Martin Michlmayr wrote:
* Joey Hess [EMAIL PROTECTED] [2008-02-10 14:55]:
Since dd is run with ibs= the full image size, quite a lot of ram is
needed to run flash-kernel. I just had it die when running it from
within d-i without swap mounted. I think that using a smaller input
buffer size
Martin Michlmayr wrote:
* Joey Hess [EMAIL PROTECTED] [2008-02-10 17:22]:
Do you have a patch? :)
Not yet. Do you remember why it uses ibs?
Here's what svn log says, but I really don't understand dd very well.
This was based on a comment on a mailing list or something.
It does makes
Martin Michlmayr wrote:
* Joey Hess [EMAIL PROTECTED] [2008-02-10 17:58]:
It does makes sense that a larger buffer size than the default 512
bytes (not 1 byte AFAIK) would speed it up somewhat, by saving syscall
overhead.
If there's a common divisor, like a block size, and it's
Package: os-prober
Severity: normal
Support needs to be written to parse the /boot/grub/grub.cfg file of
grub2..
-- System Information:
Debian Release: lenny/sid
APT prefers unstable
APT policy: (500, 'unstable'), (500, 'testing'), (500, 'stable'), (1,
'experimental')
Architecture: i386
Tomas Tintera wrote:
1. We need
ide/ide-core.ko,
ide/ide-disk.ko and
ide/ide-generic.ko
modules on 2nd (root) floppy (in initrd) in order to get
HDDs work. The space for it can be made by bzipping 2nd
floppy's initrd instead of gzipping it. Also the following
lines must be added to
Package: discover-data
Version: 2.2008.01.12
Severity: normal
discover-pkginstall wants to install 915resolution for my video card.
This package is obsolete with the X server for this card; I don't need
it. Suggest removing it from the package list, at least for this card
and others that are
Package: discover-data
Version: 2.2008.01.12
Severity: normal
There are several firmware packages in non-free. I suspect there will be
more soon, given all the ethernet driver firmware that has been removed
from the kernel lately. These packages are low-hanging fruit that it
would be easy for
I suggest that for the next beta we focus on dealing with non-free
firmware better in d-i. An increasing number of machines are difficult
to install with d-i, or don't have all hardware working fully
post-install due to non-free firmware issues.
There are clearly two parts to the problem;
Otavio Salvador wrote:
+ apt-install laptop-detect || true
If apt-install fails here, the script should be allowed to fail.
--
see shy jo
signature.asc
Description: Digital signature
maximilian attems wrote:
transmission is an easy and fast lightweight BitTorrent client.
gnome-btdownload lacks many features in comparison.
Another reason to consider this change would be that transmission
doesn't include an init script, which bittorrent does (by default it
does nothing,
Frans Pop wrote:
Have you ever actually _seen_ someone with a sparc laptop?
No. However, I have seen someone with a mips laptop. IIRC they got a
week's battery life running Debian. There are also plenty of things in
the laptop to pda spectrum using arm. mips, mipsel, arm, armel, armeb
should all
Gordon Farquharson wrote:
Attached is a patch to the nslu2-utils initramfs nslu2 hook that
updates the nslu2-utils package to work with Krzysztof's driver. I
have also taken the liberty of adding some logic to the initramfs
local-top nslu2 script to remove the module loading message that only
Martin Michlmayr wrote:
* Joey Hess [EMAIL PROTECTED] [2008-02-02 16:16]:
I've uploaded with this patch. Is ixp4xx_eth built into the 2.6.24
kernels in unstable?
No, it's a module.
So something still needs to be done to get udev to load it?
--
see shy jo
signature.asc
Description
Otavio Salvador wrote:
We have 2.6.22 as a safe bed on lenny now and their udebs are there
too however since EtchAndHalf intends to release with 2.6.24 and it
has been uploaded to sid already I'm considering a better option to
us to release with it.
linux-2.6 has been built
Otavio Salvador wrote:
Today while talking about the removal of laptop-detect from tasksel
depends list, specially for s390,
Why is s390 so special, and why do the s390 porters get away with
setting dependencies of important prority packages not-for-us, without
talking to anyone? Those are the
David Ayers wrote:
Package: tasksel
Version: 2.71
Followup-For: Bug #391356
Why did you post this as a followup to bug #391356? Please don't do
that, you've now cluttered up an unrelated bug report with your
unrelated mail, and you've hidden your problem away inside this other
bug report where
Robert Millan wrote:
Please move packages from Priority: important to tasksel so that user has the
option to de-select them.
Why? There should be nothing in important that a user should want to
de-select, by definition.
Also, there are things in important that are necessary for d-i and
tasksel
Robert Millan wrote:
Also, there are things in important that are necessary for d-i and
tasksel to work *at all*, including aptitude,
What's wrong with 'apt-install aptitude'?
It's not in the patch you sent.. (It would also need to install
tasksel.)
debian-archive-keyring, and
gnupg
Frans Pop wrote:
There really hasn't been any relevant change in choose-mirror that could
explain that.
Yes there was; I uploaded 2.20 from a clean vcs checkout, which is not
usual (I had to replace my hard drive recently).
But that said, Joey has made uploaded a change that may well
fix
Lucas Nussbaum wrote:
During a rebuild of all packages in sid, your package failed to build on i386.
Btw, should I report build failures of linux-kernel-di-i386-2.6 and
linux-modules-di-i386-2.6, or you don't care about them? I don't mind
reporting them when they occur, but if that's just
I installed sid (armel) to a thecus, and the installed system has no
/etc/environment, which isn't really a problem, except:
Jan 11 21:54:43 turtle sshd[17587]: pam_env(ssh:setcred): Unable to open env
file: /etc/environment: No such file or directory
So, localechooser used to create it .. what
Frans Pop wrote:
So the question still is _why_ ssh drops the connection in your case.
Some crappy routers and firewalls do this to TCP connections that pass
through them.
Also, the solution you propose is on the _client_ side, so is not something
we can fix in the installer. The only thing
The version of udev that supports the rtc symlink is 116.
No, it does not make sense to try to modprobe rtc-dev and only make the
symlink if the modprobe succeeds. rtc-dev is built into the kernel on
the nslu2 and some other hardware.
--
see shy jo
signature.asc
Description: Digital signature
Vincent McIntyre wrote:
Package: base-config
Version: 2.73
Chosen at random, I assume? Not even present in etch, version number
never distributed in debian, and never had anything to do with
/etc/host.conf.
On examining all the hosts I have currently installed with etch, I found
all of them
Frans Pop wrote:
Wasn't this change intended to show time remaining in the progress bar?
Only if used with a version of apt or aptitude that communicates it for
update. Apt doesn't (#448908); aptitude has other bugs (#448958) that
don't make it a good idea to use it here.
Also, if I select
Colin Watson wrote:
On Fri, Jan 04, 2008 at 05:23:24PM +0100, Frans Pop wrote:
On Friday 04 January 2008, Colin Watson wrote:
Err, how so? Surely this is #430545 (xresprobe breaks display at the
middle of installation).
I'd forgotten about that one. Yes, could very well be.
I
tzdata 2007b-1 (in etch) does not use debconf.
(BTW, several people in this thread have seemed to not be aware that
debconf-set-selections automatically sets the seen flag of preseeded
values.)
--
see shy jo
signature.asc
Description: Digital signature
Per Olofsson wrote:
So perhaps we should simply standardise on pm-utils. According to the Ubuntu
wiki page, acpi-support still needs to be around to send key events (for
IBM/Lenovo laptops which sends key events through ACPI, I presume). What we
can
do then is to replace hibernate with
Frans Pop wrote:
As I was extremely tired last night I decided to quit the discussion until
after some sleep.
On Monday 17 December 2007, Joey Hess wrote:
Frans Pop wrote:
Because a dpkg-reconfigure needs to ask the question even if a keymap
is installed.
if [ ! -e /etc
Frans Pop wrote:
On Tuesday 18 December 2007, Joey Hess wrote:
preseed's own base-installer script uses debconf-set-selections.
OK. That works, although we need to set the value too as otherwise there
isn't a question to set the seen flag for. So the patch becomes:
+# Avoid displaying
Frans Pop wrote:
Log:
After recent changes in apt-install, we need to tell console-data not to
display its keymap/policy question when it is being installed during a D-I
installation. console-data (= 2:1.04-2) will check for the existence of
the file
Frans Pop wrote:
See #456029 and especially #452331. Both BRs were discussed on the list and
I did ask for alternative suggestions.
I had hoped for a reaction from either you or Colin to my proposal. However,
Christian did not leave very much time before applying my patch which
probably
Frans Pop wrote:
We are talking about a debconf setting in the _installed_ system after all,
and one that should _only_ be valid while D-I is running.
preseeding console-data/keymap/policy seen should work ok. If
console-common is later dpkg-reconfigured, the seen flag will be ignored.
Frans Pop wrote:
Because a dpkg-reconfigure needs to ask the question even if a keymap is
installed.
if [ ! -e /etc/whatever_file ] || [ $1 = reconfigure ]; then
# ask question
fi
I decided on /tmp as the file really _is_ a temporary file: exists only for
the duration of the
Frans Pop wrote:
anna should probably check the status of a package before looking for it in
the repository.
anna does of course consider package status:
for (node = (*packages)-list.head; node; node = node-next) {
package = node-data;
Martin Michlmayr wrote:
* Gordon Farquharson [EMAIL PROTECTED] [2007-12-03 00:12]:
nslu2-utils is version 0.10+r71-13, but I don't see any changes to the
initramfs hook that should cause this problem. I am using a DFSG image
that does not contain the NPE-B firmware.
I guess the best way
Frans Pop wrote:
eth0: foo bar description, eth0: mac address: xxx:xxx... [slot 1]
That would be one way to do it without modifying debconf. You could also
get rid of the eth0: prefix if you wanted to by using Choices-C.
I'm probably just being thick, but what exactly are you proposing
dann frazier wrote:
* Modify cdebconf to support multi-line choice fields. Make each
interface choice be a multi-lined option that includes things like
vendor, model, mac, slot.
eth0: foo bar description, eth0: mac address: xxx:xxx... [slot 1]
That would be one way to do it without
Geert Stappers wrote:
Package: auto-install
Version: 1.2
Severity: normal
Hello,
In the install log are many lines like:
Nov 16 16:33:27 main-menu[1139]: INFO: Falling back to the package
description for auto-install
Nov 16 16:38:50 main-menu[1139]: INFO: Falling back to the
hasan murad wrote:
I am trying to build Build images with a custom kernel
$dpkg-buildpackage
it gets following error
debian/nic-extra-modules-2.6.8.18-386-di
lib/modules/2.6.18.8-2-386/kernel/drivers/base/firmware_class.ko
debian/irda-modules-2.6.8.18-386-di
Josef Wolf wrote:
d-i anna/choose_modules ppp-udeb
That's not formatted correctly.
d-i anna/choose_modules string ppp-udeb
--
see shy jo
signature.asc
Description: Digital signature
Eddy Petrișor wrote:
-Description: Standard system
+Description: Command-line environment
I agree with Robert; I have been asked quite a few times by friends (while
installing, what is the difference between standard and desktop system). This
change would make it a lot clearer what's
Frans Pop wrote:
You should probably change that to be a bare repository. There is no need
to keep all files checked out.
git-svn fetch fails on a bare repo.
--
see shy jo
signature.asc
Description: Digital signature
Frans Pop wrote:
We could probably make the debootstrap udeb arch all by moving pkgdetails
to base-installer and passing '--arch $(udpkg --print-architecture)', but
I'm not sure that's worth it.
It _would_ be handy if debootstrap could quickly be updated for breakage
w/o needing it to be
Joey Hess wrote:
There is no way to start aptitude interactively and at the same time
tell it to remove some tasks, so tasksel has to run aptitude twice when
you tell it to do both things.
Seems this is no longer true, so I'll implement that.
--
see shy jo
signature.asc
Description
Nis Martensen wrote:
Tasksel had Print Server, File Server, and Mail server selected.
Since there was no such task that I was looking for, I deselected the
above
I'm not sure why it wasn't clear to you that it had these selected since
it detected you already had those tasks installed, and that
Frans Pop wrote:
The patch may well introduce regressions, probably mainly in arches other
than i386/amd64.
The most likely regression looks to be ide-floppy, which AFAICS is not
autoloadable by udev. Richard Hirst at one point had ia64 hardware that
used ide-floppy (see #215455). Rather than
I've set up a git mirror of the svn repo, this is updated whenever a
commit is made:
git://git.debian.org/~joeyh/d-i.git
Total size of a checkout of all of d-i's history is about 400 mb.
The git-svn run to create this took 2+ days on a 2.3 ghz Xeon.
This looks basically nothing like real switch
Frans Pop wrote:
The create_devices function in base-installer has been simplified a lot
because it no longer actually has to create any device nodes anymore.
It now takes care of the bind mount to /target/dev and does some apt-install
calls for RAID, LVM and crypto installs.
(Suggestions
Frans Pop wrote:
# XXX This won't be needed once a new udev that handles the
# symlink gets into Debian.
if [ -e /dev/rtc0 ]; then
ln -sf rtc0 /dev/rtc
fi
If rtc-dev is now a generic driver,
Frans Pop wrote:
apt-install: support CD/DVD changing by using debconf-apt-progress if
multiple CDs are defined in sources.list.
Isn't debconf-apt-progress used unconditionally now?
+config=$(chroot /target debconf-apt-progress --config | sed s/$/;/)
+in-target sh -c $config
the progress bar.
+ * Allow the progress bar to be canceled (needs debconf 1.5.17 for proper
+operation).
+ * 50mirror: Call choose-mirror -n to avoid trashing the progress bar.
+
+ -- Joey Hess [EMAIL PROTECTED] Fri, 02 Nov 2007 16:33:10 -0400
+
apt-setup (1:0.31) unstable; urgency=low
Frans Pop wrote:
It's logical from the installer's pov, but not from a _user's_ pov.
How the hell is he supposed to guess that he will be asked about additional
CDs later when confronted with a question about a network mirror?
If I (as a newbie) were confronted by that question I would
Christian Perrier wrote:
Frans' proposal adds an extra question when doing CD installs, to
prompt users whether they want to use another CD or not. This, even if
they did choose a network mirror before.
It asks for all the CDs first, there's no path through the install that
asks about a
Praveen A wrote:
I think even now if you select a mirror and installing from CD it
downloads even those packages in the CD, correct me if I am wrong
here.
Incorrect.
(As is much of the rest of your mail. Please lose the chip on your
shoulder if you'd like to influence me *at *all*.)
--
see
Frans Pop wrote:
On Thursday 01 November 2007, Joey Hess wrote:
If you're not on broadband and you choose to use a mirror, it takes
half an hour or more just to download the Packages file. Apt can
estimate this reaonably well so you shouldn't even need to wait, you can
just see
Frans Pop wrote:
If you have both CDs and a mirror in your sources list and a package is
available from both and the version of the package on both is the same,
apt(itude) will always take the package from the CD.
Yes, he wasn't saying this, but the opposite, AIUI.
--
see shy jo
Frans Pop wrote:
This basically means that the current time-out is just too long for
practical use as is.
An alternative option could be to make it possible to cancel the action,
just like we do for looking for a DHCP server.
Yep, all network-facing progress bars in d-i need a cancel
block 448871 436497
thanks
The clock-setup progress bar cannot safely be made cancelable until this
bug in newt is fixed. To make the progress bar cancelable, clock-setup
would need to use PROGRESS INFO or PROGRESS_STEP periodically to poll
for a return code indicating cancel was hit. But both of
Joey Hess wrote:
I don't feel this will be too hard to implement, but obviously there's
no point on blocking your changes on it.
With the debconf-apt-progress from debconf 1.5.17, if the progresscancel
capb is set in d-i, hitting cancel on the progress bar will send first a
SIGINT
Rick Thomas wrote:
Using ntp to set the time should be a short operation. If it takes a long
time, the validity of the time that results is in question -- by virtue of
the process[*] being used.
Careful, you're confusing SNTP and NTP. You're also oversimplifying. NTP
is very good at
Otavio Salvador wrote:
Joey Hess [EMAIL PROTECTED] writes:
Try this:
debconf-apt-progress aptitude update
You can't use apt-get because it doesn't output the timing info to the
status-fd, but happily aptitude works. And for results that best mimic
d-i, you should add a new mirror
+to specify how it should advance the progress bar.
+ * Allow the progress bar to be canceled (needs debconf 1.5.17 for proper
+operation).
+
+ -- Joey Hess [EMAIL PROTECTED] Thu, 01 Nov 2007 17:17:18 -0400
+
apt-setup (1:0.30) unstable; urgency=low
[ Frans Pop ]
Index: apt-setup
Rick Thomas wrote:
Ahhh... So the installer time-setter uses rdate, not NTP? A wise choice,
now that I think about it. Much lighter weight protocol, and the client is
much *much* smaller.
Infact, rdate doesn't even use SNTP.
rdate, as used by the installer, uses SNTP.
The point
Frans Pop wrote:
On Friday 02 November 2007, Joey Hess wrote:
Here'a an attempt at cancelable progress bars for apt-setup. If apt's
bug gets fixed the progress bar will also include download speed
and time estimates.
The behavior when cancel is pressed could be improved, it currently
Frans Pop wrote:
OK, but in your proposal the option to scan additional CDs is _not asked at
all_ if they choose to use a mirror. A significant difference I'd say.
Asking the CD question in expert mode if a mirror has been configured is
not significantly different than allowing lilo to be
Gordon Farquharson wrote:
Oct 27 06:56:31 main-menu[2481]: INFO: Menu item 'nobootloader' selected
Oct 27 06:56:46 main-menu[2481]: INFO: Menu item 'finish-install' selected
finish-install depends on bootable-system, which is provided by both
nobootloader and by flash-kernel-installer.
Otavio Salvador wrote:
Log:
* Change tzsetup-udeb.install to avoid .svn dirs on post-base-installer.d.
If you're building direct from svn, set DH_ALWAYS_EXCLUDE=.svn and it will
automatically avoid this problem in all cases.
Or just svn export and build from the export..
--
see shy jo
Anthony Towns wrote:
To get it working for d-i uploads, I need a very reliable script that
will be invoked as:
Well, I stopped when I discovered the tar on ries is still apparently
vulnerable to #439335. I don't feel it's possible to make a very
reliable script with an insecure tar..
(Does dak
Frans Pop wrote:
Works fine inside D-I too.
One nitpick: should the new option be listed in the SYNOPSIS of the manpage?
That's currently only listing the more important options.
Otherwise, please upload.
Done.
--
see shy jo
signature.asc
Description: Digital signature
I wonder if you've considered the wrinkle of anna-install being run when
a CD other than the boot CD is in the drive? anna-install is rarely called
after base-installer, but it can happen.
Maybe cdrom-retreiver should store a volume ID of the CD and check that ID
and handling a media change.
--
Frans Pop wrote:
+db_input high apt-setup/cdrom/set-first || true
+db_go || exit 10
+db_get apt-setup/cdrom/set-first
If this generator ran after the 50mirror generator it could check to see
if network sources were enabled in sources.list. If they are, and
a DVD is available, there's probably
Frans Pop wrote:
My thought was that the user should first be allowed to load additional
CD/DVDs or not and based on that we _could_ change the defaults/priority
for using a mirror.
OTOH, as you've argued yourself in the past, having additional CDs available
does not guarantee that all
Frans Pop wrote:
I beg to disagree.
As argued above, loading multiple CDs in no way conflicts with using a
mirror. IMO we have to always ask both questions.
Neither does installing lilo to one disk and grub to another, but d-i
does not prompt the user to choose if they want to use lilo, and
Geert Stappers wrote:
I miss the why. Right now I have the unplesant feeling it is apt-cdrom
add re-invented. Someone willing to enlighten me?
The sole use-case for multiple CDs in d-i is a scenario where the user
is not connected to (or chooses not to use) a fast network and is going
to select
Frans Pop wrote:
When base-config was integrated into D-I, we had find a way to make tasksel
install from CD images for hd-media installations. This was solved by bind
mounting the CD into target and telling apt-cdrom not to mount/unmount CDs.
In the patch for apt-setup this is reverted:
Frans Pop wrote:
On Thursday 25 October 2007, Frans Pop wrote:
The prompt for CD changing works during pkgsel, but apparently not for a
normal apt-install. During initial tests, installation of grub failed
because CD2 was still mounted and I never got the prompt to change CDs.
This is
Frans Pop wrote:
What's the reason we made apt-install non-interactive? Was it just that we
did not have the passthrough frontend at the time? If so, I guess we should
probably just make apt-install interactive in all cases.
Both that we lacked working passthrough, and that none of these
Martin Michlmayr wrote:
* Frans Pop [EMAIL PROTECTED] [2007-10-24 23:02]:
Suggest you try a custom installer with Etch's main-menu to see if that
fixes it. If it does, a serious BR against main-menu would be in order.
I copied th etch main-menu binary to the image and saw the same
I was reviewing #62 and it explains at least part of the behavior.
main-menu would defer running localechooser until it was run to satisfy
a dependency, in the case where the boot image didn't have localechooser
on it. This was because localechooser was downloaded and added to the
top of the
Martin Michlmayr wrote:
INIT: Entering runlevel: 2
Starting system log daemon
Starting kernel log daemon
Starting OpenBSD Secure Shell server: sshd.
Starting periodic command scheduler: crond.
Below are the last few messages in syslog. I ran out of time before I
was able to fully
Joey Hess wrote:
+ in-target sh -c $config debconf-apt-progress --logstderr -- apt-get -o
APT::Install-Recommends=false -y --no-remove install $packages ||
ERRCODE=$?
This resets the progress bar to 0, advances it to 100, and then STOPs
it, which is not desirable if there is already
Martin Michlmayr wrote:
Do you intend to upload 1.22 in the near future?
If the DIRMs don't mind, I think it's in an uploadable state. I didn't
know for sure of any instances of the bug still left in the wild, so
didn't upload it.
--
see shy jo
signature.asc
Description: Digital signature
dann frazier wrote:
The Packages file isn't being overwritten; its being appended to.
Ah, sorry for missing that point.
The name mirror/udeb/override_suite is easily confused with mirror/udeb/suite,
especially since both are basically used to override the udeb suite.
dann frazier wrote:
+Template: pkgsel/progress/fallback
+Type: text
+_Description: Running ${SCRIPT}...
Index: debian/postinst
This will very briefly display Running popcon, followed by
popularity-contest asking if you want to participate. I can envision a
user wondering if it already
How is setting mirror/udeb/override_suite any different than setting
the existing mirror/udeb/suite?
dann frazier wrote:
+ for suite in $codename $override_suite; do
+ fetch dists/$suite/Release $Release || exit $?
+ fetch dists/$suite/$pkgfile $Packages || continue
Otavio Salvador wrote:
Joey Hess [EMAIL PROTECTED] writes:
How is setting mirror/udeb/override_suite any different than setting
the existing mirror/udeb/suite?
I think Dann's idea is to get both and validate them.
Why? Apt already validates all Release files it uses. d-i can only use
dann frazier wrote:
I want to use the udeb in the primary suite if there is no
corresponding udeb in the override_suite. The override_suite only
includes udebs that require customization. Does that answer your
question?
No, because I don't see how it works. Like I said, it downloads the
dann frazier wrote:
This patch adds an /usr/lib/pre-pkgsel.d/ directory where hooks can be
added in a run-parts fashion. popcon handling has been factored out
into its own script in this directory.
My main motivation for this patch is that I wanted the ability to
install a package
dann frazier wrote:
OK. Just for my benefit, what's the difference? I assume the 'echo -n'
above my change should get changed as well?
It's not POSIX and is grandfathered in in policy. So don't worry about
it, really..
--
see shy jo
signature.asc
Description: Digital signature
dann frazier wrote:
+ # If DHCP provides multiple servers, ignore all but the first
+ dhcp_ntp=`echo $RET | cut -d' ' -f1`
Please quote $RET here to make it less easy for a rogue NTP server to
(accidentially) root d-i. :-)
Index: dhclient-script
Frans Pop wrote:
I agree with Otavio that using a tab as column separator is not a good
idea, but for a completely different reason.
I can imagine you'd have a case where you'd want a string including
columns to be translated. If you use a tab, this is almost guaranteed to
result in broken
Steve Langasek wrote:
On Mon, Oct 15, 2007 at 05:51:38PM -0700, Don Armstrong wrote:
Package: debian-installer
Severity: normal
The current daily builds (10/15/2007) have the 2.6.21 kernel in
vmlinuz and the initrd, but the udebs for 2.6.22.
For which architecture? This is going to
Neil Williams wrote:
OK. I've done the doclifter thing and updated the XML, generated a new
manpage and compared it with the old. I'm assuming you don't want the
build-dependency on docbook-xsl so I've included a patch to create a
README that documents how to use xsltproc to generate the
Micha Beyer wrote:
the debianinstaller mean I have no dhcpclient (no dhclient or pump) in
the netinstallimage for i386 (lenny).
What's the problem with the dailybuildimage?
You're using a testing_d-i image, which uses etch's d-i still. Download
a sid_d-i image instead. Any of the links on
Otavio Salvador wrote:
This isn't accurate enough, I'm sorry. I think that any char we choose
as a separator needs to have a way to be escaped.
A tab *means* advance to the next column. Therefore, it would do
exactly what it's intended to do, therefore there is no point in
escaping it.
--
see
Jerry Quinn wrote:
Right now it's
Linux cerberus 2.6.22-2-amd64 #1 SMP Thu Aug 30 23:43:59 UTC 2007 x86_64
GNU/Linux
but I had 2.6.21 installed by the installer (from aptitude):
ii linux-image-2.6.21-2-amd64 2.6.21-6Linux 2.6.21 image on AMD64
What is the output of: ls -l
Andrew Lee wrote:
-kde-i18n-zhtw
Why installs this kde message on GNOME desktop?
Because these tasks are also used to localise the kde-desktop, until
someone splits them.
+icedove-locale-zh-tw
New package for icedove's zh-tw translation.
icedove is not installed by any desktop tasks, so I
Denys Vlasenko wrote:
Please send testcases where GNU gzip works and busybox gzip doesnt - to me.
I will try to fix those.
It's easy enough to reproduce for me:
[EMAIL PROTECTED]:~echo foo | gzip test.gz
[EMAIL PROTECTED]:~echo bar | gzip test.gz
[EMAIL PROTECTED]:~gunzip test.gz
foo
bar
Peter Rock wrote:
Hello, I hope I'm contacting the right folks!
I'm trying to find out the copyright status of the elilo-installer
package to see if it qualifies as free software or not. Below is a bit
of a description and I was advised by debian-legal to ask the package
maintainers. Hope
701 - 800 of 4877 matches
Mail list logo
