Re: Debian using USB stick on diskless machine

[David Christensen]

On 12/28/2015 04:08 PM, Ross Boylan wrote:

I have a diskless machine and figured it would be quicker to get it
running off a USB thumb drive than PXE boot.  I'd appreciate any
suggestions or advice.


That's one of my favorite tricks.  I typically boot and install from an 
optical disk to avoid confusion.  I prefer SanDisk Ultra Fit 16 GB USB 
3.0 flash drives because they are compact, fast, and inexpensive.  I use 
the d-i to partition them manually -- 0.5 GB ext4 boot, 0.5 GB random 
key LUKS swap, and 13.4 GB password LUKS ext4 or btrfs root.



On some machines:

1. You may need to go into CMOS setup after the Debian installer runs 
and configure the BIOS to boot from the USB flash drive (details vary by 
vendor and revision).


2.  Not all the USB ports are created equal.  If the one you installed 
to won't work for boot, try another.




I stuck Debian live on the thumb drive but discovered changes I made
in that environment were not persistent.


There are specific methods for persistence on Debian Live:


https://debian-live.alioth.debian.org/live-manual/html/live-manual.en.html#556


David

Re: HAL

[kamaraju kusumanchi]

>
> Because, criminal offence though I fully understand it to be, I want to use my
> television to watch UK Channel 4 catch-up TV and Amazon Prime Video.  For
> that I need the computer attached to my television to run Flash-player.
>
> I know that I can get the setup running in Wheezy.  The question is whether I
> can do so it Jessie.  But the stumbling block, I think is that I need HAL.
>

May I ask what browser you are using to watch Amazon Prime Video? When
I tried it in Iceweasel, I get the following error

Unsupported Browser
This web browser isn't compatible with Amazon Video. Please use one of
the following web browsers:
  Google Chrome (latest version)

I tried switching the user agent in Tools -> Browsers - Linux to
Chrome 43.0 (64 bit), Chrome 44.0 (32 bit), Firefox 40.0 (32 bit),
Firefox 38.0 (64 bit). None of them worked and I get the same error.


> As I say, I know that such behaviour is a criminal offence, so please, people,
> can we not go there.

I do not think it is a criminal offense to watch videos through Amazon
Video. But I am not interested in debating that either.

-- 
Kamaraju S Kusumanchi | http://raju.shoutwiki.com/wiki/Blog

Re: HAL

[kamaraju kusumanchi]

On Sun, Dec 27, 2015 at 9:30 PM, Sven Hartge  wrote:
> Sven Hartge  wrote:
>
>> But I see there is a package named libhal1-flash available from
>> Christian Marillat in his deb-multimedia.org repository. It contains the
>> code from https://github.com/cshorler/hal-flash.
>
> Hmm, libhal1-flash is only available for Stretch and Sid. I wonder why.
>
> Maybe you can ask Christian if he could rebuild the package for Jessie.
>

I submitted a request
https://www.deb-multimedia.org/lurker/message/20151229.072257.84867e7b.en.html

-- 
Kamaraju S Kusumanchi | http://raju.shoutwiki.com/wiki/Blog

Debian using USB stick on diskless machine

[Ross Boylan]

I have a diskless machine and figured it would be quicker to get it
running off a USB thumb drive than PXE boot.  I'd appreciate any
suggestions or advice.

I stuck Debian live on the thumb drive but discovered changes I made
in that environment were not persistent.  Ran the installer on the
live CD and installed to the rest of the thumb drive.  However, grub
said it could not install to the drive (which, of course, was in use
by the installer system).

I tried chroot'ing into the new system from the live system.  But,
even after mounting /dev and others for the chroot, grub-install says
root@debian:/# grub-install /dev/sda
Installing for i386-pc platform.
grub-install: warning: Attempting to install GRUB to a disk with
multiple partition labels.  This is not supported yet..
grub-install: error: embedding is not possible, but this is required
for RAID and LVM install.

There might be two issues: one, accessing sda, and the second that I
am using one big filesystem on top of LVM; there  is no separate
partition outside of lvm for grub.  This was mostly an oversight,
though I had hoped that current grub could cope.

Because this was a live image the partition table is msdos.  Which is
maybe not so great for grub, which needs more room.  The live image is
about 1G and the thumb drive is 32G.

Thanks for any help.
Ross Boylan

Re: s-nail fails with "Unable to (dot) lock mailbox, aborting operation: Permission denied"

[Rick Thomas]

On Dec 28, 2015, at 2:31 AM, Rick Thomas  wrote:

> With recent Stretch installations the “mail” (or “mailx”) command is 
> satisfied by the s-nail package.
> As it comes fresh out of the box, s-nail has a problem with dotlock files.  
> For example:
> 
>> rbthomas@half:~$ mail
>> Creating dotlock for "/var/mail/rbthomas" .
>> Unable to (dot) lock mailbox, aborting operation: Permission denied
>> Creating dotlock for "/var/mail/rbthomas" .
>> Unable to (dot) lock mailbox, aborting operation: Permission denied
>> rbthomas@half:~$ 
> 
> I don’t know if this is deliberate, or even whether this is the right fix, 
> but the error can be made to go away by doing:
> 
>> sudo chown root:mail /usr/lib/s-nail/s-nail-privsep
>> sudo chmod g+s /usr/lib/s-nail/s-nail-privsep
> 
> If the consensus is that it’s a bug, not a feature, I’ll submit a bug report. 
>  I guess s-nail is the appropriate package?
> 
> Enjoy!
> Rick

Turns out, that this is bug #806858.

Rick

Re: is twitter useful for debian?

[Gener Badenas]

On Mon, Dec 28, 2015 at 7:48 PM, Michael Fothergill <
michael.fotherg...@googlemail.com> wrote:

>
> ​Dear Folks,
>
> I don't use twitter much.   But I noticed that debian is on twitter ie The
> Debian Project  @debian.  ​
>
> What benefit is there in that?
>

Could be useful for announcements


>
> I tried tweeting a bit on twitter but I did not get much response.  So I
> got bored with it.
>
> Some exchanges on twitter can be interesting but with the 140 character
> limit it would not be easy to discuss bugs etc.
>
> Apparently 71% of postings on twitter are ignored.  That is not very
> interactive.
>
> Comments appreciated.
>
> Regards
>
> Michael Fothergill
>
>
>
>
>
>
>
>
>
>
>
>
>
>


-- 
Java  and Groovy

Re: Funny integer problem in wide-dhcpv6-client (20080615-12)?

[Henrique de Moraes Holschuh]

On Mon, 28 Dec 2015, Atle Solbakken wrote:
> I've noticed something strange in dhcp6c where the preferred lifetime of a
> delegated prefix is set to a very high value apparently for no reason. It
> looks like an issue with signing, integer size or such.
> 
> I had look into the source code to debug it, but after building from source,
> the problem disappears so it's kind of hard to figure it out.

Can you give us more information about your enviornment?  Which version of
Debian you're using (and building dhcp6c in), for example...

> Debug output from amd64 binary package built by Debian:
> update_prefix: create a prefix 2001:4660:bffd::/48 pltime=51110824,
> vltime=3600
> 
> Debug output after building from source, pltime is correct:
> update_prefix: create a prefix 2001:4660:bffd::/48 pltime=3600, vltime=3600
> 
> Have anybody else seen this problem or have any clue what might be causing
> this? If the problem only exists at my end there might not be any point

This kind of crap usually happens due to a broken build environment and/or
compiler.  But it could also be caused by ABI breakage involving dynamic
linked objects (libraries, plugins, etc).

> investigating this any further, as it's OK when built from source. Still
> annoying, though :)

Well, this is the sort of thing we *do* want to track down.

-- 
  "One disk to rule them all, One disk to find them. One disk to bring
  them all and in the darkness grind them. In the Land of Redmond
  where the shadows lie." -- The Silicon Valley Tarot
  Henrique Holschuh

Funny integer problem in wide-dhcpv6-client (20080615-12)?

[Atle Solbakken]

Hi

I've noticed something strange in dhcp6c where the preferred lifetime of 
a delegated prefix is set to a very high value apparently for no reason. 
It looks like an issue with signing, integer size or such.


I had look into the source code to debug it, but after building from 
source, the problem disappears so it's kind of hard to figure it out.


Debug output from amd64 binary package built by Debian:
update_prefix: create a prefix 2001:4660:bffd::/48 pltime=51110824, 
vltime=3600


Debug output after building from source, pltime is correct:
update_prefix: create a prefix 2001:4660:bffd::/48 pltime=3600, vltime=3600

Have anybody else seen this problem or have any clue what might be 
causing this? If the problem only exists at my end there might not be 
any point investigating this any further, as it's OK when built from 
source. Still annoying, though :)


Cheers
Atle

Re: Debian on Dell XPS 15

[Henrique de Moraes Holschuh]

On Fri, 25 Dec 2015, Dwijesh Gajadur wrote:
> Hi everyone..I am going to get a new Dell XPS 15 laptop. I want to know if
> I will be able to install and use Debian on it properly without getting any
> issues.

A new Dell XPS 15 will be Intel Skylake-based.  Unless you are feeling
extremely lucky, update it to the very latest BIOS/UEFI *BEFORE* you even
attempt to install Debian.

Also, you likely benefit from keeping that firmware up-to-date for a while
yet.  If you cannot flash updates from Linux, keep a Windows partition
around to do it.

You have been warned.

(BTW, the above applies to any Intel Skylake-based system you want to run
Debian on).

> Has anyone ever used Debian on Dell XPS 15 laptop? Please share your
> experience.

You may have to boot with the kernel parameter 'intel_idle.max_cstate=0' if
you experience crashes.  This will drain the battery a bit more, though.

-- 
  "One disk to rule them all, One disk to find them. One disk to bring
  them all and in the darkness grind them. In the Land of Redmond
  where the shadows lie." -- The Silicon Valley Tarot
  Henrique Holschuh

Re: Need a calculator that knows about coulombs

[Gene Heskett]

On Monday 28 December 2015 12:35:01 John Hasler wrote:

> Gene writes:
> > And I haven't carved up a pcb to hold it all yet either.
>
> I never do pcbs for this sort of thing.  I just solder point to point
> on a breadboard.  Printed circuit boards are for production.

Which, since I was in the environs of a surviving radio shack this 
morning, I stopped in and picked up a couple breadboards.
>
> > The qucs simulation is hampered by its lack of two part models, a
> > decent hexfet, and a total lack of anything that looks like a 40 Amp
> > SSR.
>
> The circuit simulators in the Debian archive have libraries of models
> for all these things.
>
> BTW many SSRs can be driven directly by logic.  Here's one:
> https://www.sparkfun.com/products/13015

What I have are clones of the crydom SSR, 3-32 volts & around 10 mills to 
fire them. I have it making about 1.35 volts into a 100k load right now, 
and since the shack didn't have any small hexfets, I've used some 
2na I had, and have put 100 ohms between the collector of a 2na 
and the - bolt, with about 18 volts connected to the + bolt.  Not tested 
yet, my back gave out, but probably by the time the day is done I'll 
know if it works.  Response time looks to be around 100 milliseconds, 
plenty fast enough when driven by my function generator set at 4.5 v 
p-p.  The Breakout Board can swing to nearly 5 volts, but probably 
capable of less current than the generator, which could slow the 
response, but this job can tolerate a seconds lag if it has to.

> And here's an application note:
> http://www.wrcakron.com/catalog/10b_SSR_Applications.pdf

I'll grab that, it could be educational.  Thanks.

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: zfs list shows no space while only 50% real size of files.

[David Christensen]

On 12/28/2015 07:00 AM, Mimiko wrote:

I use zfs to create zvols. This is the zpool list:

...

Here logicalused display 1.85TB of data. Why rest of the space is not
available?


For detailed ZFS on Linux questions, I'd suggest the zfs-discuss mailing 
list:


http://zfsonlinux.org/

http://zfsonlinux.org/lists.html

http://list.zfsonlinux.org/mailman/listinfo/zfs-discuss


David

Re: is twitter useful for debian?

[John Hasler]

Michael Fothergill writes:
> I don't use twitter much. But I noticed that debian is on twitter ie The
> Debian Project @debian. ​

> What benefit is there in that?

Nails down the name "debian" so that a troll can't get it.
-- 
John Hasler 
jhas...@newsguy.com
Elmwood, WI USA

Re: Need a calculator that knows about coulombs

[John Hasler]

Gene writes:
> And I haven't carved up a pcb to hold it all yet either.

I never do pcbs for this sort of thing.  I just solder point to point on
a breadboard.  Printed circuit boards are for production.
 
> The qucs simulation is hampered by its lack of two part models, a decent 
> hexfet, and a total lack of anything that looks like a 40 Amp SSR.

The circuit simulators in the Debian archive have libraries of models
for all these things.

BTW many SSRs can be driven directly by logic.  Here's one:
https://www.sparkfun.com/products/13015

And here's an application note:
http://www.wrcakron.com/catalog/10b_SSR_Applications.pdf
-- 
John Hasler 
jhas...@newsguy.com
Elmwood, WI USA

zfs list shows no space while only 50% real size of files.

[Mimiko]

Hello.

I use zfs to create zvols. This is the zpool list:
NAME  SIZE  ALLOC   FREE  EXPANDSZ   FRAGCAP  DEDUP  HEALTH  ALTROOT
zfspool  14.5T  10.7T  3.75T -13%74%  1.00x  ONLINE  -

I have 3.75T free to create more zvols or extend existing vols.

I have several zvols created as blockdevices, zfs list -o space:
NAMEAVAIL  USED  USEDSNAP  USEDDS  USEDREFRESERV  USEDCHILD
zfspool 0  11.5T0335K  0  11.5T
zfspool/1   57.4G  103G 0   45.8G  57.4G  0
zfspool/2   120G   4.13T0   4.01T   120G  0
zfspool/3   50.3G  103G 0   52.9G  50.3G  0
zfspool/4   51.1G  51.6G0445M  51.1G  0
zfspool/5   1008G  1.03T0   48.6G  1008G  0
zfspool/6   353G   413G 0   59.4G   353G  0
zfspool/7   1.03T  1.03T0   45.1M  1.03T  0
zfspool/8   75.4G  103G 0   27.7G  75.4G  0
zfspool/9   0  4.58T0   4.58T  0  0

As you can see, zfspool hase no free space available. This volume was 
created with this:

zfs create -V 4T zfspool/9
zfs set compression=lz4 zfspool/9
mkfs.ext4 -b 4096 -q /dev/zvol/zfspool/9
mkdir -p /mnt/zfs/9
mount /dev/zvol/zfspool/9 /mnt/zfs/9

The problem is that while zfs list reports 0 free space, actual data 
ocupies only about a half of the space (~1.5TB). Also df -h /mnt/zfs/9:


Filesystem  Size  Used Avail Use% Mounted on
/dev/zd48   4.0T  1.6T  2.3T  41% /mnt/zfs/oracle

Which is correct.

I don't use deduplication or snapshots.

This is the output of zfs get all zfspool/9
NAMEPROPERTY  VALUE  SOURCE
zfspool/oracle  type  volume -
zfspool/oracle  creation  Thu Mar 26  9:33 2015  -
zfspool/oracle  used  4.58T  -
zfspool/oracle  available 0  -
zfspool/oracle  referenced4.58T  -
zfspool/oracle  compressratio 1.00x  -
zfspool/oracle  reservation   none   default
zfspool/oracle  volsize   4T local
zfspool/oracle  volblocksize  8K -
zfspool/oracle  checksum  on default
zfspool/oracle  compression   lz4local
zfspool/oracle  readonly  offdefault
zfspool/oracle  copies1  default
zfspool/oracle  refreservation4.13T  local
zfspool/oracle  primarycache  alldefault
zfspool/oracle  secondarycachealldefault
zfspool/oracle  usedbysnapshots   0  -
zfspool/oracle  usedbydataset 4.58T  -
zfspool/oracle  usedbychildren0  -
zfspool/oracle  usedbyrefreservation  0  -
zfspool/oracle  logbias   latencydefault
zfspool/oracle  dedup offinherited 
from zfspool

zfspool/oracle  mlslabel  none   default
zfspool/oracle  sync  standard   default
zfspool/oracle  refcompressratio  1.00x  -
zfspool/oracle  written   4.58T  -
zfspool/oracle  logicalused   1.85T  -
zfspool/oracle  logicalreferenced 1.85T  -
zfspool/oracle  snapshot_limitnone   default
zfspool/oracle  snapshot_countnone   default
zfspool/oracle  snapdev   hidden default
zfspool/oracle  context   none   default
zfspool/oracle  fscontext none   default
zfspool/oracle  defcontextnone   default
zfspool/oracle  rootcontext   none   default
zfspool/oracle  redundant_metadataalldefault
zfspool/oracle  shareiscsioffdefault

Here logicalused display 1.85TB of data. Why rest of the space is not 
available?


Thank you.

Re: Need a calculator that knows about coulombs

[Gene Heskett]

On Monday 28 December 2015 04:30:06 Jörg-Volker Peetz wrote:

> Gene Heskett wrote on 12/23/15 16:12:
> > On Wednesday 23 December 2015 08:49:34 Jörg-Volker Peetz wrote:
> >> e.g., Qucs
> >> https://en.wikipedia.org/wiki/Quite_Universal_Circuit_Simulator
> >>
> >> Regards,
> >> jvp.
> >
> > Sounds like exactly what might be useful here. But no qucs exists in
> > the wheezy repos. Dowloaded src, will see if it will build on
> > wheezy.
>
> Could you get Qucs working?

Yes and no. I've drawn a schematic but cannot make it simulate it.  Going 
out to the workbench and building it with my limited junk box parts 
assortment results in a working circuit, if you don't mind your coffee 
being lukewarm by the time it fires. :(  I have extracted some more 
parts from scrap electronics, which may result in too fast a response, 
some 0.01 uF caps for the pump cap, and a couple .22 uF's for the 
storage cap, and 1n914 si diodes for test, but I have some schottky's, 
an assortment of tantalum caps, a few hexfets & such coming from 
aliexpress via China Post, but they'll be into 2016 by the time they get 
here.  And I haven't carved up a pcb to hold it all yet either. PCB I 
have, but I'll have to carve isolation islands by hand and surface mount 
the leaded parts. Thats time consuming when the mills spindle is 100x 
too slow to spin an etching tool correctly. 2" per minute or less. I 
have done it, using eagle for the drawing source, but the etching is 
similar to watching paint dry. I will likely use a 1/16" carbide mill so 
I don't have to retrace so many times to get an adequate separation of 
the copper. The std v tip etch tool only cuts a thou deep and wide 
unless you want to wear it out cutting the glass under the copper.
 
The qucs simulation is hampered by its lack of two part models, a decent 
hexfet, and a total lack of anything that looks like a 40 Amp SSR.  I 
tried to make an enhancement mode hexfet from an n channel MOSFET, but 
it wants the blanks filled in with info only the mask designer and 
silicon foundry would know.  In that regard, its obvious I don't know 
what the heck I am doing, darn it. 

> By the way, I just remember there's the Emacs builtin calculator
> "calc" which a.o. can handle expressions with units.

I've looked at units, and it could probably be scripted to do what I 
want, but I'd need a better grasp of math to do it and get results out 
of gnuplot by 1/31/16. :(  I can capture it on a digital scope in 10 
seconds once the soldering iron has pasted a test circuit together. :)

> Regards,
> jvp.


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: Need a calculator that knows about coulombs

[Gene Heskett]

On Monday 28 December 2015 01:59:44 Gener Badenas wrote:

Back on the list where this belongs.

> On Wed, Dec 23, 2015 at 9:09 AM, Gene Heskett  
wrote:
> > Greetings all;
> >
> > Do we have such a beast?
> >
> > Thanks.
>
> There might be some projects in github that does this.

Qucs seems to be the default plaything for such these days.  However I am 
having a heck of a time making it work to actually do the simulation of 
the charge-pump detection circuit I have drawn, basically 2 caps, 2 
diodes and a loading R on the output.

I have joined the qucs-help mailing list, but it appears I may be the 
only subscriber.  No other activity in about 18 hours anyway.

I've looked at git-hub a few times but it doesn't take long to get lost 
in that corn maze.

> > 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)
> > Some mill pix are at:
> > Genes Web page 


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: KTorrent: Turn off computer when finished

[Gene Heskett]

On Monday 28 December 2015 01:14:55 Allan Aguilar wrote:

> On 12/28/2015 06:06 AM, Brandon Vincent wrote:
> > On Sun, Dec 27, 2015 at 10:20 PM, Allan Aguilar 
 wrote:
> >> Hello,
> >>
> >> Is there a way to set up KTorrent in a way that the computer turns
> >> off when all downloads are complete?
> >
> > Hi Allan,
> >
> > If you go to the "View" menu, there should be a "Plugins" option.
> > One of the plugins is "Shutdown" that does what you are looking for.
> >
> > Brandon Vincent
>
> That's exactly what I was looking for. Thanks, Brandon. I really
> appreciate it.

It should be mentioned that the torrent worked as well as it did because 
others did NOT do that, but left their machines running so that your 
machine could pull snippets from theirs, thereby effectively multiplying 
the original source sites available bandwidth by the upload bandwidth 
each site has times the number of sites shareing the torrent, which in 
turn made your download quite a few times faster.  Shutting your machine 
off slows the torrent for others, and this isn't fair, nor the FOSS 
attitude.  You gained from using the torrent, you should pay it back.

This process actually began, your machine sharing back what it has of the 
requested file, after the first 64 kilobytes had been pulled and stored 
on your machine.  That isn't a full explanation of coarse because the 
torrent is a random access protocol, and theoretically the first 64k 
block your machine receives, could in fact haver a small chance of being 
the last 64k of the file.  A torrent tracks where the individual block 
goes in the file, taking any block advertised by others that it needs 
until it has every piece of the file and gets the same crc check value 
as the source had, on a per 64k packet.  Any missing pieces can be 
requested, and whoever has it will send it.


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 

is twitter useful for debian?

[Michael Fothergill]

​Dear Folks,

I don't use twitter much.   But I noticed that debian is on twitter ie The
Debian Project  @debian.  ​

What benefit is there in that?

I tried tweeting a bit on twitter but I did not get much response.  So I
got bored with it.

Some exchanges on twitter can be interesting but with the 140 character
limit it would not be easy to discuss bugs etc.

Apparently 71% of postings on twitter are ignored.  That is not very
interactive.

Comments appreciated.

Regards

Michael Fothergill

Build Bacula 7.2.x on Debian Jessie

[Daniel Bareiro]

Hi all!

In a virtual machine with Debian Jessie I am testing Bacula 7.2.x.
Initially I had downloaded the tarball of here [1] (published in
2015-08-13), but here I got some errors when running the configuration
script:

---
make[2]: Entering directory '/usr/src/bacula-7.2.0/platforms/systemd'
Makefile:24: *** missing separator.  Stop.
make[2]: Leaving directory '/usr/src/bacula-7.2.0/platforms/systemd'
Makefile:88: recipe for target 'clean' failed
make[1]: *** [clean] Error 2
make[1]: Leaving directory '/usr/src/bacula-7.2.0/platforms'
Makefile:280: recipe for target 'clean' failed
make: *** [clean] Error 1
---

According I read on this message [2], it seems that this was fixed by
Kern Sibbald on 2015-08-18 in the Git repository, but the tarball was
not updated. This makes me wonder if the branch 7.2 is recommended for
production or there is another that we consider stable branch.

After cloning the "Branch-7.2", I was able to compile without these
errors. But I'm having some difficulties with the startup script for the
Storage Daemon and Director:

---
root@baculatest:~# /etc/init.d/bacula-dir start
Starting Bacula Director: start-stop-daemon: unable to start
/usr/sbin/bacula-dir (Permission denied)
bacula-dir
---
root@baculatest:~# /etc/init.d/bacula-sd start
Starting Bacula Storage Daemon: start-stop-daemon: unable to start
/usr/sbin/bacula-sd (Permission denied)
bacula-sd
---

In the first case, the following line is executed:

---
start-stop-daemon --start --quiet --pidfile /var/run/bacula-dir.9101.pid
--chuid bacula:bacula --exec /usr/sbin/bacula-dir -- -c
/etc/bacula/bacula-dir.conf
---

And in the second runs the following line:

---
start-stop-daemon --start --quiet --pidfile /var/run/bacula-sd.9103.pid
--chuid bacula:bacula --exec /usr/sbin/bacula-sd -- -c
/etc/bacula/bacula-sd.conf
---

I've created the "bacula" user and group.

I was making a comparison in a Jessie server using Storage Daemon an
Director with Bacula 5.2.6 packaged. Here the binaries have the same
owners (root:root), but there are some differences in how the binaries
are invoked:

---
start-stop-daemon --start --quiet --pidfile $PIDFILE --oknodo --exec
/usr/sbin/bacula-sd -- -c /etc/bacula/bacula-sd.conf -u bacula -g tape
---
start-stop-daemon --start --quiet --pidfile $PIDFILE --oknodo --exec
/usr/sbin/bacula-dir -- -c /etc/bacula/bacula-dir.conf -u bacula -g bacula
---

Beyond that difference, I found another difference from the binaries
permissions, so I changed the permissions for "others":

--
root@baculatest:/etc/bacula# ll /usr/sbin/bacula-*
-rwxr-x--- 1 root root 2929712 Dec 24 19:50 /usr/sbin/bacula-dir
-rwxr-x--- 1 root root  940720 Dec 24 19:50 /usr/sbin/bacula-fd
-rwxr-x--- 1 root root 1928928 Dec 24 19:50 /usr/sbin/bacula-sd

root@baculatest:/etc/bacula# chmod o+rx /usr/sbin/bacula-*
--

Now I do not get the permissions error, but for some reason the
processes are not initiated:

--
root@baculatest:~# systemctl start bacula-sd.service
root@baculatest:~# systemctl status bacula-sd.service
● bacula-sd.service - LSB: Start Bacula Storage daemon at boot time
   Loaded: loaded (/etc/init.d/bacula-sd)
   Active: active (exited) since Sat 2015-12-26 11:30:12 ART; 9s ago
  Process: 2001 ExecStop=/etc/init.d/bacula-sd stop (code=exited,
status=0/SUCCESS)
  Process: 3026 ExecStart=/etc/init.d/bacula-sd start (code=exited,
status=0/SUCCESS)

Dec 26 11:30:12 baculatest bacula-sd[3026]: Starting Bacula Storage
Daemon: bacula-sd
--
root@baculatest:~# systemctl start bacula-dir.service
root@baculatest:~# systemctl status bacula-dir.service
● bacula-dir.service - LSB: Start Bacula Director daemon at boot time
   Loaded: loaded (/etc/init.d/bacula-dir)
   Active: active (exited) since Sat 2015-12-26 11:31:04 ART; 5s ago
  Process: 2010 ExecStop=/etc/init.d/bacula-dir stop (code=exited,
status=0/SUCCESS)
  Process: 304

s-nail fails with "Unable to (dot) lock mailbox, aborting operation: Permission denied"

[Rick Thomas]

With recent Stretch installations the “mail” (or “mailx”) command is satisfied 
by the s-nail package.
As it comes fresh out of the box, s-nail has a problem with dotlock files.  For 
example:

> rbthomas@half:~$ mail
> Creating dotlock for "/var/mail/rbthomas" .
> Unable to (dot) lock mailbox, aborting operation: Permission denied
> Creating dotlock for "/var/mail/rbthomas" .
> Unable to (dot) lock mailbox, aborting operation: Permission denied
> rbthomas@half:~$ 

I don’t know if this is deliberate, or even whether this is the right fix, but 
the error can be made to go away by doing:

> sudo chown root:mail /usr/lib/s-nail/s-nail-privsep
> sudo chmod g+s /usr/lib/s-nail/s-nail-privsep

If the consensus is that it’s a bug, not a feature, I’ll submit a bug report.  
I guess s-nail is the appropriate package?

Enjoy!
Rick

Re: Need a calculator that knows about coulombs

[Jörg-Volker Peetz]

Gene Heskett wrote on 12/23/15 16:12:
> On Wednesday 23 December 2015 08:49:34 Jörg-Volker Peetz wrote:
> 
>> e.g., Qucs
>> https://en.wikipedia.org/wiki/Quite_Universal_Circuit_Simulator
>>
>> Regards,
>> jvp.
> 
> Sounds like exactly what might be useful here. But no qucs exists in the 
> wheezy repos. Dowloaded src, will see if it will build on wheezy.
> 

Could you get Qucs working?

By the way, I just remember there's the Emacs builtin calculator "calc" which
a.o. can handle expressions with units.

Regards,
jvp.

Re: Recovering data from a Raid 1Sata HD

[tomas]

-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1

On Sun, Dec 27, 2015 at 02:48:45PM +0100, Pascal Hambourg wrote:
> to...@tuxteam.de a écrit :
> > 
> > On Wed, Dec 23, 2015 at 07:41:00AM -0300, Renaud OLGIATI wrote:
> > 
> >> Or do I have to first create a (failed) array with mdadm ?
> > 
> > For RAID 1 it shouldn't be necessary, AFAIR
> 
> You cannot mount directly a RAID 1 partition with format 1.1 or 1.2
> because the RAID superblock gets in the way. You must mount it with
> loop,offset (see my other reply in this thread) or assemble it as a
> degraded array and mount the array.

Ah -- thanks. That was the missing piece. Seems my RAID knowledge is
a bit rusty, sorry.

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

iEYEARECAAYFAlaA6fEACgkQBcgs9XrR2kabagCeOLlyX+aUKGLlZ96MPyQJ7rXf
trIAnR7vJQFLnNX7YNRUr60FmXMtMWB5
=u/ht
-END PGP SIGNATURE-