante el servicio de correo electronico que 
> ofrece Infomed para respaldar el cumplimiento de las misiones del Sistema 
> Nacional de Salud. La persona que envia este correo asume el compromiso de 
> usar el servicio a tales fines y cumplir con las regulaciones establecidas
>
> Infomed: http://www.sld.cu/
>



-- 
Julian

Re: Carte WiFi PCI Belkin F5D7000 : RÉSOLU

[andre_debian]

On Tuesday 02 May 2017 14:24:40 err...@free.fr wrote:
> On 05/02/2017 02:19 PM, andre_deb...@numericable.fr wrote:
> > Parmi vous, qui a réussi à installer une carte WiFi PCI Belkin Wireless G 
> > F5D7000 sous Linux (Jessie) ?
> > Sur le site de Belkin. il n'existe pas de pilote pour Linux, que pour
> > Windows. 
> > C'est une  carte assez ancienne, si je peux l'installer tant mieux, 
> > sinon tant pis.

> Pour beaucoup de prériphériques dont il n'y a pas de drivers libres, 
> ou pas tout à fait libre, il faut installer le paquet firmware-nonfree :

Ce que j'ai fait,

la carte est bien détectée, la WiFi marche très bien (100%) avec "wicd".

Finies les câbles Ethernet qui se déroulent dans mon appartement :-)

Étonnant, pas de pilote pour Linux mais elle est bien détectée et 
fonctionnelle.

Grand merci,

André

Re: Problemas creando lanzador invocando gksudo desde el menú

[Maikel Enrique Pernía Matos]

Julian:

Particularmente cuando hago una aplicación que necesite ejecutarse con
privilegios de «root» es en ella donde hago el chequeo y no en los
lanzadores de la app o en un script en bash intermedio, ahora si lo que
deseas es crear un lanzador con estas características quizás te sirva
este ejemplo para lanzar una app como root:


[Desktop Entry]
Name=MiApp
Version=1.0
Exec=su-to-root -X -c /usr/bin/miapp
Comment=Comentario generico de la app
Icon=/usr/share/icons/miapp.svg
Type=Application
Terminal=false
StartupNotify=true
Encoding=UTF-8
Categories=System;Development;


Puedes copiar el anterior ejemplo modificar los valores correspondientes
para que se ejecuta tu aplicación y guardarlo como: «nombre.desktop».

También puedes utilizar en la variable «Exec» gksu como bien te han
comentado otros colegas o si en el fichero «sudoers» has definido que no
verifique la contraseña para tu usuario podrías utilizar sudo, esto
ultimo (no chequear contraseña) no te lo recomiendo.


Saludos,
Maikel


El sáb, 29-04-2017 a las 19:05 +0200, Julian Daich escribió:
> Hola,
> 
> Creé un lanzador en el menú de XFCE4 que invoca el siguiente comando
> gksudo 'java -jar ~/bin/WIFIAuditor.jar'
> 
> Después de pedir la contraseña no hace nada. El mismo resultado
> creando un lanzador en el panel.
> 
> Si uso la terminal o Alt+ F2 sí funciona¿ alguna idea?
> 
> Saludos,
> 



--
Este mensaje le ha llegado mediante el servicio de correo electronico que 
ofrece Infomed para respaldar el cumplimiento de las misiones del Sistema 
Nacional de Salud. La persona que envia este correo asume el compromiso de usar 
el servicio a tales fines y cumplir con las regulaciones establecidas

Infomed: http://www.sld.cu/

Fwd: Carte WiFi PCI Belkin F5D7000

[bernard . schoenacker]

> > Bonjour,
> > 
> > Parmi vous, qui a réussi à installer une carte WiFi PCI Belkin
> > Wireless G
> > F5D7000 sous Linux (Jessie) ?
> > 
> > Sur le site de Belkin. il n'existe pas de pilote pour Linux, que
> > pour
> > Windows.
> > 
> > C'est une  carte assez ancienne, si je peux l'installer tant mieux,
> > sinon tant pis.
> > 
> > Merci,
> > 
> > André
> > 
> > 
> 
> bonjour,
> 
> 
> pour "dédé" :
> 
> apt-cache policy $(apt-cache search ndiswrapper |awk '{print $1}')
> 
> rtfm : https://doc.ubuntu-fr.org/ndiswrapper
> 
> slt
> bernard
 
 

Re: Carte WiFi PCI Belkin F5D7000

[bernard . schoenacker]

- Mail original -
> De: "andre debian" 
> À: debian-user-french@lists.debian.org
> Envoyé: Mardi 2 Mai 2017 14:19:17
> Objet: Carte WiFi PCI Belkin F5D7000
> 
> Bonjour,
> 
> Parmi vous, qui a réussi à installer une carte WiFi PCI Belkin
> Wireless G
> F5D7000 sous Linux (Jessie) ?
> 
> Sur le site de Belkin. il n'existe pas de pilote pour Linux, que pour
> Windows.
> 
> C'est une  carte assez ancienne, si je peux l'installer tant mieux,
> sinon tant pis.
> 
> Merci,
> 
> André
> 
> 

bonjour,


pour "dédé" :

apt-cache policy $(apt-cache search ndiswrapper |awk '{print $1}')

rtfm : https://doc.ubuntu-fr.org/ndiswrapper




slt
bernard

IDE to USB Converters

[Martin McCormick]

If one buys an IDE to USB converter for an older chassis
that currently can only boot from the IDE interface, will a
bootable usb thumb drive work with such a setup?

I actually have been using an IDE-SATA converter on these
systems but 2 of 4 SATA drives have completely died after 7 or 8
years of service and thumb drives are cheaper and can be
written to and read from on anything from a raspberry Pi to a
P.C.

The SATA drives that died simply don't even act like they are
plugged in. Oh, they get slightly warm but they don't show up as
a device.

Re: Carte WiFi PCI Belkin F5D7000

[err404]

On 05/02/2017 02:19 PM, andre_deb...@numericable.fr wrote:
> Bonjour,
> 
> Parmi vous, qui a réussi à installer une carte WiFi PCI Belkin Wireless G 
> F5D7000 sous Linux (Jessie) ?
> 
> Sur le site de Belkin. il n'existe pas de pilote pour Linux, que pour Windows.
> 
> C'est une  carte assez ancienne, si je peux l'installer tant mieux, 
> sinon tant pis.
> 
> Merci,
> 
> André
> 

Pour beaucoup de prériphériques dont il n'y a pas de drivers libres, ou pas 
tout à fait libre, il faut installer le paquet firmware-nonfree

Carte WiFi PCI Belkin F5D7000

[andre_debian]

Bonjour,

Parmi vous, qui a réussi à installer une carte WiFi PCI Belkin Wireless G 
F5D7000 sous Linux (Jessie) ?

Sur le site de Belkin. il n'existe pas de pilote pour Linux, que pour Windows.

C'est une  carte assez ancienne, si je peux l'installer tant mieux, 
sinon tant pis.

Merci,

André

Re: converting my local site to be https only access

[tomas]

-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1

On Tue, May 02, 2017 at 11:35:19AM +0100, Jonathan Dowland wrote:

[...]

> one of the instruction sets you were following was suggesting to use
> mod_rewrite. Personally, I think that's overkill [...]

Good advice. Debugging Apache's mod_rewrite turns out to be a black
art in itself [1].

[elided, agree]

> This is all from memory as I haven't used apache2 myself for many years (and
> looking back, having since used things like lighttpd and more recently nginx,
> the configuration language is much worse; stockholm syndrome whilst I was a
> user perhaps?)

I don't understand that: do you find Apache's config worse, or
lighttpd's or nginx's?

Personally I *strongly* prefer lighttpd (I don't know nginx enough
to bother anyone with my opinion). Apache config's "looks-like-XML-
but-really-isn't" is downright ugly, but one can cope with that
(perhaps holding one's nose while editing. But the semantics ("looks-
like-declarative-but-really-isnt" -- see a pattern?) is horribly
error prone, and you've got to internalize that seven-phase model
and the hooks each module gets a stab at to understand the somewhat
counter-intuitive interaction of different configuration directives.

The result is that most end up cargo-culting some random snippets
off the Tubes, mixing them into their distro's default config and
beating on the resulting mess until it seems to work. Maintainability
and security... less good.

At work, Apache (they want it badly and it's not mine anyway). At
home, lighttpd (it's mine, after all).

Regards

[1] Written about an older mod_rewrite (Apache 1.3), but mod_rewrite has
   *grown* since then:

``The great thing about mod_rewrite is it gives you all the
  configurability and flexibility of Sendmail. The downside
  to mod_rewrite is that it gives you all the configurability
  and flexibility of Sendmail.''
-- Brian Behlendorf
Apache Group

`` Despite the tons of examples and docs, mod_rewrite is voodoo.
   Damned cool voodoo, but still voodoo. ''
-- Brian Moore
b...@news.cmc.net

   in http://mx.demos.su/manual/mod/mod_rewrite.html. I'd keep that
   reference around anyway, since it helps a lot to wrap one's head
   around Apache's "phases", without which config will stay a mystery
   forever.

- -- tomás
-BEGIN PGP SIGNATURE-
Version: GnuPG v1.4.12 (GNU/Linux)

iEYEARECAAYFAlkIbHoACgkQBcgs9XrR2kZsnQCfcsfSNnV+QLy3FB5NR7aTx+gp
188An05nuNZWwBDgR1ZgaQQ8jQ/II6JU
=YIX4
-END PGP SIGNATURE-

Re: converting my local site to be https only access

[Jonathan Dowland]

On Fri, Apr 28, 2017 at 11:35:06PM -0400, Gene Heskett wrote:
> Greetings all;
> 
> My web site (see the sig) is local, on this machine, in a pretty tight 
> sandbox, but not running https.
> 
> Where can I find a tut that is a complete instruction set to have it do 
> an auto-redirect to itself, but using the "s" stuff regardless of the 
> accessing client as long as the client can handle the https stuff this 
> conversion will return to the client?

Inferring that you are using apache2, a few notes from my observations of
the thread as it currently stands.

one of the instruction sets you were following was suggesting to use
mod_rewrite. Personally, I think that's overkill. Achieving what you
want is possible using a simple Redirect, which is provided by
mod_rewrite, which is very likely already enabled (but if not, you can
enable it in the same way, via symlink, which you can create via a2enmod,
or by hand).

We don't know how you have your site set up already, in particular whether
or not you are using VirtualHosts. My advice would be to do so if you are
not, so a "pre-step" would be migrating to them.

Once using VirtualHosts, you can configure one to bind to port 80, and
another to 443. From memory, it would look something like this

 
   RedirectMatch permanent ^(.*)$ https://$1
 
 
   # configuration for your website (now) lives here
 

Substituting  appropriately.

This is all from memory as I haven't used apache2 myself for many years (and
looking back, having since used things like lighttpd and more recently nginx,
the configuration language is much worse; stockholm syndrome whilst I was a
user perhaps?)

-- 
⢀⣴⠾⠻⢶⣦⠀ 
⣾⠁⢠⠒⠀⣿⡁ Jonathan Dowland
⢿⡄⠘⠷⠚⠋⠀ https://jmtd.net
⠈⠳⣄ Please do not CC me, I am subscribed to the list.