The problem is actually not in find, it's in ash. When invoked from hw-detect
as (e.g.)
find /lib/modules/2.4.21-5-386/kernel/ | grep -q ide-scsi
grep exits before reading the entire output from find, and find gets killed
with SIGPIPE. Normally, this would not be a problem (ash would see that t
Package: cdebconf-udeb
Version: 0.41
Severity: normal
When cdebconf (at least cdebconf with the newt-frontend, which I'm using) is
set to display a template with a choice containing at least one slash character
('/'), it appears to freeze totally. I've tried this both with the latest CVS
as well a
Hmm, actually I can't reproduce this in "make demo" on my other machine...
I'll try to look into it again and see if I can track it down (ie. I'll try
triggering the bug again and take it from there :-) )
/* Steinar */
--
Homepage: http://www.sesse.net/
--
To UNSUBSCRIBE, email to [EMAIL PROTE
Package: usb-discover
Version: 0.02 (not installed)
Severity: important
Tags: patch
If /lib/debian-installer.d/S45usb-linux is run on a system without PCI,
it does an "exit 0":
if [ ! -e /proc/bus/pci/devices ]; then
# PCI not present. quit
exit 0
fi
However, since this script is
Package: base-installer
Version: 0.029 (not installed)
Severity: important
Tags: patch
If multiple extra packages are queued for installation, base-installer
sends them all to apt-install at the same time, which fails if one of
them is not installable. Please see the included patch, which makes it
Package: main-menu
Version: 0.037 (not installed)
Severity: normal
Tags: patch
main-menu is supposed to skip a package if a "parallel package" already
is installed (for instance, grub-installer if lilo-installer is already
installed), but fails to do so. The reason is that the isdefault()
heuristi
Package: autopartkit
Version: 0.57 (not installed)
Severity: wishlist
Tags: patch d-i
Before asking autopartkit/confirm (the message that asks the user to
confirm the deletion of everything on all hard drives), autopartkit
fsets autopartkit/confirm's "seen" attribute to false, which prevents
(amon
Package: prebaseconfig
Version: 0.034 (not installed)
Severity: normal
Tags: d-i patch
prebaseconfig.d/99reboot contains the following line:
db_fset prebaseconfig/reboot_in_progress seen false || true
Which prevents an automatic install (in my case, the user has already
been prompted to remove
Hi,
The included patch removes all internationalized template strings from the
d-i installer disks. It's quite rough and should of course not be applied
unconditionally, but if you're really cramped for space, it could be very
useful to gain those extra few kilobytes -- at least, it was the thing
It turns out the previous patch misinterpreted exit codes from menutest
scripts -- an exit value of 1 (true) means "don't care", not "don't
select". (languagechooser, for one, is affected by this.) This patch
(which should be applied on the top of the last one) should make things
work again. Apolog
On Tue, Sep 16, 2003 at 12:58:40AM +0100, Daniel Silverstone wrote:
> I fired up a bochs machine and booted the netinst CD
>
> It sat looping saying "unable to load vesafb.o"
This is a bug in usb-discover, which should be fixed in the latest upload,
ie. it should reach the netinst CD soon. :-)
>
On Tue, Sep 16, 2003 at 12:30:44AM +0200, Petter Reinholdtsen wrote:
> Was this bug fixed with the following upload?
>
> busybox-cvs (0.60.99.cvs20030819-3) unstable; urgency=low
>
>* shell/ash.c
> - fix signal handling (upstream)
I haven't seen the "EXIT EXIT EXIT" bug today, at leas
On Tue, Sep 16, 2003 at 02:44:29AM +0200, Sebastian Ley wrote:
> This is a known bug, although I do not remember which package is
> responsible. I believe it is busybox, but I can't find a bug report.
> Will investigate tomorrow.
#208810 for busybox-cvs-udeb, #211009 for dash.
/* Steinar */
--
H
Package: autopartkit
Version: N/A; reported 2003-09-16
Severity: minor
Tags: d-i
The description for autopartkit claims it leaves and resizes existing
FAT partitions, which is wrong -- currently it destroys everything on
disk. This should probably be fixed. :-)
-- System Information
Debian Releas
Package: rootskel
Version: 0.36
Severity: normal
Tags: patch d-i
/linuxrc has a line saying
exec chroot . /linuxrc /dev/console 2>&1
On my own images, I've had problems with this script not finding chroot.
Changing the line to
exec /usr/sbin/chroot . /linuxrc /dev/console 2>&1
fixes the pr
Please file all bug reports about udebs to either "installation" or the udeb
in question -- boot-floppies does not know about udebs and is unrelated to
them.
Second, I can't see the problem -- scsi-modules-2.4.21-1-386-udeb does
indeed have both aic7xxx/aic7xxx.o and aic7xxx_old.o.
/* Steinar */
Package: netcfg-dhcp
Version: 0.25 (not installed)
Severity: normal
Tags: patch d-i
netcfg-dhcp does not make an /etc/network directory during package build
nor runtime, yet tries to write to /etc/network/interfaces. This causes
an error (although a non-fatal one, d-i can live without that file) i
Since I've already filed this as #211260, I'm merging the two bug reports and
setting ownership to rootskel.
/* Steinar */
--
Homepage: http://www.sesse.net/
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Package: cdebconf
Version: unavailable; reported 2003-09-17
Severity: normal
Tags: d-i
When kbd-chooser is run (at least from the netinst CDs), and gets to the
point "input medium console-tools/arch", the frontend crashes with
SIGSEGV, which makes it impossible to use kbd-chooser. The bug is not
On Tue, Sep 16, 2003 at 09:56:08PM +0100, Daniel Silverstone wrote:
> On Tue, 2003-09-16 at 19:56, Steinar H. Gunderson wrote:
> > Severity: minor
> [snip]
> > currently it destroys everything on disk.
>
> Perhaps I'm alone in thinking this, but surely this is kinda
Package: busybox-cvs-udeb
Version: 0.60.99.cvs20030819 (not installed)
Severity: normal
Tags: patch d-i
busybox-cvs-udeb should be built with CONFIG_FEATURE_WGET_STATUSBAR.
Rationale:
1) It is very useful during large downloads (or on slow links) to
display some kind of progress information to
Sorry, the patch did not get included last time -- attaching it now.
/* Steinar */
--
Homepage: http://www.sesse.net/
diff -ur busybox-cvs-0.60.99.cvs20030819/debian/config-udeb
busybox-cvs-patched/debian/config-udeb
--- busybox-cvs-0.60.99.cvs20030819/debian/config-udeb 2003-09-17 23:15:05.00
The patch has been applied in current CVS.
/* Steinar */
--
Homepage: http://www.sesse.net/
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
I've updated the package description in current CVS, so this should be fixed
in next upload.
/* Steinar */
--
Homepage: http://www.sesse.net/
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Hi,
I've fixed the problem with kbd-chooser crashing the cdebconf frontend
(making it impossible to run it). The problem in this case was that the
template in question (console-tools/archs) didn't have an extended
description, and thus returned NULL. The frontends expected
question_get_field() to
On Thu, Sep 18, 2003 at 02:45:37AM +0200, Steinar H. Gunderson wrote:
> I can't find any bugs filed against this (there is one for kbd-chooser, but
> that's a different issue)
Doh, I did file one myself a while ago (#211315). Added the correct Closes:
line to CVS; apologies fo
The previous patch didn't quite work as intended -- try this one instead (it
has actually been tested in a live d-i environment :-) ).
/* Steinar */
--
Homepage: http://www.sesse.net/
Index: tools/autopartkit/autopartkit.c
===
RCS f
On Thu, Sep 18, 2003 at 02:03:59PM +0100, Richard Hirst wrote:
> My guess was that 'id' might require more
> libraries (*nss* stuff), but I havn't invesigated further yet.
An strace seems to indicate that, yes -- it looks for libnss_files.so.1 but
can't find it. Without it, it finds and parses /et
On Thu, Sep 18, 2003 at 09:54:33PM +0200, Steinar H. Gunderson wrote:
> An strace seems to indicate that, yes -- it looks for libnss_files.so.1 but
> can't find it. Without it, it finds and parses /etc/passwd correctly, but not
> /etc/group for some reason. busybox has a libpwdg
On Thu, Sep 18, 2003 at 09:38:42PM +0200, jan colpaert wrote:
> I recently created iso DVD's for sid and sarge.
> When booting the CD/DVD, I cannot get past the choose keyboard option.
> The kbd-chooser program crashes always, before you have the chance to
> choose a keyboard.
> It is impossible t
On Fri, Sep 19, 2003 at 12:13:56AM +0200, VETSEL Patrice wrote:
> I don't know how the iso are compiled each days.
They are compiled from the latest packages in the Debian repository.
> I think the d-i stay the same, and only packages are updates.
That is true, but once a package is uploaded, it
On Thu, Sep 18, 2003 at 10:29:44PM +0200, Steinar H. Gunderson wrote:
> Simply adding "./libpwdgrp/libpwdgrp.a ./libbb/libbb.a" (libpwdgrp.a needs
> stuff from libbb.a, too) to the end of the command line linking busybox fixes
> the problem, but there must surely be a more
Package: busybox-cvs-udeb
Version: 0.60.99.cvs20030819
Severity: normal
Tags: patch upstream d-i
busybox has its own internal libpwdgrp, providing standard functions
like getgrnam() etc. without having to resort to external libc (which
uses NSS, which does not work during a d-i install since the N
Forgot to add the patch... again. :-) *sigh*
/* Steinar */
--
Homepage: http://www.sesse.net/
diff -ur busybox-cvs-0.60.99.cvs20030819/Makefile
busybox-cvs-0.60.99.cvs20030819-patched/Makefile
--- busybox-cvs-0.60.99.cvs20030819/Makefile2003-09-17 23:15:05.0 +0200
+++ busybox-cvs-0.
Just wanted to mention that it was pointed out that one could solve this in a
slightly different (and uglier, but more portable) way: Instead of doing
gcc -o busybox -Wl,--start-group lib1.a lib2.a lib3.a -Wl,--end-group
one could do
gcc -o busybox lib1.a lib2.a lib3.a lib1.a lib2.a lib3.a
When testing the latest netinst images today, I noticed that even after
selecting a language, main-menu shows only English descriptions for all menu
items (but the right language is used for everything else). Further
investigation revealed that the translations simply don't exist. kbd-chooser
has a
On Mon, Sep 22, 2003 at 05:37:14PM +0200, Simon Hürlimann wrote:
> The control file of base-installer has XB-Description-nb entries. As I
> can't
> find any other XB-Descriptions I'm not sure if you still support that (or
> maybe not yet?). If so, I could write some XB-Description-de entries.
I as
Hi,
syslog reveals the following error when trying to load fr-latin1:
installer[9985]: kbd_chooser: setting keymap fr-latin1
installer[9985]: kbd-chooser: unknown keysym 'Control_braceleft'
installer[9985]: kbd-chooser: /usr/share/keymaps/i386/azerty/fr-latin1.kmap:18:
syntax error
insta
On Fri, Feb 28, 2003 at 10:38:34PM +0100, Petter Reinholdtsen wrote:
> The slang frontend is missing progress bar support. The screen go
> blue, but no text appear. The progress bar should be implemented. It
> works in the text frontend.
I've implemented preliminary progress bars (not very pret
A fix for this bug has been checked into CVS.
/* Steinar */
--
Homepage: http://www.sesse.net/
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Hi,
I can't reproduce #195704. I set TERM=dumb and run the progress bar test with
the text frontend in cdebconf, but everything appears normal. Has this been
fixed, or am I missing anything?
/* Stienar */
--
Homepage: http://www.sesse.net/
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a
Package: busybox-cvs-udeb
Version: 0.60.99.cvs20030819-3 (not installed)
Severity: important
Tags: patch d-i upstream
A typo (OPTIMIZATION is used some places and OPTIMIZATIONS others)
prevent busybox-cvs-udeb from being built with optimization (only
-fomit-frame-pointer is activated, no -O option
On Tue, Aug 26, 2003 at 10:36:44PM +0200, Sebastian Ley wrote:
> The image built for the floppy type is too large:
With fixed busybox-cvs-udeb (see bug #212485) the floppy images appear to
build again, so once #212485 is resolved and upload we can just continue as
usual without having to slash ude
On Wed, Sep 24, 2003 at 10:37:20AM +0200, Bastian Blank wrote:
> The text frontent don't look at TERM and you have a terminal which
> supports \r.
Ah, so the real bug is that the text frontend doesn't look at TERM? :-)
/* Steinar */
--
Homepage: http://www.sesse.net/
--
To UNSUBSCRIBE, email
Package: base-installer
Version: unavailable; reported 2003-09-26
Severity: normal
Tags: d-i
The fix for #210915 seems to have introduced a new bug. Multiple
packages seem to apt-install discover, which then gets put in the queue
multiple times. Since we do separate installation of every package i
Could it be that /sbin/grub-install needs a working /dev? Do we have a
working /target/dev at grub installation time?
/* Steinar */
--
Homepage: http://www.sesse.net/
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
I've implemented solution #2 (it was a two-line fix) in rootskel CVS --
tagging this bug as pending and reassigning it to rootskel.
/* Steinar */
--
Homepage: http://www.sesse.net/
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
The changes done in CVS before autopartkit 0.58 upload seem to have broken my
patch -- at least I'm still getting asked even when fset seen=true and
value=true, which I didn't with my patch. Reopening.
/* Steinar */
--
Homepage: http://www.sesse.net/
--
To UNSUBSCRIBE, email to [EMAIL PROTECTE
OK, here's a slightly different proposal for how to handle "fset seen" stuff
(there is code in d-i ATM that just does "fset seen false" on a question
before asking it, which is very very bad in an automated install...):
If "seen=true", then simply lower the priority one notch before deciding to
as
On Fri, Sep 26, 2003 at 06:18:40PM +0200, Petter Reinholdtsen wrote:
> There should be an noninteractive frontend in cdebconf, to handle
> automatic installation. It should use the currently set value for all
> debconf questions, and behave like the noninteractive debconf
> frontend.
Does this me
Joey and I had a discussion on IRC today, and we agreed on what we think is a
relatively sane way to handle the "fset seen" issues:
No package should "fset seen=false" unless it has a very good reason -- that
is, it has to show some critical error message or such. However, there are
also situation
On Tue, Sep 30, 2003 at 03:09:25PM -0700, Matt Kraai wrote:
> I think the Depends field for partitioner should be
>
> Depends: cdebconf-udeb, fdisk-udeb, hw-detect-full
>
> on hppa and left as
>
> Depends: cdebconf-udeb, parted-udeb, hw-detect-full
>
> on other architectures. I'm not sure ho
On Tue, Sep 30, 2003 at 04:32:20PM -0700, Matt Kraai wrote:
>> Make it depend on "partitioning-program", and make both fdisk-udeb and
>> parted-udeb "Provide[s]:" that. Should solve the problem.
> No, it won't. anna could still install just parted-udeb since
> that would satisfy the dependency. I
On Tue, Sep 30, 2003 at 04:52:51PM -0700, Matt Kraai wrote:
>> If parted-udeb is unusable on a particular platform, I'd take it it was
>> never built/uploaded for that platform at all...?
> Nope, it's
>
> Architecture: any
>
> so it's built for hppa even though it doesn't fully work there.
Well
On Tue, Sep 30, 2003 at 05:22:21PM -0700, Matt Kraai wrote:
>> Well, if it isn't usable (remember, it's an udeb, so we're only talking in
>> the d-i context here) for hppa, hppa should not be in the list of
>> architectures for that package, IMO.
> OK. partitioner still needs to Depend on the part
On Tue, Sep 30, 2003 at 07:28:37PM -0700, Matt Kraai wrote:
> Because it can't use just any partioning program, it uses a
> particular one (or, actually, a particular one, two, or three)
> depending on the architecture. If we depend on a virtual
> partitioning-program package, it could install par
On Wed, Oct 01, 2003 at 07:01:49AM -0700, Matt Kraai wrote:
> On New World PowerPC systems, we use mac-fdisk. On PREP PowerPC
> systems, we use fdisk.
OK, so it's a subarchitecture problem, not an architecture problem. That was
what I asked about four mails ago. :-)
/* Steinar */
--
Homepage: h
Package: cdebconf-newt-udeb
Version: N/A; reported 2003-10-01
Severity: normal
With the "progress bar status text wrapping" patch that was made at
debcamp, the dialog changes size all the time (it changes height) to
make room for the new text. When the text changes frequently (like
installing pack
On Thu, Oct 02, 2003 at 10:29:10AM +0200, Sven Luther wrote:
> On the same issue, i maintain the (non-free) unicorn driver for Bewan
> ADSL modem lines.
>
> I can if need be provide udebs for said driver, but i am not sure about
> the possibility to use non-free udebs in d-i ?
Without being a Deb
On Thu, Oct 02, 2003 at 05:48:42AM -0700, Matt Kraai wrote:
> This isn't such a problem on my 128-column console. :)
>
> I'd prefer to only let it grow, rather than truncate the text.
But when the progress meter changes (since a new action is being started, for
instance) it will go back to the s
I've looked around a bit since my last post on this issue, and I've found out
how it's implemented: cdebconf heeds "fset seen" if and only if
"debconf/showold" is set to "false", _or_ the environment variable
DEBCONF_SHOWOLD exists and is set to "false". (Or both, of course.)
Does anybody know why
On Sun, Oct 05, 2003 at 02:14:49PM +0200, Goswin von Brederlow wrote:
> cdebconf build-depends on d-shlib twice
Fixed in CVS, thanks.
/* Steinar */
--
Homepage: http://www.sesse.net/
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECT
On Tue, Oct 07, 2003 at 10:05:56PM +0800, Eva Yip wrote:
> I am an architecture student and I am doing a project which needs to make
> big mirrors, could you tell me how I could DIY a mirror?
For Debian mirrors, try the package debmirror.
/* Steinar */
--
Homepage: http://www.sesse.net/
--
T
On Wed, Oct 08, 2003 at 08:34:05PM -0400, Sean Finney wrote:
> when running base-config after installing sid, the user gets prompted
> whether or not to use security updates from security.d.o. while it's
> very thoughtful, it's not too much help for those who use sid, and does
> put extra strain o
On Thu, Oct 09, 2003 at 02:58:01PM +0200, Sebastian Ley wrote:
> In partitioner and partconf you have a select question generating some
> kind of menu, together with a "Finish" entry. However it turned out that
> people believed that the "Weiter" button would have the meaning of the
> Finish button
On Thu, Oct 09, 2003 at 03:17:25PM +0200, Sebastian Ley wrote:
>> What about simply dropping the "Finish" and "Abort" selections? They should
>> be the "Finish" and "Back" button anyhow, not entries in a selection box.
> And what will you do with that button in languagechooser? Or in the
> filesyst
On Fri, Oct 10, 2003 at 07:52:42PM -0400, Joey Hess wrote:
> The current syslinux stuff for the boot floppies is quite bare. I don't
> remember about the cdrom. We need a nicely designed set of boot screens
> that is easy to ready and not offputting, but also provides all the
> information a user
On Sat, Oct 11, 2003 at 03:18:46PM +0200, Javier Fernández-Sanguino Peña wrote:
> Some come to mind, either global: criteria based on the
> percentage of population that talks a given language (based on ethnologic
> data), criteria based on the languages things (books, movies..) get
> translated in
Hi,
The businesscard .iso is now about ~60MB, which probably won't even fit on
most "businesscard"-sized CDs. Could somebody please answer:
- Why is the entire Debian keyring (6MB) on the CD?
- Why are the net_drivers, cd_drivers and other floppies (at least 7-8MB) in
/install? They do not seem
On Sat, Oct 11, 2003 at 01:41:08PM -0400, Joey Hess wrote:
>> - Why is the entire Debian keyring (6MB) on the CD?
> Probably to support checking signatures of Release files.
Do we use that at all? :-)
>> - Why are the net_drivers, cd_drivers and other floppies (at least 7-8MB) in
>> /install? T
On Sat, Oct 11, 2003 at 08:28:51PM +0200, Geert Stappers wrote:
> Sort the lang. list on international telephone country code.
> People do known their country code, that makes skipping easier.
Languages may very often map to multiple countries, and sometimes countries
map to multiple languages as
Relocation errors in d-i usually means some new binary was brought in (a new
version of busybox, it appears), but libc-udeb (containing the full,
unreduced libc.so.6) was not. Could you check the logs and see if you ever
get to download libc-udeb before anna gives up?
/* Steinar */
--
Homepage: h
As this (progress bars for slang frontend) has been fixed in CVS and nobody
seems to complain about its uglyness, I'm closing this (old) bug. :-)
/* Steinar */
--
Homepage: http://www.sesse.net/
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EM
On Wed, Oct 15, 2003 at 11:44:44AM -0400, Joey Hess wrote:
> See attached log, shows it loading busybox-udeb just before relocation
> hell.
busybox-udeb should not be loaded (priority extra), yet the debug log claims
busybox-udeb is priority ">= standard". This does not happen with my network
flop
On Wed, Oct 15, 2003 at 03:03:47PM -0400, Joey Hess wrote:
> No matter what I do, language-chooser is always the default in main-menu
> now. I suppose this is continuing libd-i upgrade breakage.
Just an update: There is no blank line after tha last template (and yes, it
is debian-installer/country
On Wed, Oct 15, 2003 at 03:03:47PM -0400, Joey Hess wrote:
> No matter what I do, language-chooser is always the default in main-menu
> now. I suppose this is continuing libd-i upgrade breakage.
This is probably caused by
22:53 < Sesse> Oct 15 15:36:28 (none) user.debug frontend: --> SET
On Thu, Oct 16, 2003 at 02:43:41PM -0400, Joey Hess wrote:
> Debian-cd procudes CDs for woody which contain udebs.
s/woody/sarge/
> In that case, busybox-cvs-udeb was incorrectly included on the cd
s/busybox-cvs-udeb/busybox-udeb/
Apart from that, I fully agree with the bug report.
/* Steinar
On Thu, Oct 16, 2003 at 01:12:26PM -0700, Chris Tillman wrote:
> Yes, this part has always been slow, since debootstrap was invented.
> It is running the 'pkgdetails' script for each package in Packages. I
> looked once to see if some progress could be implemented, but it's
> subsidiary to debootst
On Mon, Nov 03, 2003 at 11:24:23AM -0500, Joey Hess wrote:
> * Progress bar when running mkfs.
FWIW, this has been fixed in autopartkit for some time now.
> CD installs to i386 and (some) powerpc are working, with many
> successful installs (and many bugs).
> Some success with
On Mon, Nov 03, 2003 at 06:48:34PM -0500, Joey Hess wrote:
> Do we have any languages that are say 90% done for d-i yet? I've tried
> various languages and they always seem closer to the 20% mark, but maybe
> I have been trying the wrong ones.
Norwegian (bokmål) is rather good, at least it was the
On Sat, Nov 08, 2003 at 12:05:02PM -0200, Andre Luis Lopes wrote:
> - netcfg
> Working, pending rewrite for IPv6 and multiple devices
> support.
>
> is now half done ? I mean, only IPv6 support is lacking now ?
>
> Now *that* would be cool :-)
Looks like it -- I have no clue, I'm just helpin
Package: autopartkit
Severity: normal
Tags: d-i
If autopartkit can't find any drives to partition (say, a SCSI-only
system where the SCSI card was not properly detected), it dies with the
following (rather confusing) error message:
Could not stat device autopartkit/device_name doesn't exist - N
On Sun, Nov 09, 2003 at 06:27:47PM +0100, Denis Barbier wrote:
> Recently cdebconf behavior has been changed to not display already seen
> questions, in order to help non-interactive installations.
> But this change makes back up quite useless, because one cannot go
> back and forth without trouble
On Sun, Nov 09, 2003 at 06:37:25PM +0100, Karsten Merker wrote:
> during the "install base system" step in d-i, i.e. while debootstrap
> is running, there is one annoying effect. The newt frontend redraws
> the whole status box (the progress bar and the text about which file
> is retrieved respecti
On Sun, Nov 09, 2003 at 07:08:49PM -0500, Joey Hess wrote:
> I've just sent a release announcement to debian-boot. I am actually
> still waiting for tomorrow's dinstall to get the official non-cd images
> updated in the debian archive, but everything else is in place for the
> beta now. I've learne
On Sun, Nov 09, 2003 at 06:46:01PM +0100, Steinar H. Gunderson wrote:
>> during the "install base system" step in d-i, i.e. while debootstrap
>> is running, there is one annoying effect. The newt frontend redraws
>> the whole status box (the progress bar and the
Package: hw-detect
Severity: important
Tags: d-i
This has been discussed on IRC earlier, but deserves a bug report so we
won't forget it before it's fixed :-)
If one first runs a module that fails (for instance, install LILO on a
non-existant drive, or install base from network with cable unplugg
On Mon, Nov 10, 2003 at 02:46:00AM +0100, Steinar H. Gunderson wrote:
> We should probably fix the web pages so
> - The errata is a little more complete (most notably "you can't select mirror
> at priority=high, and manual mirror selection is broken")
Add to that "m
Package: anna
Severity: minor
File: anna
Tags: d-i
(I was quite sure I had reported this before, but can't find it.
Apologies if this is a dupe. :-) )
In the list of modules to fetch, anna seems to cut off the last
character from some of the descriptions. This might be a bit confusing
to the user
On Mon, Nov 10, 2003 at 03:49:51AM +0100, Jeremie Koenig wrote:
> Thanks, i didn't know debconf was offering this possibility. Here is the
> new version :
Applied in CVS, thanks.
(Please send your patches as attachment in the future; it makes them a lot
easier to apply.)
/* Steinar */
--
Homepa
On Mon, Nov 10, 2003 at 04:00:12AM +0100, Steinar H. Gunderson wrote:
> In the list of modules to fetch, anna seems to cut off the last
> character from some of the descriptions. This might be a bit confusing
> to the user. :-)
Actually this seems to be a libd-i bug. src/package_parser.
On Sun, Nov 09, 2003 at 10:45:54PM -0500, Joey Hess wrote:
>> If one first runs a module that fails (for instance, install LILO on a
>> non-existant drive, or install base from network with cable unplugged)
>> and then run hw-detect, weird stuff happens:
> Can you provide an easier way to reproduce
On Mon, Nov 10, 2003 at 07:43:09PM +0100, Jeremie Koenig wrote:
> BTW, anna does :
>
> debconf_fset(debconf, ANNA_CHOOSE_MODULES, "seen", "false");
>
> before the question is asked. I think there are other instances of this.
> Maybe this sould be moved after we input the question ?
Definitel
A quick test indicates that most interesting mkfs-variations (ext2, ext3,
reiserfs; vfat and jfs are so quick it doesn't need progress, and xfs seems
only a bit slower than that) has at least some kind of progress output, so we
_could_ theoretically grok that. It will be ugly, though...
libparted1
What is the current status on alpha? (The ports page say that the images are
bootable, but not much more.) I have access to an AS2100 (SRM-based, in other
words) and could probably help testing/fixing bugs if the Alpha guys need it
to get on the beta 2 release.
/* Steinar */
--
Homepage: http://w
Could somebody add the URL for the daily-built images
(http://people.debian.org/~sjogren/d-i/images/daily/) back to the website?
They seem to have been removed, and they are useful for people who want to
test the d-i beta via netboot or floppy.
Also, the last built images are from the 9th; isn't t
Wasn't this bug fixed in autopartkit 0.64?
/* Steinar */
--
Homepage: http://www.sesse.net/
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
On Tue, Nov 11, 2003 at 03:30:30PM +0100, Gaudenz Steinlin wrote:
>> It would definitely help compared to what we have today, but it would still
>> suck.
> Yes, but I don't think it's worth to invest more time in that now. This
> would be an easy and acceptable solution.
Well, it's not a matter of
On Tue, Nov 11, 2003 at 11:03:53AM +0100, Gaudenz Steinlin wrote:
> This depends on the size of the filesystem and if dma is enabled or not.
> enabling DMA speeds up filesystem creation a lot on big disks. With DMA
> disabled it can take several minutes to create a big filesystem.
It can, even wit
On Mon, Nov 10, 2003 at 10:08:43PM -0500, Joey Hess wrote:
> In my tests it looks like cdebconf writes the database quite frequently
> (suprisingly frequently, really), so getting the full db should not be a
> problem.
To make cdebconf write the database, kill the frontend with SIGUSR1.
main-menu
1 - 100 of 192 matches
Mail list logo
