Re: Virtual Machines: Newbie / novice questions

[Kent West]

On May 14, 2017 06:51,  wrote:


   * Is it right to consider that there is somehow a base OS and the other
VMs somehow are set up on top of it--you boot into that base OS first, and
then have the ability to switch to another VM.  OK, I'm thinking that is
not the case--I think I've heard of things like a hypervisor (??), and now
I'm guessing that the machine boots first into the hypervisor and then from
there boot into any of one or more VMS, one of which would contain what I
consider my main or base OS?


I'm unsure about the hypervisor route, but in my case, I'm running Debian
Sid on a 12 MiB Dell with a TB SSD. Then on top of that I'm running
VirtualBox (aptitude install virtualbox). Within that app, I've created
several virtual machines (4 GiB RAM, 120 GiB drive, 64M video RAM, etc).
For each one, you turn it on, and it's just like turning on a physical PC
(with a VirtualBox firmware (BIOS/UEFI) instead of a Dell firnware). The
firmware splashscreen comes up, and complains that it can't find an OS to
boot. I then load a Windows7 or 10 or XP or DOS or Debian netinstall or etc
.ISO (or similar) file into the virtual machine's virtual CD drive (or a
real CD into the real drive, mapping that to the VB's virtual CD drive),
and boot and install that OS. Now I can boot into that OS at will, in a
window (or full-screen), and have a fully-functioning second (or third, or
tenth, with enough resources) sitting on top of my Debian Sid base. I can
leave those VMs running, or shut them down, or suspend them, as I wish.
With VirtualBox I can share a folder/partition, and/or the clipboard for
cut/paste, either direction.


   * I know (or I'm 99.9% sure) that the software for a VM is on separate
partitions (except maybe some are or could be shared between the VM and the
"base" machine, such as for my documents and such (things that I'd want to
be
accessible from any of the VMs).


Not necessarily a separate partition. The virtual machine is really just a
few files, but the main file is essentially the entire VM. You can move it
to a separate partition, but no need. You can copy/move/backup that file to
copy/move/restore the entire virtual machine, even to another physical box.


   * Can I keep a VM in a running state for instant switching to it?  (I
presume that in that case, it will be using RAM, on the other hand, it could
be in a non-running (either non-booted or suspended-to-disk) and not be
using
RAM?


Yes.


   * I presume that booting a VM takes  about the same amount of time,
maybe a
little longer than booting the base machine /  OS (well, unless the OS in
the
VM is significantly "smaller" in terms of footprint and services started at
boot)?


Pretty close to the same amount of time as physical hardware, with a little
bit of overhead.


   * Can I C between (applicatons in) different VMs and/or the base OS?


With VirtualBox, yes. VMWare, also, if I recall correctly. I'm not sure
about other vm apps, like Qemu.


Thanks!


Lots of RAM and a roomy SSD make virtual machines handy to have around. A
lack of those resources might present you with a miserable experience.

-- 
Kent

Re: Update Notifier

[Matthew McKinnon]

On Sun, May 14, 2017 at 05:03:31PM +0900, Mark Fletcher wrote:
> On Fri, May 05, 2017 at 04:32:44PM +, sare...@att.net wrote:
> > Why doesn't Debian 8 Cinnamon notify when updates are ready to install 
> > after all these years Debian has existed? Don't tell me there is one, 
> > because after installing Debian I waited a long time to see if a 
> > notification would pop up. It never did. I know about doing apt-get 
> > update && apt-get upgrade,but why should we have to use a command 
> > line? It makes me wonder about Debian security. Would you please put 
> > an update notifier in all your versions of Debian.
> > 
> 
> Who do you think you are talking to?
> 

In Debian Jessie install pk-update-icon from debian-backports

In Debian Stretch install pk-update-icon.

This package works both on Cinnamon, XFCE4 and Gnome. Not sure on the other DEs.

Thanks
Matt

Re: how to copy files

[Long Wind]

archive just means that?
Thank Dan Ritter!
l will try it later on.

Connection to ubcsecure.

[peter]

I'm interested to have Debian 8 connect to ubcsecure.
https://autoconnect.it.ubc.ca/xc/page4_download.html 

The Cloupath app works in Android.  
https://autoconnect.it.ubc.ca/xc/page4_download.html 
The analogous gadget works in the Fedora based XO system but 
in Jessie it detects no network.

To configure at a more fundamental level the most detailed 
instructions I've found are here.
https://ubc.service-now.com/kb_view_customer.do?sysparm_article=KB0013556

Wicd is installed.  Among the 15 possibilities in the encryption 
menue, "PEAP with TKIP/MSCHAPV2" is correct?  

The UBC instructions cited above refer to Thawte_Premium_Server_CA.pem 
whereas Jessie has 
/etc/ssl/certs/thawte_Primary_Root_CA{.pem,_-_G2.pem,_-_G3.pem} .  
The Thawte Web site lists the "Premium Server CA" under "Retired 
Roots".  The instructions are outdated?

Any tips are welcome.

Thanks,   ... Peter E.



-- 

123456789 123456789 123456789 123456789 123456789 123456789 123456789
Tel: +1 360 639 0202  Pender Is.: +1 250 629 3757
http://easthope.ca/Peter.html  Bcc: peter at easthope. ca

Re: how to copy files

[Dan Ritter]

On Mon, May 15, 2017 at 04:57:35AM +0800, Long Wind wrote:
> I have a folder in hard drive and a backup of it is made in a USB disk
> Now some new files are added to the folder in hard drive
> how to make the folder in USB disk the same as in hard drive?
> 
> I want a program that can check the two folders
> if file size and file name are same, they are regarded as same
> if not, update is needed
> 
> I have try rsync, but can't find such options

That's the default for rsync.

rsync /path/to/originals/  /path/to/copies/

add -r for recursion into subdirectories
add -v for verbosity
add -a (archive) to keep ownership, groups, and permissions
   identical. That usually needs root privileges.

There are a lot of other options to rsync; you can have
unrecognized files in the destination ignored or deleted.

You can require the contents of all files to match, rather than
just their name and size.

You can have special exclusions and inclusions.

If you need changes on both directories to be reflected in each
other, you want unison instead of rsync.

-dsr-

Re: how to copy files

[Lisi Reisz]

On Sunday 14 May 2017 21:57:35 Long Wind wrote:
> I have a folder in hard drive and a backup of it is made in a USB disk
> Now some new files are added to the folder in hard drive
> how to make the folder in USB disk the same as in hard drive?
>
> I want a program that can check the two folders
> if file size and file name are same, they are regarded as same
> if not, update is needed
>
> I have try rsync, but can't find such options

cp -Rpu /home/. /backup/

will copy only new stiff, but all new stuff.

That is for everything in home.  But you could do the same for one folder or 
sub-folder, or one user.

Lisi

Re: Virtual Machines: Newbie / novice questions

[Ben Caradoc-Davies]

On 14/05/17 23:51, rhkra...@gmail.com wrote:

* Can I C between (applicatons in) different VMs and/or the base OS?


While you can do this in emulators like VMWare Player, QEMU does not 
support it. A popular solution for QEMU is Spice 
. A simpler but less 
feature-rich solution is connecting with ssh, as described earlier in 
this thread.


Kind regards,

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

Re: Virtual Machines: Newbie / novice questions

[Ben Caradoc-Davies]

On 15/05/17 02:27, Thomas Schmitt wrote:

For my VMs which shall be operated without graphics screen, i let qemu
forward a host port (here 4567) to port 22 of the guest, where a SSH
server will be set up by installation to offer login:
  qemu-system-x86_64 ... \
 -net nic -net user,hostfwd=tcp::4567-:22


And if you want QEMU to listen only on localhost and not allow remote 
network connections, this becomes:


qemu-system-x86_64 ...
-net nic -net user,hostfwd=tcp:127.0.0.1:4567-:22

Kind regards,

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

how to copy files

[Long Wind]

I have a folder in hard drive and a backup of it is made in a USB disk
Now some new files are added to the folder in hard drive
how to make the folder in USB disk the same as in hard drive?

I want a program that can check the two folders
if file size and file name are same, they are regarded as same
if not, update is needed

I have try rsync, but can't find such options

Re: Bug#860543: 'Gave up waiting for suspend/resume device'

[Larry Dighera]

Hello Ben,

Thank you for your prompt response to my inquiry, and your kind assistance.

My comments in-line below:


On Sat, 13 May 2017 00:40:47 +0100, Ben Hutchings 
wrote:

>On Fri, 2017-05-12 at 21:26 +, Larry Dighera wrote:
>> Dear Mr. Hutchings,
>> Todays update appears to have caused my Debian Stretch Linux system
>> to fail to boot with this error message: 
>> 
>>     "Gave up waiting for suspend/resume device..."
>
>That (by itself) does not indicate a boot failure; it only means that
>resume from disk (hibernation) won't work.  
>

Previous to the update, hibernation worked.  

>
>Do any messages appear afterward?
>

Yes.  There is a bit after that.

"Gave up waiting for root file system device"
or

"Gave up waiting for suspend/resume device"

Then some suggestions presumably about how to address the issue:

"Boot Args. Cat /proc/cmdline"
something about "check root delay" perhaps not long enough... 
missing modules..."
"UUID 2d182fb2... doesn't exist."

>
>> The Grub menu is still accessible, but I'm clueless how to return the
>> system to operational status.
>> Please provide complete step-by-step instructions to restore
>> successful system booting through the Grub menu.  
>
>One of the 'recovery mode' item under Advanced Options will enable
>verbose logging to the screen and will perhaps get you to a shell.
>

Unfortunately, selecting the 'recovery mode' grub menu option is useless.
There is something about 'busybox' and a '(initramfs)' prompt(?), but the
system fails to respond to keyboard or mouse input at that point.  So
'recovery mode' appears to be useless.

However, by entering 'e' at the grub menu, I can edit the 'command line.  At
your suggestion above, I changed 'quite' to 'verbose', but nothing seems to
have changed when rebooting.

By entering 'c' at the grub menu, I can get a 'Grub>' prompt from which I am
able to 'ls' and 'cat' and probably a hundred or more other commands that
can be displayed with the 'help' command.  It is necessary to enter 'set
pager=1' to see them before they scroll off screen.  Entering 'help
' provides some additional terse (cryptic) information about the
command.

I don't find any sort of editing command, and redirection '>>' attempts
fail.  

Can you give me a clue to necessary commands to go about returning my system
to operational status from this 'Grub>' prompt?  

>
>In any case, you should open a *new* bug report rather than appending
>to this one.
>
>Ben.

It was not my intent to open a bug report at all.  I apologize for any issue
I may have caused.

In the information at this link:
https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=860543 you stated:

"Is there a resume device specified in
/etc/initramfs-tools/conf.d/resume and does it exist?"

I found /etc/initramfs-tools/conf.d/resume.  It contains the single entry:

"RESUME=UUID=f5aa19f3-7707-4bab-bd18-15d6124a0147"

I have no idea where to find that file/device/?.

I just want my system to boot again.  Please assist me in reaching that
goal.

Best regards,
Larry

PS:
On that page, you also stated: 

> ...
> /etc/initramfs-tools/conf.d/resume exists and did not contain a valid
> device. The problem still persists, if I enter a valid UUID, or remove
> the file.

This issue should only affect systems that have a nonexistent device
specified in the configuration, or that have a swap device that isn't
suitable as a resume device because it's not set up early enough.
Previously we would only look for the resume device once, so this
didn't hurt, but it also meant that resume was broken on some systems.

Unfortunately, as you've seen, there isn't currently a way to
completely disable use of a resume device when building the initramfs. 
It *is* possible to do so at boot time (kernel parameter 'noresume' or
'resume=').

I think I need to add:

- The option to disable use of a resume device in the configuration
  (e.g. RESUME=none)
- A warning on upgrade if the configured resume device doesn't exist or
  is unlikely to be available

Ben.

Is that what I need to do?  If so, please provide step-by-step keystrokes I
need to enter at the "Grub>" prompt. 

Re: Virtual Machines: Newbie / novice questions

[Patrick Bartek]

On Sun, 14 May 2017 07:51:18 -0400 rhkra...@gmail.com wrote:

> I've never used a VM but I wanted to ask some newbie questions about
> them.  

Then first you should read the manual and the FAQs.  Here's the link to
the Documentation page for VirtualBox, one of the more popular and
easy to set up VMs.

https://www.virtualbox.org/wiki/Documentation

B

Re: Virtual Machines: Newbie / novice questions

[Thomas Schmitt]

Hi,

rhkra...@gmail.com wrote:
>  * Is it right to consider that there is somehow a base OS

Normally there is a single real computer with its OS that runs virtual
computers as "guests".
(The Cloud might have its own habits to host VMs.)


> I'm guessing that the machine boots first into the hypervisor and then from
> there boot into any of one or more VMS, one of which would contain what I
> consider my main or base OS?

I use my real machine with its Debian 8 as my home system and have several
VMs for various development purposes. I operate them via SSH from terminal
windows of the host system.

The VMs are normal processes on the host OS which may have a host graphics
window to show their own graphics screen and to get user input. They may
have access to some hardware components of the host system (e.g. a DVD drive).
Most important is their connection to the host system and the internet
via a virtual network card.


> * I know (or I'm 99.9% sure) that the software for a VM is on separate 
> partitions

It is typically on virtual disks which on the host system are data files
or disk partitions. Sharing of files may e.g. be done via services over
the virtual network.


>* Can I keep a VM in a running state for instant switching to it?

Yes. As long as your real machine has enough memory to serve all processes
of the host OS and as long as you don't need a piece of hardware which
the VM uses exclusively.


> I presume that in that case, it will be using RAM, on the other hand,
> it could be in a non-running (either non-booted or suspended-to-disk)
> and not be using RAM?

If the real RAM of the host system does not suffice, then swapping will
begin. If the swap space is exhausted, processes will die.


>* I presume that booting a VM takes  about the same amount of time, maybe
> a little longer than booting the base machine

If the VM has the same CPU type as the host system and can use kvm, then
it is not supposed to compute much slower than on real hardware. Access to
virtual disks might be slow compared to the disk speed on the host system.


>* Can I C between (applicatons in) different VMs and/or the base OS?

Afaik, the graphics windows of VM do not show text but rather mindless
pixels. So no copy+paste.
But you may use the display of your host system as X terminal which starts
applications on the VMs. (As said, the virtual network is essential in most
use cases.)

-

Try it.

For example, on a modern 64 bit machine with 8 GiB of RAM, get one of the
ISOs at
  https://cdimage.debian.org/debian-cd/current-live/amd64/iso-hybrid/
(let's say: debian-live-8.8.0-amd64-xfce-desktop.iso )
and start qemu+kvm with 4 GiB of RAM by

  qemu-system-x86_64 -enable-kvm -m 4096 -cdrom 
debian-live-8.8.0-amd64-xfce-desktop.iso  

A graphics window should pop up and show you the boot menu.

If you create a file for a virtual disk by command qemu-img

  qemu-img create vm_64gb_disk.qemu 64G

you may attach it to the starting VM by

  qemu-system-x86_64 ... \
 -hda vm_64gb_disk.qemu

For my VMs which shall be operated without graphics screen, i let qemu
forward a host port (here 4567) to port 22 of the guest, where a SSH
server will be set up by installation to offer login:

  qemu-system-x86_64 ... \
 -net nic -net user,hostfwd=tcp::4567-:22 

When the system is installed, i add option -nographic and access the
then headless system by

  ssh -p 4567 my_user@localhost

The option -cdrom and the ISO image are then not used any more.


There are VM control tools by which one can create and manage VMs.
But i have no experience with them.


Have a nice day :)

Thomas

Re: How stable is the frozen stretch?

[Brian]

On Sun 14 May 2017 at 08:57:55 -0500, Richard Owlett wrote:

>   1. Can I retroactively choose "include all available drivers" and
>  if so, how?

https://wiki.debian.org/Initrd

Re: How stable is the frozen stretch?

[Richard Owlett]

Thank you.

On 05/14/2017 07:09 AM, Joe wrote:

On Sun, 14 May 2017 04:57:07 -0500
Richard Owlett  wrote:


On 05/14/2017 02:40 AM, Joe wrote:

On Sat, 13 May 2017 20:54:04 -0400
RavenLX  wrote:


On 05/13/2017 12:40 PM, Richard Owlett wrote:

I have a partition whose label is "common".


I could almost smack myself in the head. I had done that when I
used to dual-boot Windows / Linux (now you can see why I'm not a
fan of dual-boot, I guess! LOL!). I also used to dual boot SolydX
and SolydK distributions.

In my VMs, I do use the guest additions and have a shared directory
(necessity for what I need to do as well).

But that was one way I had shared data.

However, I still balk at dual-booting (like you balk at VMs :) )



Another option I've used for some years: a pocket-sized USB hard
drive with an i386 all-modules installation, which boots on
practically anything that isn't an ARM.



Can you clarify what you mean by the phrase "i386 all-modules
installation"? TIA




Question during installation: 'Do you want all driver modules or just
the ones that apply to this hardware' or something similar. On every
other occasion I go for 'just the ones needed'.

Re: How stable is the frozen stretch?

[Richard Owlett]

On 05/14/2017 06:46 AM, David Wright wrote:

On Sun 14 May 2017 at 04:57:07 (-0500), Richard Owlett wrote:

On 05/14/2017 02:40 AM, Joe wrote:

On Sat, 13 May 2017 20:54:04 -0400
RavenLX  wrote:


On 05/13/2017 12:40 PM, Richard Owlett wrote:

I have a partition whose label is "common".


I could almost smack myself in the head. I had done that when I used
to dual-boot Windows / Linux (now you can see why I'm not a fan of
dual-boot, I guess! LOL!). I also used to dual boot SolydX and SolydK
distributions.

In my VMs, I do use the guest additions and have a shared directory
(necessity for what I need to do as well).

But that was one way I had shared data.

However, I still balk at dual-booting (like you balk at VMs :) )



Another option I've used for some years: a pocket-sized USB hard drive
with an i386 all-modules installation, which boots on practically
anything that isn't an ARM.



Can you clarify what you mean by the phrase "i386 all-modules installation"?


Recall this d-i screen?


I do now ;)
Actually initrd and initramfs and were on my "google today" list.
That may that may explain something I've recently noticed.
As I experiment with many Debian versions and configurations I routinely 
can multi-boot among a a half dozen (or more) choices.


As I've had some spectacularly failed installs, my practice is to 
install Grub *ONLY* with the *FIRST* install. That way the default 
action is to boot a known good system and the questionable one will 
generally be last on the list (Grub insists on listing /dev/sda10 before 
/dev/sda2).


What I'm investigating is an apparent anomaly:
  1. My primary machine had Jessie installed first with Squeeze
 and Stretch later. It happily will boot all.
  2. I've set up a designated EXPERIMENTAL machine. Squeeze was
 installed first and thus Grub is associated with it.
 Wheezy and Jessie were installed without Grub.
 Squeeze runs fine and will browse the web.
 However, update-grub does not recognize either Wheezy or Jessie.
 Also, a netinst of Stretch failed at the network setup step -
 haven't done any trouble shooting yet.

Two questions:
  1. Can I retroactively choose "include all available drivers" and
 if so, how?
  2. How much free space would be required? As I was already planning
 on some Grub experiments, I had left 10 MB un-partitioned before
 /dev/sda1. That can easily be expanded.

TIA

Re: How stable is the frozen stretch?

[Joe]

On Sun, 14 May 2017 04:57:07 -0500
Richard Owlett  wrote:

> On 05/14/2017 02:40 AM, Joe wrote:
> > On Sat, 13 May 2017 20:54:04 -0400
> > RavenLX  wrote:
> >  
> >> On 05/13/2017 12:40 PM, Richard Owlett wrote:  
> >>> I have a partition whose label is "common".  
> >>
> >> I could almost smack myself in the head. I had done that when I
> >> used to dual-boot Windows / Linux (now you can see why I'm not a
> >> fan of dual-boot, I guess! LOL!). I also used to dual boot SolydX
> >> and SolydK distributions.
> >>
> >> In my VMs, I do use the guest additions and have a shared directory
> >> (necessity for what I need to do as well).
> >>
> >> But that was one way I had shared data.
> >>
> >> However, I still balk at dual-booting (like you balk at VMs :) )
> >>  
> >
> > Another option I've used for some years: a pocket-sized USB hard
> > drive with an i386 all-modules installation, which boots on
> > practically anything that isn't an ARM.
> >  
> 
> Can you clarify what you mean by the phrase "i386 all-modules
> installation"? TIA
> 
> 

Question during installation: 'Do you want all driver modules or just
the ones that apply to this hardware' or something similar. On every
other occasion I go for 'just the ones needed'.

-- 
Joe

Virtual Machines: Newbie / novice questions

[rhkramer]

I've never used a VM but I wanted to ask some newbie questions about them.  

Currently, I don't see much use for one--I haven't had to run Windows in a 
very long time, which is a good thing--I used to have to occasionally run 
Windows to do something that, at the time, I couldn't do in Linux--oh, I guess 
I still have to update my Garmin maps in Windows, which I do very rarely--only 
before a long road trip or when I discover some discrepancies based on new 
road construction.

I often have parts to build all or most of my next machine around, and, 
sometimes (as now) that machine is built and has the next generation of 
software on it for "testing and tuning" (as now, I have a 2nd machine running 
Jessie while my everyday "user" is running Wheezy.

I have some assumptions / guesses about the answers to these questions, and I 
could google for answers, but maybe somebody will indulge me with answers.

   * Is it right to consider that there is somehow a base OS and the other VMs 
somehow are set up on top of it--you boot into that base OS first, and then 
have the ability to switch to another VM.  OK, I'm thinking that is not the 
case--I think I've heard of things like a hypervisor (??), and now I'm 
guessing that the machine boots first into the hypervisor and then from there 
boot into any of one or more VMS, one of which would contain what I consider 
my main or base OS?

   * I know (or I'm 99.9% sure) that the software for a VM is on separate 
partitions (except maybe some are or could be shared between the VM and the 
"base" machine, such as for my documents and such (things that I'd want to be 
accessible from any of the VMs).  

   * Can I keep a VM in a running state for instant switching to it?  (I 
presume that in that case, it will be using RAM, on the other hand, it could 
be in a non-running (either non-booted or suspended-to-disk) and not be using 
RAM?

   * I presume that booting a VM takes  about the same amount of time, maybe a 
little longer than booting the base machine /  OS (well, unless the OS in the 
VM is significantly "smaller" in terms of footprint and services started at 
boot)?

   * Can I C between (applicatons in) different VMs and/or the base OS?

Thanks!

Re: How stable is the frozen stretch?

[David Wright]

On Sun 14 May 2017 at 04:57:07 (-0500), Richard Owlett wrote:
> On 05/14/2017 02:40 AM, Joe wrote:
> >On Sat, 13 May 2017 20:54:04 -0400
> >RavenLX  wrote:
> >
> >>On 05/13/2017 12:40 PM, Richard Owlett wrote:
> >>>I have a partition whose label is "common".
> >>
> >>I could almost smack myself in the head. I had done that when I used
> >>to dual-boot Windows / Linux (now you can see why I'm not a fan of
> >>dual-boot, I guess! LOL!). I also used to dual boot SolydX and SolydK
> >>distributions.
> >>
> >>In my VMs, I do use the guest additions and have a shared directory
> >>(necessity for what I need to do as well).
> >>
> >>But that was one way I had shared data.
> >>
> >>However, I still balk at dual-booting (like you balk at VMs :) )
> >>
> >
> >Another option I've used for some years: a pocket-sized USB hard drive
> >with an i386 all-modules installation, which boots on practically
> >anything that isn't an ARM.
> >
> 
> Can you clarify what you mean by the phrase "i386 all-modules installation"?

Recall this d-i screen?

  ┌─┤ [?] Install the base system 
├─┐   
  │ 
│   
  │ The primary function of an initrd is to allow the kernel to mount the root  
│   
  │ file system. It therefore needs to contain all drivers and supporting   
│   
  │ programs required to do that.   
│   
  │ 
│   
  │ A generic initrd is much larger than a targeted one and may even be so 
large│   
  │ that some boot loaders are unable to load it but has the advantage that it  
│   
  │ can be used to boot the target system on almost any hardware. With the  
│   
  │ smaller targeted initrd there is a very small chance that not all needed
│   
  │ drivers are included.   
│   
  │ 
│   
  │ Drivers to include in the initrd:   
│   
  │ 
│   
  │generic: include all available drivers   
│   
  │targeted: only include drivers needed for this system
│   
  │ 
│   
  │
│   
  │ 
│   
  
└─┘
   

Cheers,
David.

Re: Problemen met Debian Live

[Paul van der Vlis]

Op 14-05-17 om 11:54 schreef Ernst van der Storm:
> Beste...,
> 
> Ik werk nu al anderhalf jaar met Debian8 Jessie op mijn pc, kreeg hulp
> van een collega om het goed te krijgen met twee schermen.
> Ik noem een paar punten:
> 
>   * arandr lijkt makkelijk, maar ik kon er eigenlijk niets mee
>   * met xrandr krijg je een overzicht van aangesloten schermen
>   * grondig googelen levert ook cvt op, met meerdere voorbeelden
> waarvoor je dat nodig hebt (je prepareert daarmee een schermmodus
> met de juiste waardes (volgens een strikte reeks parameters) en dan
> kun je met xrandr het scherm activeren
>   * ik gebruik geen speciale drivers, alles loopt volgens mij met
> nouveau die 'uit debian komt'

Mijn ervaring met nouveau is, dat het regelmatig niet stabiel werkt.
Maar dat verschilt nogal per type videokaart. Omdat ik niet zo goed weet
welke videokaart het goed doet, gebruik ik Nvidia zo min mogelijk.

>   * *maar*:
>   * ik werk op een oude pc met agpkaarten, die zowel een vga- en een
> dvi-connector hebben
>   * in deze opzet lukte het alleen met *Nvidia*kaarten
>   * ATI-videokaart geeft wel twee schermen met beeld, maar die zijn
> hetzelfde qua inhoud, ik kreeg die niet goed werkend

Wellicht had je de firmware niet geïnstalleerd? Dan kan zo'n ATI kaart
maar heel erg weinig.

Verder heb je voor een nieuw type videokaart een nieuwe kernel en een
redelijk nieuwe xorg driver nodig.

> Zie bijv. https://wiki.archlinux.org/index.php/xrandr

Archlinux is inderdaad erg goed in de documentatie, daar kunnen we nog
wat van leren.

Zelf gebruik ik veelal het grafische tooltje wat bij de desktop omgeving
komt.

Groet,
Paul

> Groet, Ernst van der Storm
> 
> 
> Op 12-05-17 om 17:52 schreef Patrick Kox:
>> Ik gebruik maar 1 scherm dus ik heb er geen ervaring mee, maar heb je
>> daarvoor misschien de officiële ATI/NVIDIA drivers nodig?
>>
>> Oh, en kijk ook eens naar xrand ofzo (kan me de naam niet exact
>> herinneren).
>>
>>
>>
>> Cecil Westerhof  schreef op 12 mei 2017 16:37:36 CEST:
>>
>> Ik heb een nieuw systeem waar ik Debian op wil gaan installeren. Ik
>> gebruik twee monitoren, maar Debian Live gebruikt er slechts één. Wat
>> kan hier de reden van zijn? Instellingen ziet er ook maar één. En
>> teruggaan naar één monitor vind ik niet echt een optie. (Ik heb zelfs
>> liever drie, of vier.)
>>
>> Waar ik ook tegenaan loop is dat de Hardware Detection Tool niet
>> werkt: geeft ‘Failed to load libmenu.32.
>> Ook de Memory Diagnostic Tool werkt niet: als ik daarop enter geef
>> gebeurt er helemaal niets.
>>
>>
>> -- 
>> Verstuurd vanaf mijn Android apparaat met K-9 Mail. Excuseer mijn
>> beknoptheid. 
> 



-- 
Paul van der Vlis Linux systeembeheer Groningen
https://www.vandervlis.nl/

Re: Debian Live en twee monitoren

[Paul van der Vlis]

Op 13-05-17 om 18:46 schreef Geert Stappers:

> Mij ontgaat waarom het Debian Live moet zijn.
> 
> Wat ik in de thread nog _niet_ voorbij heb zien komen,
> is welke versie Debian Live gebruikt wordt.

Er is nog geen Debian live 9 volgens mij. Het moet dus wel 8 of ouder zijn.

Groeten,
Paul


-- 
Paul van der Vlis Linux systeembeheer Groningen
https://www.vandervlis.nl/

Re: Migration 32 bits vers 64 bits gros problème avec dpkg testing

[MERLIN Philippe]

Merci Étienne, 
Actuellement j'ai du arrêter mon travail de la migration 32 bits--64 bits car 
je suis en dehors de chez moi.
Je n'ai pas tenu au courant la liste des progrès dans cette migration .
J'avais réussi à passer ce problème en modifiant comme l'indique  Étienne le 
script /var/lib/dpkg/info/python3-pil:i386.prerm et tout alors à fonctionner à 
merveille, sauf  et c'est là que je me suis arrêter.
Quand j'ouvre une session avec Kdm j'obtiens un écran gris et le plus étrange 
c'est que la souris fonctionne, en examinant  attentivement on voit que le 
serveur X marche également
En regardant .Xsession-errors on voit une erreur que je retranscris de mémoire 
Kdm[1401] pam_ck_connector (kde:session) : noX11 mode 
Je me suis dit puisque Kdm me joue un tour essayons Sddm et là un autre 
problème impossible d'ouvrir une session car impossible de passer la phase 
authentification tous mes comptes sont refusés.
Remarque : J'avais déjà ce problème en 32 bits mais comme Kdm fonctionnait je 
me disais que je regarderais ce problème plus tard.
Voilà ou j'en suis de ma migration et étant actuellement  loin du poste 
incriminé, je ne peux continuer mes recherches.
Encore merci.
Philippe Merlin


Le vendredi 12 mai 2017, 23:07:25 CEST Étienne Mollier a écrit :
> On 05/06/2017 06:50 PM, MERLIN Philippe wrote:
> > Bonsoir,
> > J'avance péniblement dans cette migration et j'espérais arriver
> > à la fin
> > Las !!!
> > J'arrive à obtenir un apt-get -f install qui semble cohérent
> > sauf  que python3 a frappé et me bloque, je ne sais pas comment
> > m'en sortir.
> 
> Bonsoir Philippe,
> 
> J'ai pris un peu de temps pour sauvagement torturer une VM Debian
> afin de voir si j'arrivais à me retrouver dans la même situation.
> Plein de problèmes se sont produits, mais celui-ci je n'ai pas
> réussi à le voir passer, faute d'avoir pu réussir à atteindre la
> fin de mon second crossgrade sans tout casser.  Ceci dit, ça m'a
> permis d'avoir peut-être une idée pour décoincer la situation...
> 
> > Les scripts de python3 semblent ne pas avoir envisager la
> > possibilité d'avoir un instant donné deux versions d'un paquet
> > i386 et amd64 si bien que dpkg -L paquet sort en erreur, à la
> > main un dpkg -l paquet:i386 marche très bien.
> > 
>
> #!/bin/sh
> set -e
> 
> # Automatically added by dh_python3:
> if which py3clean >/dev/null 2>&1; then
> py3clean -p python3-pil
> else
> dpkg -L python3-pil | perl -ne
> 's,/([^/]*)\.py$,/__pycache__/\1.*, or next; unlink $_ or die $!
> foreach glob($_)'
> find /usr/lib/python3/dist-packages/ -type d -name
> __pycache__ -empty -print0 | xargs --null --no-run-if-empty rmdir
> fi
> 
> # End automatically added section
> 
> La ligne qui produit l'erreur est très probablement celle en
> `dpkg -L python3-pil | ...' qui gagnerait à être rédigée sous la
> forme `dpkg -L python3-pil:i386 | ...'.  La correction peut être
> effectuée dans le script enregistré par `dpkg' à l'installation
> de python3-pil, et qui devrait se nommer
> 
> /var/lib/dpkg/info/python3-pil:i386.prerm
> 
> ou si l'architecture n'est pas présente (mais je crains une
> confusion avec le paquet 64bits dans ce cas) :
> 
> /var/lib/dpkg/info/python3-pil.prerm
> 

> À plus,

Re: How stable is the frozen stretch?

[Richard Owlett]

On 05/14/2017 02:40 AM, Joe wrote:

On Sat, 13 May 2017 20:54:04 -0400
RavenLX  wrote:


On 05/13/2017 12:40 PM, Richard Owlett wrote:

I have a partition whose label is "common".


I could almost smack myself in the head. I had done that when I used
to dual-boot Windows / Linux (now you can see why I'm not a fan of
dual-boot, I guess! LOL!). I also used to dual boot SolydX and SolydK
distributions.

In my VMs, I do use the guest additions and have a shared directory
(necessity for what I need to do as well).

But that was one way I had shared data.

However, I still balk at dual-booting (like you balk at VMs :) )



Another option I've used for some years: a pocket-sized USB hard drive
with an i386 all-modules installation, which boots on practically
anything that isn't an ARM.



Can you clarify what you mean by the phrase "i386 all-modules installation"?
TIA

Re: Problemen met Debian Live

[Ernst van der Storm]

Beste...,

Ik werk nu al anderhalf jaar met Debian8 Jessie op mijn pc, kreeg hulp 
van een collega om het goed te krijgen met twee schermen.

Ik noem een paar punten:

 * arandr lijkt makkelijk, maar ik kon er eigenlijk niets mee
 * met xrandr krijg je een overzicht van aangesloten schermen
 * grondig googelen levert ook cvt op, met meerdere voorbeelden
   waarvoor je dat nodig hebt (je prepareert daarmee een schermmodus
   met de juiste waardes (volgens een strikte reeks parameters) en dan
   kun je met xrandr het scherm activeren
 * ik gebruik geen speciale drivers, alles loopt volgens mij met
   nouveau die 'uit debian komt'
 * *maar*:
 * ik werk op een oude pc met agpkaarten, die zowel een vga- en een
   dvi-connector hebben
 * in deze opzet lukte het alleen met *Nvidia*kaarten
 * ATI-videokaart geeft wel twee schermen met beeld, maar die zijn
   hetzelfde qua inhoud, ik kreeg die niet goed werkend

Zie bijv. https://wiki.archlinux.org/index.php/xrandr

Groet, Ernst van der Storm


Op 12-05-17 om 17:52 schreef Patrick Kox:
Ik gebruik maar 1 scherm dus ik heb er geen ervaring mee, maar heb je 
daarvoor misschien de officiële ATI/NVIDIA drivers nodig?


Oh, en kijk ook eens naar xrand ofzo (kan me de naam niet exact 
herinneren).




Cecil Westerhof  schreef op 12 mei 2017 16:37:36 CEST:

Ik heb een nieuw systeem waar ik Debian op wil gaan installeren. Ik
gebruik twee monitoren, maar Debian Live gebruikt er slechts één. Wat
kan hier de reden van zijn? Instellingen ziet er ook maar één. En
teruggaan naar één monitor vind ik niet echt een optie. (Ik heb zelfs
liever drie, of vier.)

Waar ik ook tegenaan loop is dat de Hardware Detection Tool niet
werkt: geeft ‘Failed to load libmenu.32.
Ook de Memory Diagnostic Tool werkt niet: als ik daarop enter geef
gebeurt er helemaal niets.


--
Verstuurd vanaf mijn Android apparaat met K-9 Mail. Excuseer mijn 
beknoptheid. 

Re: How to run Debian from SD card but booting from USB stick?

[Błażej Popławski]

2017-05-13 14:28 GMT+02:00 Brian :
> On Sat 13 May 2017 at 05:36:56 -0500, Richard Owlett wrote:
>
>> On 05/13/2017 04:59 AM, Brian wrote:
>> >On Fri 12 May 2017 at 09:08:50 -0500, Richard Owlett wrote:
>> >
>> >>In your case "Why install part of the OS to the SD card if you have to have
>> >>the flash drive at all times to launch the OS?"
>> >
>> >In your case "Why install part of the OS to the SD card if you have to have
>> >the hard disk drive at all times to launch the OS?"
>> >
>>
>> *ROFL* *!!!*
>> Please *READ* content before snipping.
>> I strongly suggest reading paragraph beginning "In my case I needed ..." in
>> potion you snipped ;/
>
> So the OP has no desire to test or demonstrate whether or not an SD card
> would function in his machine?
>
> --
> Brian.
>
>

There is a desire.
But I also have other things to do. I'm sorry but I cannot keep up to
date with the thread.

Re: How to run Debian from SD card but booting from USB stick?

[Błażej Popławski]

2017-05-13 20:44 GMT+02:00 Brian :
>
> On Sat 13 May 2017 at 13:07:31 +0200, Błażej Popławski wrote:
>
> > I don't have to do this for purpose. I was just wondering if it can be done
> > without much work spent on it.
> >
> > This laptop is an old Toshiba Satellite A300.
> > Main OS is Win7.
> > I have plenty of space on my hard drive. As I wrote, I'm just curious about
> > the answer.
>
> Does your curiosity extend to trying out any of the ideas expressed in
> this thread? Those who participated (and others) would find it useful to
> know how you went about booting into a Debian OS from a USB stick. The
> work involved seems very slight.
>
> --
> Brian.
>
>

As for now I've tried to do the same thing (for Ubuntu) as the author
of the wiki page which I linked at the beginning. But it didn't work
for me.
I saw that someone wrote, that the drivers for SD are loaded at the
booting time, so I assume that it should be straightforward: install
OS (Debian) on SD with /boot and bootloader on USB stick. Without any
additional work.
I've done it once but it didn't started and in the act of anger I
formatted both, SD and USB stick. Then I realized that I didn't set
the boot flag for USB. :)
Installation took long hours (I read something about that issue) and
for now I don't have the time to repeat that process.

Re: Update Notifier

[Mark Fletcher]

On Fri, May 05, 2017 at 04:32:44PM +, sare...@att.net wrote:
> Why doesn't Debian 8 Cinnamon notify when updates are ready to install 
> after all these years Debian has existed? Don't tell me there is one, 
> because after installing Debian I waited a long time to see if a 
> notification would pop up. It never did. I know about doing apt-get 
> update && apt-get upgrade,but why should we have to use a command 
> line? It makes me wonder about Debian security. Would you please put 
> an update notifier in all your versions of Debian.
> 

Who do you think you are talking to?

Re: How stable is the frozen stretch?

[Joe]

On Sat, 13 May 2017 20:54:04 -0400
RavenLX  wrote:

> On 05/13/2017 12:40 PM, Richard Owlett wrote:
> > I have a partition whose label is "common".  
> 
> I could almost smack myself in the head. I had done that when I used
> to dual-boot Windows / Linux (now you can see why I'm not a fan of 
> dual-boot, I guess! LOL!). I also used to dual boot SolydX and SolydK 
> distributions.
> 
> In my VMs, I do use the guest additions and have a shared directory 
> (necessity for what I need to do as well).
> 
> But that was one way I had shared data.
> 
> However, I still balk at dual-booting (like you balk at VMs :) )
> 

Another option I've used for some years: a pocket-sized USB hard drive
with an i386 all-modules installation, which boots on practically
anything that isn't an ARM.

-- 
Joe

Re: Serveur LAMP qui ne fonctionne plus

[Georges]

Le Sat, 13 May 2017 21:49:33 +0200,

TBP a écrit :

> Je pense que le package à installer était libapache2-mod-php5.
> Il faudrait vérifier selon les informations du package php5filter
> https://packages.debian.org/jessie/libapache2-mod-php5filter
> Là, on sort du cadre de mes connaissances, il faudrai que quelqu'un
> prenne le relai.
> Je *pense* que si la version Apache2 installée n'est pas la mdm-
> prefork, il faut choisir libapache2-mod-php5. Dans le cas contraire,
> libapache2-mod-php5filter.

 1- j'ai enlevé libapache2-mod-php5filter alors localhost/phpmyadmin me
 renvoi la page de texte et localhost/webtrees de même

 2- j'ai installé libapache2-mod-php5 alors les deux URL fonctionnent

 3- et je ne vois pas de différence avec libapache2-mod-php5filter ou
 libapache2-mod-php5

 J'ai trouvé sur Internet ceci :
Quelle est la différence entre le module php5 (libapache2-mod-php5) et
 le module de filtrage php5 (libapache2-mod-php5filter) pour apache2?

Libapache2-mod-php5filter est utilisé pour les filtres php internes
d'apache. Je suppose que vous pouvez exécuter des applications de cette
façon, mais il est loin d'être recommandé car ce n'est pas son
intention. Le mod-php5 d'autre part est un module de contenu qui génère
du contenu (à partir de votre application php) et c'est ce qu'il faut
utiliser.

 Ouf ! sa dépasse mes modeste compétence ;-)

> 
> Le samedi 13 mai 2017 à 21:36 +0200, Georges a écrit :
> > Le Sat, 13 May 2017 21:03:37 +0200,
> > 
> > TBP a écrit :
> > 
> > > Que dit la commande suivante ?
> > > dpkg -l | grep libapache2-mod-php
> > 
> >  ne répond rien et j'installe
> >  libapache2-mod-php5filter
> > 
> >  alors localhost/phpmyadmin m'affiche "enfin" la page bienvenue
> > 
> > > La commande indiquera si l'interpréteur PHP d'Apache est
> > > installé. Si oui, faites cette seconde commande
> > > grep -ir "Handler" /etc/apache2/*
> > 
> >  j'ai une réponse assez similaire avec Jessie
> > 
> > > Elle est censée retourner des lignes proches de ceci
> > > :/etc/apache2/mods-available/php7.0.conf:SetHandler
> > > application/x-
> > > httpd-php/etc/apache2/mods-available/php7.0.conf:SetHandler
> > > application/x-httpd-php-source
> > > (ces exemples sont sous Sid)
> > 
> >  Yes !! la commande : http://localhost/webtrees/setup.php sa marche
> > je
> >  vais pouvoir configurer ma Généalogie
> > 
> >  Un pour tous, tous pour un C'est la magie des Logiciels Libre.
> > 
> >  Merci, merci, merci à Thierry Bugier Andre Debian Jean michel Oltra
> > et
> >  à vous tous  Georges