Re: Best practice vsftpd writable root inside chroot()

[Bob Proulx]

Maik Stubbe wrote:
> I just came accross this new security feature of vsftpd whith which
> you cannot have your $HOME writable for a chrooted user.

If the chroot file system is writable then that has the same security
issues as a writable root filesystem.  Which is to say it has no
security.  Simply write your own /etc/passwd and other files into it
and then you have root on the system.

> Since this ftp server is a public accessable server with ~60 users
> and partial nested homedirs, there is no chance that there is no
> writable homedirectory. Another big point is, the directory
> structure increased over years. So there is even no possibility to
> change it and give users an writable directory inside their $HOME.

Sounds important.  Sounds very constrained.  If you have no degrees of
freedom then you probably can't do it. :-(

> So what is the best practice to avoid security problems?

Use sftp instead of ftp.  Use http for downloading instead of ftp.

> 1. Using no ftp -> no chance

You didn't say but are you using ftp for upload access?  Since ftp
sends passwords in the clear it is unsuitable in these days of a
hostile Internet for any purpose other than anonymous downloading.

> 2. Avoid using chroot is in my opinion a bad idea on a public
>accessable ftpd.

But you are already using ftp.  Talking about security at that point
is like putting a heavy duty lock on a screen door.  No matter how
good the lock it is still a screen door and won't stop anyone who
wants to run through it.

> 3. Compiling vsftpd >3.0 from source and using
>allow_writeable_chroot=YES: This would lead in using non Debian
>packages and watching them seperatly.

Use a Debian watch file.  See the 'uscan' program for details.  But
you can have it automatically notify you when new versions arrive.

> 4. Using packages from Jessie: My preffered choice. But how to
>control security updates?

Does the Jessie vsftpd allow writable chroots?  Sounds like a bug to
be filed to me.

> Any suggestions from your side?

Use sftp upload instead.  All operating systems support it in various ways.
Use http download instead.  All operating systems easily support it.

Bob


signature.asc
Description: Digital signature

Re: battery

[Bob Proulx]

Darac Marjal wrote:
> Perhaps so, but if the charge isn't increasing, then that's basically
> "full".

Probably so.

> Batteries usually note three main values in their "fuel gagues" (a small
> chip in the battery): current charge, current maximum charge and
> designed maximum charge. The "designed maximum charge" is written at
> manufacture time and the "current maximum charge" is updated every so
> often.
> 
> The ratio between current charge and current maximum charge give your
> your percentage full. The ratio between current maximum charge and
> designed maximum charge gives an indication of battery health (that is,
> if the battery was designed to hold 100Wh but now can only hold 20Wh,
> it's basically dead).

Install 'acpitool' and then see what it says for your battery.  Mine
says this at this time.  The information available will vary from
vendor to vendor.  Yours might have a different set of fields.

  rwp@dismay:~$ acpitool --battery
  Battery #1 : present
Remaining capacity : 65260 mWh, 100.0%
Design capacity: 71280 mWh
Last full capacity : 65260 mWh, 91.55% of design capacity
Capacity loss  : 8.446%
Present rate   : 0 mW
Charging state : Full
Battery type   : Li-ion 
Model number   : 92P1137
Serial number  : 772

You can also get this data from /proc and-or /sys but recent kernels
have changed the locations so now I find it easier to use acpitool to
dump this information.

The battery above is in good shape and the last full capacity is 91%
of the design capacity.  But a worn out battery may report something
down to 20% of capicity or worse.

It is possible that the reporting on your battery is reporting the
percentage of design capacity rather than the current capability.

> Resetting the fuel gauge varies from battery to battery, but the typical
> way is to drain the battery (so, take the charger out and run the laptop
> until it shuts down. DON'T try to over-drain the battery by starting it
> at this point), then charge the battery to full again

I think it is a good idea to do this once a month.  A battery used
well will last longer.

> (plug back in, but
> don't start the laptop - let the battery charge at full rate until it's
> at 100%. You'll need to estimate how long this usually takes). The idea
> is that the fuel gauge monitors the increase in charge and, when the
> increase tapers off, that's your maximum charge.

I think it is okay to plug it in and also start the laptop.  I have
always monitored the charging this way.  But for the reset it is
important to allow it to charge all of the way to full in one cycle.

> Don't try this too often, though. Deep discharges are also bad for
> Lithium Ion batteries. If in doubt, read the manual :)

"A ship in harbor is safe, but that is not what ships are built for."
The same is true for laptop batteries.  I use mine and that means I
will cycle it.  As long as I have gotten good use from something then
I don't feel bad if I have eventually worn it out.

Bob


signature.asc
Description: Digital signature

Re: an other question about installer

[Jerome BENOIT]

Bonjour,

On 08/08/13 23:34, François Patte wrote:
> Bonsoir,
> 
> I encounter a difficulty while partitionning.
> 
> The situation:
> 
> 4 disks:
> 
> 2 "old" HDD with the previous install (fedora 10). These 2 disks contain
> 2 raid(1) arrays, one with the / partition and one with the system
> (/usr, /var, /tmp, /tmp, /home) using LVM.
> 
> 2 SSD that I intend to configurate in the same way as the HDD: 2 raid(1)
> arrays (/boot and the rest (except /home, /var, /tmp), including /),
> using LVM.
> 
> The installer reads:
> 
> device RAID1 n°0  ext3 (old HDDs, first array)
> device RAID1 n°1 lvm (old HDDs, second array)
> device RAID1 n°2 ext4 (SSDs, first array)
> device RAID1 n°3 lvm (SSDs, second array)
> 
> SCSI1 (sda) with 2 raid devices (1st SSD)
> SCSI2 (sdb) with 2 raid devices (2nd SSD)
> SCSI5 (sdc)  with 2 raid devices (1st HDD)
> SCSI6 (sdd)  with 2 raid devices (2nd HDD)
> 
> I could configurate device RAID n°2 giving the mount point (/boot) and
> ext4 formatting.
> 
> When I want to configurate the lvm on device RAID n°3, the installer
> returns:
> 
> before the LVM could be configurated, the actual scheme of partitionment
> must be set down. *These changes will be irreversible*.
> 
> <..>
> 
> The following partition tables will be changed:
> RAID1 n°2:
>   SCSI5 (0,0,0) (sdc)
>   SCSI6 (0,0,0) (sdd)
> 
> The following partitions will be formated:
>   partition 1 on device RAID1 n°2, type ext4
> 
> 
> I am upset by this:
> 
> 1- RAID device n°2 is supposed to be the first array on SSDs
> corresponding to
> sda and sdb (and not sdc and sdd)
> 
> 2- I don't want to change anything in RAID devices n°0 and n°1 (sdc and
> sdd) because I want to keep the old system untill the new one will be
> installed on the SSD (sda and sdb) *and* I want to use the /home
> partition of these old disks as the /home of the new installation
> 
> (Everything is backed up but I prefer to understand what is going on
> here!)
> 
> 
> I hope to have been clear enough: I am not a computer scientist and
> English language is not my mother tongue... So I could have done some
> mistakes and did not use the right vocabulary.

I am not a computer scientist either, but I am somehow familiar with computers 
in labs.

Given the situation, to simplify your life:
1] make a backup of your /home on an external haddrive (an effective external 
hard drive, on an other computer, ...);
2] forget the Fedora stuff and install Debian from _scratch_ ;
3] after installation, backup back you /home folder (and check that the former 
privilege policy of your /home is
compatible with the Debian policy).

I can not help on what is really going on, but by experience, to put it simply, 
I know that Fedora is ``messy'' (compare to Debian):
if you keep your Fedora material, you may encounter minor but annoying 
difference (as different version, different default options, ...)
that may pollute you life not only during the installation but also afterwards.

Installing from scratch may seem a waste of time, but it is not in fact because 
you de facto discard a long list
of potential highly annoying and subtil issues. On the list, you will even find 
some people that encourage
installation from scratch for Debian release upgrading. Anyway, the waste of 
time will be due mainly to backups:
during backups, take a break, drink cups of tea, or read Debian manuals.

Bon courage,
Jerome

> 
> 
> Thank you for attention.
> 
> 
> 


-- 
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Archive: http://lists.debian.org/52047ac1.6020...@rezozer.net

Re: replacement for ide-generic in Wheezy?

[Jerome BENOIT]

Hello,

On 09/08/13 02:31, Gary Dale wrote:
> On 08/08/13 12:20 PM, Darac Marjal wrote:
>> On Thu, Aug 08, 2013 at 12:06:23PM -0400, Gary Dale wrote:
>>> On 08/08/13 06:28 AM, Jerome BENOIT wrote:
 Hello Gary,

 On 08/08/13 11:39, Gary Dale wrote:
> The various ide modules (ide-generic, ide-cdrom, etc.) seem to be missing 
> in the Wheezy kernel. However they still compile a 486 kernel so I'm 
> assuming there is some intent to continue to support older hardware.
>
> In my case, I have an old laptop that I could run Squeeze on by including 
> ide-generic in the /etc/initramfstool/modules file. However the same 
> thing doesn't work in Wheezy because there is no ide-generic module.
>
> Googling around, I found one post that suggested it's been superseded by 
> libata but I can't find that either.
>
> Is 486-era hardware still supported in the newer kernels or is this a 
> lost cause?
>

 I guess that it is still supported by the kernel, but not by the kernel 
 deb package as distributed by debian.
 You may build your own kernel, and the following link is a good place to 
 start:

 http://kernel-handbook.alioth.debian.org

 hth,
 Jerome
>>>
>>> Thanks. This would be a truly bizarre situation - creating a 486
>>> kernel package for people but omitting the basic hardware drivers
>>> for hardware of the that vintage.
>>>
>>> The basic difference between the 486 and 686 kernels is support for
>>> PAE. Pentium processors lack PAE but would generally need the IDE
>>> drivers.
>>>
>>> In my case, I'm talking about Pentium-MMX @ 200MHz, which is far
>>> less powerful than a Raspberry Pi but still useful. However the
>>> CD-ROM and disk drives of that time need IDE drivers.
>>>
>>> It wasn't long after that PAE was added (in the Pentium Pro and
>>> later). Why would anyone make a 486 kernel and not include the
>>> drivers? I could understand, but not agree with, dropping 486
>>> support as some distros have done. But to simply not compile the
>>> drivers seems silly.
>>
>> Newer kernels, as you've mentioned, use libata to interface with
>> Parallel ATA devices (which is what most 486-era drives are). AIUI, this
>> is a new set of code that works both with SATA and PATA drives. Common
>> code means more people use it, less bit-rot etc. If does, however, come
>> with the side effect that /dev/hd* is now /dev/sd* (even if it's not a
>> SCSI disk. I suspect the initial S is now meaningless).
>>
>> If you need to access a drive that's pre-IDE (so MFM, RLL etc), then
>> yes, you'll need the old code.
>>
>> However, the chances are very likely that CONFIG_ATA_GENERIC (aka
>> ata_generic.ko) is the successor you want.
>>
> 
> I found the ata_generic module in Squeeze, Wheezy and Jessie. However, it 
> doesn't seem to help. While I'd expect that it would be in the initramfs, I 
> included it in /etc/initramfs-tools/modules anyway but I still can't boot 
> with the Wheezy kernel.
> 
> It's presence in Squeeze probably stems from an ongoing attempt to move all 
> IDE code into one module. I'm wondering if someone got a little overzealous 
> in deciding that they'd covered their bases when they removed ide-generic.
> 
> 

Just build your own kernel.

Jerome


-- 
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Archive: http://lists.debian.org/52046e89.2010...@rezozer.net

Re: battery

[Robert Holtzman]

On Thu, Aug 08, 2013 at 11:15:03AM -0700, cletusjenkins wrote:

snip...
> 
> The cycle only does damage when the battery is discharged beyond 50-60%, 
> which a li-on battery should not allow.

Laptop batteries can and sometimes do go to total discharge if left on
battery power and forgotten. I have done that more than once on Dell and
Lenovo laptops. If you never let the charge get below 50-60% and never
above 80% or so your charge life would be what? An hour or two? Makes
little sense. Do you have a citation for your 50-60% figure?

On the subject of overcharging, most recent chargers will stop charging
once the battery is at 100%. This, however, qualifies as overcharging to
quite a few people. 

-- 
Bob Holtzman
Your mail is being read by tight lipped 
Homeland Security agents who fail to see
the humor in Doctor Strangelove 


signature.asc
Description: Digital signature

Re: dhclient "No DHCPOFFERS received"

[Jerry Stuckle]

On 8/7/2013 4:53 PM, Sean Alexandre wrote:

On Wed, Aug 07, 2013 at 09:29:57PM +0200, Slavko wrote:

is your network like this, please:

-
|ISP|
-
  |
  |
-
|   modem   |
-
  |
--
| |  |
 -  -  -
 |  Debian I |  | Debian II |  |  TP-Link  |
 -  -  -


or like this?

-
|ISP|
-
  |
  |
-
|   modem   |
-
  |
  |
-
|  TP-Link  |
-
  |
 --
 ||
  - -
  |  Debian I | | Debian II |
  - -


My network is like your first diagram, but with only one machine connected to
the modem at a time.




Getting a public address from a private DHCP server is not wrong.  The 
DHCP server may have been given those addresses by the upstream server 
for allocation.  That takes a load off the upstream server.


One thought.  Just because there is only one machine connected at a time 
does NOT mean the previous DHCP lease has been released.  It may still 
be considered active by the DHCP server.


For instance - if you hook up one machine and get a least, that lease 
may be good for an hour to a week (or more).  If you hook a second one 
up, that may also get a lease for the same amount of time.


Then if you disconnect the first two (or even disconnect the first 
machine before connecting the second machine) and only have two leases 
available, you will see this problem.


The cable company was not necessarily incorrect when telling you to 
power off the cable modem then powering it back on.  This will probably 
reset all leases.


What happens if you do power the cable modem off like they said then 
power it back on, followed by connecting only the failing machine?


And BTW - you said all of the computers have different MAC addresses.  I 
would hope so!  Hardware MAC addresses for a port are unique in the 
entire world - every one HAS to be different.  Of course some OS's allow 
you to override the hardware MAC address, but that's another story.



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

Archive: http://lists.debian.org/52043a5b.5060...@attglobal.net

Re: replacement for ide-generic in Wheezy?

[Gary Dale]

On 08/08/13 12:20 PM, Darac Marjal wrote:

On Thu, Aug 08, 2013 at 12:06:23PM -0400, Gary Dale wrote:

On 08/08/13 06:28 AM, Jerome BENOIT wrote:

Hello Gary,

On 08/08/13 11:39, Gary Dale wrote:

The various ide modules (ide-generic, ide-cdrom, etc.) seem to be missing in 
the Wheezy kernel. However they still compile a 486 kernel so I'm assuming 
there is some intent to continue to support older hardware.

In my case, I have an old laptop that I could run Squeeze on by including 
ide-generic in the /etc/initramfstool/modules file. However the same thing 
doesn't work in Wheezy because there is no ide-generic module.

Googling around, I found one post that suggested it's been superseded by libata 
but I can't find that either.

Is 486-era hardware still supported in the newer kernels or is this a lost 
cause?



I guess that it is still supported by the kernel, but not by the kernel deb 
package as distributed by debian.
You may build your own kernel, and the following link is a good place to start:

http://kernel-handbook.alioth.debian.org

hth,
Jerome


Thanks. This would be a truly bizarre situation - creating a 486
kernel package for people but omitting the basic hardware drivers
for hardware of the that vintage.

The basic difference between the 486 and 686 kernels is support for
PAE. Pentium processors lack PAE but would generally need the IDE
drivers.

In my case, I'm talking about Pentium-MMX @ 200MHz, which is far
less powerful than a Raspberry Pi but still useful. However the
CD-ROM and disk drives of that time need IDE drivers.

It wasn't long after that PAE was added (in the Pentium Pro and
later). Why would anyone make a 486 kernel and not include the
drivers? I could understand, but not agree with, dropping 486
support as some distros have done. But to simply not compile the
drivers seems silly.


Newer kernels, as you've mentioned, use libata to interface with
Parallel ATA devices (which is what most 486-era drives are). AIUI, this
is a new set of code that works both with SATA and PATA drives. Common
code means more people use it, less bit-rot etc. If does, however, come
with the side effect that /dev/hd* is now /dev/sd* (even if it's not a
SCSI disk. I suspect the initial S is now meaningless).

If you need to access a drive that's pre-IDE (so MFM, RLL etc), then
yes, you'll need the old code.

However, the chances are very likely that CONFIG_ATA_GENERIC (aka
ata_generic.ko) is the successor you want.



I found the ata_generic module in Squeeze, Wheezy and Jessie. However, 
it doesn't seem to help. While I'd expect that it would be in the 
initramfs, I included it in /etc/initramfs-tools/modules anyway but I 
still can't boot with the Wheezy kernel.


It's presence in Squeeze probably stems from an ongoing attempt to move 
all IDE code into one module. I'm wondering if someone got a little 
overzealous in deciding that they'd covered their bases when they 
removed ide-generic.



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

Archive: http://lists.debian.org/52043856.6020...@velcom.ca

an other question about installer

[François Patte]

-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1

Bonsoir,

I encounter a difficulty while partitionning.

The situation:

4 disks:

2 "old" HDD with the previous install (fedora 10). These 2 disks contain
2 raid(1) arrays, one with the / partition and one with the system
(/usr, /var, /tmp, /tmp, /home) using LVM.

2 SSD that I intend to configurate in the same way as the HDD: 2 raid(1)
arrays (/boot and the rest (except /home, /var, /tmp), including /),
using LVM.

The installer reads:

device RAID1 n°0  ext3 (old HDDs, first array)
device RAID1 n°1 lvm (old HDDs, second array)
device RAID1 n°2 ext4 (SSDs, first array)
device RAID1 n°3 lvm (SSDs, second array)

SCSI1 (sda) with 2 raid devices (1st SSD)
SCSI2 (sdb) with 2 raid devices (2nd SSD)
SCSI5 (sdc)  with 2 raid devices (1st HDD)
SCSI6 (sdd)  with 2 raid devices (2nd HDD)

I could configurate device RAID n°2 giving the mount point (/boot) and
ext4 formatting.

When I want to configurate the lvm on device RAID n°3, the installer
returns:

before the LVM could be configurated, the actual scheme of partitionment
must be set down. *These changes will be irreversible*.

<..>

The following partition tables will be changed:
RAID1 n°2:
  SCSI5 (0,0,0) (sdc)
  SCSI6 (0,0,0) (sdd)

The following partitions will be formated:
  partition 1 on device RAID1 n°2, type ext4


I am upset by this:

1- RAID device n°2 is supposed to be the first array on SSDs
corresponding to
sda and sdb (and not sdc and sdd)

2- I don't want to change anything in RAID devices n°0 and n°1 (sdc and
sdd) because I want to keep the old system untill the new one will be
installed on the SSD (sda and sdb) *and* I want to use the /home
partition of these old disks as the /home of the new installation

(Everything is backed up but I prefer to understand what is going on
here!)


I hope to have been clear enough: I am not a computer scientist and
English language is not my mother tongue... So I could have done some
mistakes and did not use the right vocabulary.


Thank you for attention.

- -- 
François Patte
UFR de mathématiques et informatique
Laboratoire MAP5 --- UMR CNRS 8145
Université Paris Descartes
45, rue des Saints Pères
F-75270 Paris Cedex 06
Tél. +33 (0)1 8394 5849
http://www.math-info.univ-paris5.fr/~patte
-BEGIN PGP SIGNATURE-
Version: GnuPG v1.4.11 (GNU/Linux)
Comment: Using GnuPG with Fedora - http://enigmail.mozdev.org/

iEYEARECAAYFAlIEDv4ACgkQdE6C2dhV2JV4VgCdEsxNuVvqNHhzQZ/9l6ax1CMj
L4oAni4C8jXKN4f4HPcySv1LVF3DuOOE
=xuv8
-END PGP SIGNATURE-


-- 
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Archive: http://lists.debian.org/52040efe.3060...@mi.parisdescartes.fr

Re: battery

[David Guntner]

cletusjenkins grabbed a keyboard and wrote:
> 
> Thank god I never bought a samsung laptop. I am amazed that samsung
> has just recently discovered the amazing feature of not over-charging
> batteries.

Funny... my laptop is a Samsung that I've had for a few years now, and
it has a feature under battery management that lets me select if I want
it to go to full 100% charge or a more battery-saving 80%

Maybe I just got lucky. :-)

   --Dave




smime.p7s
Description: S/MIME Cryptographic Signature

Re: debian installation question

[Jerome BENOIT]

Bonsoir,

On 08/08/13 20:15, François Patte wrote:
> Bonjour,
> 
> I would like to make a fresh install of debian and I want to keep my
> previous /home partition ie. forbid the installer to reformat this
> partition.
> 
> I am sure that it is possible,

Indeed, it is possible
Nevertheless, for safety, you may make a backup of your home on an external 
drive
(something as ``cp -prd /home '' can do the work) 

 but at the end of the partitionning step,
> how cna I check which partition will be formated and which won't.

If I remember well, the installer ask for either a generic partionning or a by 
hand one:
you want to choose the by hand option.



> 
> Thank you.
> 
> 
>

hth,
Jerome
 


-- 
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Archive: http://lists.debian.org/5203e531.8030...@rezozer.net

Re: debian installation question

[cletusjenkins]

 On Thu, 08 Aug 2013 11:15:44 -0700 François 
Patte wrote  

 > -BEGIN PGP SIGNED MESSAGE- 
 > Hash: SHA1 
 >  
 > Bonjour, 
 >  
 > I would like to make a fresh install of debian and I want to keep my 
 > previous /home partition ie. forbid the installer to reformat this 
 > partition. 
 >  
 > I am sure that it is possible, but at the end of the partitionning step, 
 > how cna I check which partition will be formated and which won't. 
 >  
 > Thank you. 
 >  
 > - -- 
 > François Patte 
 > UFR de mathématiques et informatique 
 > Laboratoire CNRS MAP5, UMR 8145 
 > Université Paris Descartes 
 > 45, rue des Saints Pères 
 > F-75270 Paris Cedex 06 
 > Tél. +33 (0)1 8394 5849 
 > http://www.math-info.univ-paris5.fr/~patte 
 > -BEGIN PGP SIGNATURE- 
 > Version: GnuPG v1.4.9 (GNU/Linux) 
 > Comment: Using GnuPG with Fedora - http://enigmail.mozdev.org/ 
 >  
 > iEYEARECAAYFAlID4E8ACgkQdE6C2dhV2JVjvACePj1GZSE7elD5I7wgD6ehC2xL 
 > Hn4AoKd/q0kxBG3zui7CzeVfHBrP6Wnk 
 > =tZzb 
 > -END PGP SIGNATURE- 

If you go through manual partitioning (make sure your old home partiion is set 
to not being used, click on "Done" or "finished" (sorry I forget the exact 
verbage, but it is the last menu item in the manual partition screen) , before 
it  changes anything it will list exactly which partitions it will touch, then 
you have the option to continue or not.


--
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Archive: 
http://lists.debian.org/1904161724.1571.1375986836717.JavaMail.sas@[172.29.251.236]

Re: battery

[cletusjenkins]

 >  
 > http://www.samsung.com/us/article/tips--tricks-extending-notebook-battery-life
 >  
 >  
 > it's called "Smart charging", so I guess it should have a config option 
 > somewhere. 
 >  
 > Philip 

Thank god I never bought a samsung laptop. I am amazed that samsung has just 
recently discovered the amazing feature of not over-charging batteries.

Are there any commands that manipulate battery/acpi info? I'm not sure what to 
even search for. From what does the battery indicator get its info?


-- 
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Archive: 
http://lists.debian.org/301141475.17367.1375986485544.JavaMail.sas2@[172.29.244.248]

debian installation question

[François Patte]

-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1

Bonjour,

I would like to make a fresh install of debian and I want to keep my
previous /home partition ie. forbid the installer to reformat this
partition.

I am sure that it is possible, but at the end of the partitionning step,
how cna I check which partition will be formated and which won't.

Thank you.

- --
François Patte
UFR de mathématiques et informatique
Laboratoire CNRS MAP5, UMR 8145
Université Paris Descartes
45, rue des Saints Pères
F-75270 Paris Cedex 06
Tél. +33 (0)1 8394 5849
http://www.math-info.univ-paris5.fr/~patte
-BEGIN PGP SIGNATURE-
Version: GnuPG v1.4.9 (GNU/Linux)
Comment: Using GnuPG with Fedora - http://enigmail.mozdev.org/

iEYEARECAAYFAlID4E8ACgkQdE6C2dhV2JVjvACePj1GZSE7elD5I7wgD6ehC2xL
Hn4AoKd/q0kxBG3zui7CzeVfHBrP6Wnk
=tZzb
-END PGP SIGNATURE-


-- 
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Archive: http://lists.debian.org/5203e050.2080...@mi.parisdescartes.fr

Re: battery

[cletusjenkins]

>> 
>It's a feature of modern laptops. 
> 
>The idea is that the battery lasts longer if you avoid full 
>charge/discharge cycles. 
> 
>Mine is on 80% too - it's normal. 
> 
>As for how you tell it to charge to 100% because you're planning a bus 
>trip - anyone out there, please let me know. 
> 
>Regards, 
>Philip Ashmore 
> 

The cycle only does damage when the battery is discharged beyond 50-60%, which 
a li-on battery should not allow. It is not the charge that does the damage it 
is the discharge, so either this diagnosis is not correct or battery makers are 
morons.


-- 
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Archive: 
http://lists.debian.org/284161387.16808.1375986004867.JavaMail.sas2@[172.29.244.248]

Re: replacement for ide-generic in Wheezy?

[Darac Marjal]

On Thu, Aug 08, 2013 at 12:06:23PM -0400, Gary Dale wrote:
> On 08/08/13 06:28 AM, Jerome BENOIT wrote:
> >Hello Gary,
> >
> >On 08/08/13 11:39, Gary Dale wrote:
> >>The various ide modules (ide-generic, ide-cdrom, etc.) seem to be missing 
> >>in the Wheezy kernel. However they still compile a 486 kernel so I'm 
> >>assuming there is some intent to continue to support older hardware.
> >>
> >>In my case, I have an old laptop that I could run Squeeze on by including 
> >>ide-generic in the /etc/initramfstool/modules file. However the same thing 
> >>doesn't work in Wheezy because there is no ide-generic module.
> >>
> >>Googling around, I found one post that suggested it's been superseded by 
> >>libata but I can't find that either.
> >>
> >>Is 486-era hardware still supported in the newer kernels or is this a lost 
> >>cause?
> >>
> >
> >I guess that it is still supported by the kernel, but not by the kernel deb 
> >package as distributed by debian.
> >You may build your own kernel, and the following link is a good place to 
> >start:
> >
> >http://kernel-handbook.alioth.debian.org
> >
> >hth,
> >Jerome
> 
> Thanks. This would be a truly bizarre situation - creating a 486
> kernel package for people but omitting the basic hardware drivers
> for hardware of the that vintage.
> 
> The basic difference between the 486 and 686 kernels is support for
> PAE. Pentium processors lack PAE but would generally need the IDE
> drivers.
> 
> In my case, I'm talking about Pentium-MMX @ 200MHz, which is far
> less powerful than a Raspberry Pi but still useful. However the
> CD-ROM and disk drives of that time need IDE drivers.
> 
> It wasn't long after that PAE was added (in the Pentium Pro and
> later). Why would anyone make a 486 kernel and not include the
> drivers? I could understand, but not agree with, dropping 486
> support as some distros have done. But to simply not compile the
> drivers seems silly.

Newer kernels, as you've mentioned, use libata to interface with
Parallel ATA devices (which is what most 486-era drives are). AIUI, this
is a new set of code that works both with SATA and PATA drives. Common
code means more people use it, less bit-rot etc. If does, however, come
with the side effect that /dev/hd* is now /dev/sd* (even if it's not a
SCSI disk. I suspect the initial S is now meaningless).

If you need to access a drive that's pre-IDE (so MFM, RLL etc), then
yes, you'll need the old code.

However, the chances are very likely that CONFIG_ATA_GENERIC (aka
ata_generic.ko) is the successor you want.



signature.asc
Description: Digital signature

Re: replacement for ide-generic in Wheezy?

[Jerome BENOIT]

Hi,

On 08/08/13 18:06, Gary Dale wrote:
> On 08/08/13 06:28 AM, Jerome BENOIT wrote:
>> Hello Gary,
>>
>> On 08/08/13 11:39, Gary Dale wrote:
>>> The various ide modules (ide-generic, ide-cdrom, etc.) seem to be missing 
>>> in the Wheezy kernel. However they still compile a 486 kernel so I'm 
>>> assuming there is some intent to continue to support older hardware.
>>>
>>> In my case, I have an old laptop that I could run Squeeze on by including 
>>> ide-generic in the /etc/initramfstool/modules file. However the same thing 
>>> doesn't work in Wheezy because there is no ide-generic module.
>>>
>>> Googling around, I found one post that suggested it's been superseded by 
>>> libata but I can't find that either.
>>>
>>> Is 486-era hardware still supported in the newer kernels or is this a lost 
>>> cause?
>>>
>>
>> I guess that it is still supported by the kernel, but not by the kernel deb 
>> package as distributed by debian.
>> You may build your own kernel, and the following link is a good place to 
>> start:
>>
>> http://kernel-handbook.alioth.debian.org
>>
>> hth,
>> Jerome
> 
> Thanks. This would be a truly bizarre situation - creating a 486 kernel 
> package for people but omitting the basic hardware drivers for hardware of 
> the that vintage.
> 
> The basic difference between the 486 and 686 kernels is support for PAE. 
> Pentium processors lack PAE but would generally need the IDE drivers.
> 
> In my case, I'm talking about Pentium-MMX @ 200MHz, which is far less 
> powerful than a Raspberry Pi but still useful. However the CD-ROM and disk 
> drives of that time need IDE drivers.
> 
> It wasn't long after that PAE was added (in the Pentium Pro and later). Why 
> would anyone make a 486 kernel and not include the drivers? I could 
> understand, but not agree with, dropping 486 support as some distros have 
> done. But to simply not compile the drivers seems silly.
> 
> 
It is a very good idea to build specific kernels:
building a kernel deb package is easy within Debian. 

For the choice of driver of the distro default, you may ask to the package 
maintainer.

Jerome


-- 
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Archive: http://lists.debian.org/5203c460.4050...@rezozer.net

Re: replacement for ide-generic in Wheezy?

[Gary Dale]

On 08/08/13 06:28 AM, Jerome BENOIT wrote:

Hello Gary,

On 08/08/13 11:39, Gary Dale wrote:

The various ide modules (ide-generic, ide-cdrom, etc.) seem to be missing in 
the Wheezy kernel. However they still compile a 486 kernel so I'm assuming 
there is some intent to continue to support older hardware.

In my case, I have an old laptop that I could run Squeeze on by including 
ide-generic in the /etc/initramfstool/modules file. However the same thing 
doesn't work in Wheezy because there is no ide-generic module.

Googling around, I found one post that suggested it's been superseded by libata 
but I can't find that either.

Is 486-era hardware still supported in the newer kernels or is this a lost 
cause?



I guess that it is still supported by the kernel, but not by the kernel deb 
package as distributed by debian.
You may build your own kernel, and the following link is a good place to start:

http://kernel-handbook.alioth.debian.org

hth,
Jerome


Thanks. This would be a truly bizarre situation - creating a 486 kernel 
package for people but omitting the basic hardware drivers for hardware 
of the that vintage.


The basic difference between the 486 and 686 kernels is support for PAE. 
Pentium processors lack PAE but would generally need the IDE drivers.


In my case, I'm talking about Pentium-MMX @ 200MHz, which is far less 
powerful than a Raspberry Pi but still useful. However the CD-ROM and 
disk drives of that time need IDE drivers.


It wasn't long after that PAE was added (in the Pentium Pro and later). 
Why would anyone make a 486 kernel and not include the drivers? I could 
understand, but not agree with, dropping 486 support as some distros 
have done. But to simply not compile the drivers seems silly.



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

Archive: http://lists.debian.org/5203c1ff.3080...@velcom.ca

Re: roundcube-plugins

[Brad Alexander]

Okay, so any idea why dpkg is doing the half-installed/unpacked/configured
issue every 30 minutes?


On Thu, Aug 8, 2013 at 9:50 AM, Darac Marjal wrote:

> On Thu, Aug 08, 2013 at 09:36:43AM -0400, Brad Alexander wrote:
> >That is the usual behavior, but in my case, installing one uninstalls
> the
> >other. According to [1]Bug 714135, this was marked as fixed, but I
> still
> >get the install-uninstall behavior.
>
> Ah, yes. Roundcube-plugins 0.9.2-2 breaks *and replaces*
> roundcube-plugins-extra <= 0.7-20120110. This means the two are
> incompatible. So you don't get a file-level conflict any more, you get
> a package-level conflict instead.
>
> That is, r-p 0.9.2-1 and r-p-e 0.7-20120110 where co-installable, but
> you'd get a "Trying to overwrite "..." which is also in package "...""
> error. Now, you CAN'T install both at the same time.
>
> This is fine for the main roundcube packages: installing r-p  will
> remove the old r-p-e package beforehand. But if you want any of the
> OTHER plugins from r-p-e (personally I want the fail2ban plugin and
> don't much care about the zipdownload plugin), then you need to wait for
> an update r-p-e package available.
>
> I don't know if there IS going to be an updated roundcube-plugins-extra,
> but for now, I'm holding back and 0.9.2-1 (having forced the overwrite).
> But, hey, that's Sid for you :)
>
> >
> >And I am still getting it trying to install and uninstall whichever
> one is
> >installed.
> >
> >On Thu, Aug 8, 2013 at 9:21 AM, Darac Marjal <[2]
> mailingl...@darac.org.uk>
> >wrote:
> >
> >  On Thu, Aug 08, 2013 at 09:16:46AM -0400, Brad Alexander wrote:
> >  >Is anyone running roundcube? I upgraded this past weekend
> (08/04),
> >  and
> >  >ever since then, I have been getting these messages to
> >  /var/log/dpkg.log
> >  >every 30 minutes:
> >
> >  Both roundcube-plugins (0.9) and roundcube-plugins-extra (0.7)
> provide
> >  the zipdownload plugin. If you try to install both, you'll get a
> file
> >  conflict.
> >
> >  You can either wait for roundcube-plugins-extra 0.9 to arrive or
> you can
> >  force the overwrite of roundcube-plugins.
> >  >
> >  >I can't find why it is happening with roundcube-plugins. It
> showed
> >  similar
> >  >with roundcube-plugins-extra, but in spite of the bug report
> saying
> >  that
> >  >both could be installed together, it uninstalled -extra when I
> >  installed
> >  >-plugins.
> >  >
> >  >Any ideas?
> >  >
> >  >thanks,
> >  >--b
> >
> > References
> >
> >Visible links
> >1. http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=714135
> >2. mailto:mailingl...@darac.org.uk
>

Re: roundcube-plugins

[Darac Marjal]

On Thu, Aug 08, 2013 at 09:36:43AM -0400, Brad Alexander wrote:
>That is the usual behavior, but in my case, installing one uninstalls the
>other. According to [1]Bug 714135, this was marked as fixed, but I still
>get the install-uninstall behavior.

Ah, yes. Roundcube-plugins 0.9.2-2 breaks *and replaces*
roundcube-plugins-extra <= 0.7-20120110. This means the two are
incompatible. So you don't get a file-level conflict any more, you get
a package-level conflict instead.

That is, r-p 0.9.2-1 and r-p-e 0.7-20120110 where co-installable, but
you'd get a "Trying to overwrite "..." which is also in package "...""
error. Now, you CAN'T install both at the same time.

This is fine for the main roundcube packages: installing r-p  will
remove the old r-p-e package beforehand. But if you want any of the
OTHER plugins from r-p-e (personally I want the fail2ban plugin and
don't much care about the zipdownload plugin), then you need to wait for
an update r-p-e package available.

I don't know if there IS going to be an updated roundcube-plugins-extra,
but for now, I'm holding back and 0.9.2-1 (having forced the overwrite).
But, hey, that's Sid for you :)

> 
>And I am still getting it trying to install and uninstall whichever one is
>installed.
> 
>On Thu, Aug 8, 2013 at 9:21 AM, Darac Marjal <[2]mailingl...@darac.org.uk>
>wrote:
> 
>  On Thu, Aug 08, 2013 at 09:16:46AM -0400, Brad Alexander wrote:
>  >    Is anyone running roundcube? I upgraded this past weekend (08/04),
>  and
>  >    ever since then, I have been getting these messages to
>  /var/log/dpkg.log
>  >    every 30 minutes:
> 
>  Both roundcube-plugins (0.9) and roundcube-plugins-extra (0.7) provide
>  the zipdownload plugin. If you try to install both, you'll get a file
>  conflict.
> 
>  You can either wait for roundcube-plugins-extra 0.9 to arrive or you can
>  force the overwrite of roundcube-plugins.
>  >
>  >    I can't find why it is happening with roundcube-plugins. It showed
>  similar
>  >    with roundcube-plugins-extra, but in spite of the bug report saying
>  that
>  >    both could be installed together, it uninstalled -extra when I
>  installed
>  >    -plugins.
>  >
>  >    Any ideas?
>  >
>  >    thanks,
>  >    --b
> 
> References
> 
>Visible links
>1. http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=714135
>2. mailto:mailingl...@darac.org.uk


signature.asc
Description: Digital signature

Re: roundcube-plugins

[Brad Alexander]

That is the usual behavior, but in my case, installing one uninstalls the
other. According to Bug
714135,
this was marked as fixed, but I still get the install-uninstall behavior.

And I am still getting it trying to install and uninstall whichever one is
installed.


On Thu, Aug 8, 2013 at 9:21 AM, Darac Marjal wrote:

> On Thu, Aug 08, 2013 at 09:16:46AM -0400, Brad Alexander wrote:
> >Is anyone running roundcube? I upgraded this past weekend (08/04), and
> >ever since then, I have been getting these messages to
> /var/log/dpkg.log
> >every 30 minutes:
>
> Both roundcube-plugins (0.9) and roundcube-plugins-extra (0.7) provide
> the zipdownload plugin. If you try to install both, you'll get a file
> conflict.
>
> You can either wait for roundcube-plugins-extra 0.9 to arrive or you can
> force the overwrite of roundcube-plugins.
>
> >
> >I can't find why it is happening with roundcube-plugins. It showed
> similar
> >with roundcube-plugins-extra, but in spite of the bug report saying
> that
> >both could be installed together, it uninstalled -extra when I
> installed
> >-plugins.
> >
> >Any ideas?
> >
> >thanks,
> >--b
>

Re: roundcube-plugins

[Darac Marjal]

On Thu, Aug 08, 2013 at 09:16:46AM -0400, Brad Alexander wrote:
>Is anyone running roundcube? I upgraded this past weekend (08/04), and
>ever since then, I have been getting these messages to /var/log/dpkg.log
>every 30 minutes:

Both roundcube-plugins (0.9) and roundcube-plugins-extra (0.7) provide
the zipdownload plugin. If you try to install both, you'll get a file
conflict.

You can either wait for roundcube-plugins-extra 0.9 to arrive or you can
force the overwrite of roundcube-plugins.

> 
>I can't find why it is happening with roundcube-plugins. It showed similar
>with roundcube-plugins-extra, but in spite of the bug report saying that
>both could be installed together, it uninstalled -extra when I installed
>-plugins.
> 
>Any ideas?
> 
>thanks,
>--b


signature.asc
Description: Digital signature

roundcube-plugins

[Brad Alexander]

Is anyone running roundcube? I upgraded this past weekend (08/04), and ever
since then, I have been getting these messages to /var/log/dpkg.log every
30 minutes:

2013-08-08 07:32:33 startup packages remove
2013-08-08 07:32:33 status installed roundcube-plugins-extra:all
0.7-20120110
2013-08-08 07:32:33 remove roundcube-plugins-extra:all 0.7-20120110 
2013-08-08 07:32:33 status half-configured roundcube-plugins-extra:all
0.7-20120110
2013-08-08 07:32:33 status half-installed roundcube-plugins-extra:all
0.7-20120110
2013-08-08 07:32:33 status config-files roundcube-plugins-extra:all
0.7-20120110
2013-08-08 07:32:33 status config-files roundcube-plugins-extra:all
0.7-20120110
2013-08-08 07:32:34 startup archives unpack
2013-08-08 07:32:35 install roundcube-plugins:all 0.9.2-2 0.9.2-2
2013-08-08 07:32:35 status half-installed roundcube-plugins:all 0.9.2-2
2013-08-08 07:32:35 status unpacked roundcube-plugins:all 0.9.2-2
2013-08-08 07:32:35 status unpacked roundcube-plugins:all 0.9.2-2
2013-08-08 07:32:37 startup packages configure
2013-08-08 07:32:37 configure roundcube-plugins:all 0.9.2-2 
2013-08-08 07:32:37 status unpacked roundcube-plugins:all 0.9.2-2
2013-08-08 07:32:37 status unpacked roundcube-plugins:all 0.9.2-2
2013-08-08 07:32:37 status unpacked roundcube-plugins:all 0.9.2-2
2013-08-08 07:32:37 status unpacked roundcube-plugins:all 0.9.2-2
2013-08-08 07:32:37 status unpacked roundcube-plugins:all 0.9.2-2
2013-08-08 07:32:37 status unpacked roundcube-plugins:all 0.9.2-2
2013-08-08 07:32:37 status unpacked roundcube-plugins:all 0.9.2-2
2013-08-08 07:32:37 status unpacked roundcube-plugins:all 0.9.2-2
2013-08-08 07:32:37 status unpacked roundcube-plugins:all 0.9.2-2
2013-08-08 07:32:37 status unpacked roundcube-plugins:all 0.9.2-2
2013-08-08 07:32:38 status unpacked roundcube-plugins:all 0.9.2-2
2013-08-08 07:32:38 status half-configured roundcube-plugins:all 0.9.2-2
2013-08-08 07:32:38 status installed roundcube-plugins:all 0.9.2-2

I can't find why it is happening with roundcube-plugins. It showed similar
with roundcube-plugins-extra, but in spite of the bug report saying that
both could be installed together, it uninstalled -extra when I installed
-plugins.

Any ideas?

thanks,
--b

Re: battery

[Philip Ashmore]

On 08/08/13 12:46, Peter Hillier-Brook wrote:
> On 08/08/13 09:29, Philip Ashmore wrote:
>> On 08/08/13 04:05, cletusjenkins wrote:
>>> I'm using wheezy and in gnome 3, if I disconnect my laptop from its
>>> charger and then at some point suspend. After bringing the system
>>> back up and connect the charger the battery charge indicator in
>>> gnome shows incorrect info.
>>>
>>> I plug in the charger and the battery charges, but it never reaches
>>> 100%, it stops at around usually in the 80's or 90's. The indicator
>>> shows the battery is charging no matter how long it is plugged in,
>>> but once the percentage stops increasing it will never go higher.
>>> This inaccuracy usually persists across reboots. I can shutdown,
>>> disconnect the charger, and remove the battery. The inaccuracy
>>> remains. At least once this process lead to a correct reading, but
>>> it hasn't worked recently. I'm not sure if this has gone on since I
>>> first got it or if this is a new thing. I noticed it about 3 months
>>> ago.
>>>
>>> This laptop is less than 6 months old, when disconnected from the
>>> charger the battery holds a charge for the same length it always
>>> has.
>>>
>>> Is there some file I can erase or some other procedure to reset the
>>> the indicator?
> [cut]
> I've noted similar behaviour in that I leave my machine connected to a
> mains supply after shut-down and the 80% charge seems to become a stable
> state, but after starting the system - Kubuntu in my case - the charge
> eventually reaches 100%, so it seems to be a function of the running/not
> running state of the m/c.
> 
> Peter HB

...which means it's distro-specific.
Maybe one distro asks the battery controller some questions the other
doesn't and responds accordingly.

I think I remember seeing Ubuntu charging my battery to 100% too on a
dual boot machine with Debian.

According to


http://www.samsung.com/us/article/tips--tricks-extending-notebook-battery-life

it's called "Smart charging", so I guess it should have a config option
somewhere.

Philip


-- 
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Archive: http://lists.debian.org/52038a4c.6020...@philipashmore.com

Re: battery

[Peter Hillier-Brook]

On 08/08/13 09:29, Philip Ashmore wrote:

On 08/08/13 04:05, cletusjenkins wrote:

I'm using wheezy and in gnome 3, if I disconnect my laptop from its
charger and then at some point suspend. After bringing the system
back up and connect the charger the battery charge indicator in
gnome shows incorrect info.

I plug in the charger and the battery charges, but it never reaches
100%, it stops at around usually in the 80's or 90's. The indicator
shows the battery is charging no matter how long it is plugged in,
but once the percentage stops increasing it will never go higher.
This inaccuracy usually persists across reboots. I can shutdown,
disconnect the charger, and remove the battery. The inaccuracy
remains. At least once this process lead to a correct reading, but
it hasn't worked recently. I'm not sure if this has gone on since I
first got it or if this is a new thing. I noticed it about 3 months
ago.

This laptop is less than 6 months old, when disconnected from the
charger the battery holds a charge for the same length it always
has.

Is there some file I can erase or some other procedure to reset the
the indicator?

[cut]
I've noted similar behaviour in that I leave my machine connected to a 
mains supply after shut-down and the 80% charge seems to become a stable 
state, but after starting the system - Kubuntu in my case - the charge 
eventually reaches 100%, so it seems to be a function of the running/not 
running state of the m/c.


Peter HB


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

Archive: http://lists.debian.org/52038524@hbsys.plus.com

Re: replacement for ide-generic in Wheezy?

[Jerome BENOIT]

Hello Gary,

On 08/08/13 11:39, Gary Dale wrote:
> The various ide modules (ide-generic, ide-cdrom, etc.) seem to be missing in 
> the Wheezy kernel. However they still compile a 486 kernel so I'm assuming 
> there is some intent to continue to support older hardware.
> 
> In my case, I have an old laptop that I could run Squeeze on by including 
> ide-generic in the /etc/initramfstool/modules file. However the same thing 
> doesn't work in Wheezy because there is no ide-generic module.
> 
> Googling around, I found one post that suggested it's been superseded by 
> libata but I can't find that either.
> 
> Is 486-era hardware still supported in the newer kernels or is this a lost 
> cause?
> 

I guess that it is still supported by the kernel, but not by the kernel deb 
package as distributed by debian.
You may build your own kernel, and the following link is a good place to start:

http://kernel-handbook.alioth.debian.org

hth,
Jerome


> 


-- 
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Archive: http://lists.debian.org/520372b3.5060...@rezozer.net

replacement for ide-generic in Wheezy?

[Gary Dale]

The various ide modules (ide-generic, ide-cdrom, etc.) seem to be 
missing in the Wheezy kernel. However they still compile a 486 kernel so 
I'm assuming there is some intent to continue to support older hardware.


In my case, I have an old laptop that I could run Squeeze on by 
including ide-generic in the /etc/initramfstool/modules file. However 
the same thing doesn't work in Wheezy because there is no ide-generic 
module.


Googling around, I found one post that suggested it's been superseded by 
libata but I can't find that either.


Is 486-era hardware still supported in the newer kernels or is this a 
lost cause?



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

Archive: http://lists.debian.org/52036764.7080...@velcom.ca

Re: battery

[Darac Marjal]

On Thu, Aug 08, 2013 at 09:29:49AM +0100, Philip Ashmore wrote:
> On 08/08/13 04:05, cletusjenkins wrote:
> > I'm using wheezy and in gnome 3, if I disconnect my laptop from its charger 
> > and then at some point suspend. After bringing the system back up and 
> > connect the charger the battery charge indicator in gnome shows incorrect 
> > info.
> > 
> > I plug in the charger and the battery charges, but it never reaches 100%, 
> > it stops at around usually in the 80's or 90's. The indicator shows the 
> > battery is charging no matter how long it is plugged in, but once the 
> > percentage stops increasing it will never go higher. This inaccuracy 
> > usually persists across reboots. I can shutdown, disconnect the charger, 
> > and remove the battery. The inaccuracy remains. At least once this process 
> > lead to a correct reading, but it hasn't worked recently. I'm not sure if 
> > this has gone on since I first got it or if this is a new thing. I noticed 
> > it about 3 months ago.
> > 
> > This laptop is less than 6 months old, when disconnected from the charger 
> > the battery holds a charge for the same length it always has. 
> > 
> > Is there some file I can erase or some other procedure to reset the the 
> > indicator?
> > 
> > 
> It's a feature of modern laptops.
> 
> The idea is that the battery lasts longer if you avoid full
> charge/discharge cycles.

Perhaps so, but if the charge isn't increasing, then that's basically
"full".

Batteries usually note three main values in their "fuel gagues" (a small
chip in the battery): current charge, current maximum charge and
designed maximum charge. The "designed maximum charge" is written at
manufacture time and the "current maximum charge" is updated every so
often.

The ratio between current charge and current maximum charge give your
your percentage full. The ratio between current maximum charge and
designed maximum charge gives an indication of battery health (that is,
if the battery was designed to hold 100Wh but now can only hold 20Wh,
it's basically dead).

Resetting the fuel gauge varies from battery to battery, but the typical
way is to drain the battery (so, take the charger out and run the laptop
until it shuts down. DON'T try to over-drain the battery by starting it
at this point), then charge the battery to full again (plug back in, but
don't start the laptop - let the battery charge at full rate until it's
at 100%. You'll need to estimate how long this usually takes). The idea
is that the fuel gauge monitors the increase in charge and, when the
increase tapers off, that's your maximum charge.

Don't try this too often, though. Deep discharges are also bad for
Lithium Ion batteries. If in doubt, read the manual :)



signature.asc
Description: Digital signature

Best practice vsftpd writable root inside chroot()

[Maik Stubbe]

Hi,

I just came accross this new security feature of vsftpd whith which you cannot 
have your $HOME writable for a chrooted user.

Since this ftp server is a public accessable server with ~60 users and partial 
nested homedirs, there is no chance that there is no writable homedirectory. 
Another big point is, the directory structure increased over years. So there is 
even no possibility to change it and give users an writable directory inside 
their $HOME.

So what is the best practice to avoid security problems?

1. Using no ftp -> no chance
2. Avoid using chroot is in my opinion a bad idea on a public accessable ftpd.
3. Compiling vsftpd >3.0 from source and using allow_writeable_chroot=YES: This 
would lead in using non Debian packages and watching them seperatly.
4. Using packages from Jessie: My preffered choice. But how to control security 
updates?

Any suggestions from your side?


Regards


-- 
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Archive: http://lists.debian.org/20130808083032.gd7...@stubbi.org

Re: battery

[Philip Ashmore]

On 08/08/13 04:05, cletusjenkins wrote:
> I'm using wheezy and in gnome 3, if I disconnect my laptop from its charger 
> and then at some point suspend. After bringing the system back up and connect 
> the charger the battery charge indicator in gnome shows incorrect info.
> 
> I plug in the charger and the battery charges, but it never reaches 100%, it 
> stops at around usually in the 80's or 90's. The indicator shows the battery 
> is charging no matter how long it is plugged in, but once the percentage 
> stops increasing it will never go higher. This inaccuracy usually persists 
> across reboots. I can shutdown, disconnect the charger, and remove the 
> battery. The inaccuracy remains. At least once this process lead to a correct 
> reading, but it hasn't worked recently. I'm not sure if this has gone on 
> since I first got it or if this is a new thing. I noticed it about 3 months 
> ago.
> 
> This laptop is less than 6 months old, when disconnected from the charger the 
> battery holds a charge for the same length it always has. 
> 
> Is there some file I can erase or some other procedure to reset the the 
> indicator?
> 
> 
It's a feature of modern laptops.

The idea is that the battery lasts longer if you avoid full
charge/discharge cycles.

Mine is on 80% too - it's normal.

As for how you tell it to charge to 100% because you're planning a bus
trip - anyone out there, please let me know.

Regards,
Philip Ashmore


-- 
To UNSUBSCRIBE, email to debian-user-requ...@lists.debian.org 
with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Archive: http://lists.debian.org/520356fd.1050...@philipashmore.com