Re: debian-installer/libdebian-installer/debian changelog,1.58,1.59 rules,1.15,1.16
On Wed, Dec 10, 2003 at 06:50:52AM +0100, Bastian Blank wrote:
On Tue, Dec 09, 2003 at 11:26:15PM +0100, Thiemo Seufer wrote:
It's not. debian/rules is a makefile, and makefile variables
aren't handed down to the environment of a subshell.
it is, if not either your shell or your make is horrible broken.
There is no harm in having
env CFLAGS=${CFLAGS} ./configure
it does not break any ( if it does, please elaborate )
If the porter has a horrible broken shell or make,
he still wants to port the debian-installer.
Please make that possible.
bastian
Geert Stappers
--
Hailing frequencies open, Captain.
Engage the discussion
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: d-i calender
On Mon, Dec 08, 2003 at 06:36:53PM -0500, Joey Hess wrote: 12/08 cvs reopens 12/14 string freeze begins 12/21 string freeze ends 12/28 beta 2 release At least that's my current thinking. Err, i just hope that somewhen along those lines, the new powerpc packages get frozen out of the NEW queue. I understand that there is little chance of this happening now, since the ftp-master are busy with the intrusion or whatever, but it has now been over a month that i have not have any feedback on this from the ftp-masters, apart from the indirect feedback i got from you. Friendly, Sven Luther -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: architectures status
On Tue, Dec 09, 2003 at 04:53:10PM -0500, Joey Hess wrote: If you're involved in a d-i port, please write back and tell me what the current status is (boots to main menu, some successful installs, whatever), and what you expect the status to be in approximatly 2 weeks. Powerpc on non pmac architectures is lost in the NEW noman's land and not likely to be fixed anytime soon until the ftp-master decide to at least have a look at the new powerpc packages and provide some feedback to me about why they are not accepting them. I know there has been the intrusion and all that, but letting someone sit in the dark like that is not a correct way of handling your fellow debian developpers. Friendly, Sven Luther -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Installation report on CDROM-less machine
Hi, I've just attempted a debian-installer installation on an i386 system with no CDROM drive, using Etherboot. I admittedly used an outdated version of Debian installer (the beta 1 announced about a month ago), so I apologise if some or all of these problems have already been noticed and fixed. The machine had a standard IDE hard drive, NSC Geode CPU, 128MB RAM, and Realtek 8139 networking. - Throughout the installer, the line-drawing characters looked a bit messed up. (using ^ for horizontal lines, for corner bits) - The keyboard layout selection is confusing. It first asked me for a keymap for a USB keyboard - but I wasn't using a USB keyboard, and all of the options it gave me were for the Mac, and for some reason it defaulted to British English. It then took me back to the menu. After selecting Select a Keyboard Layout /again/, it asked me whether I was using a PC or USB keyboard, defaulting to USB; I chose PC. Why was it necessary to ask this, and why was I prompted to choose a keyboard layout twice? Finally, it gave me a selection for a PC-style keyboard layouts. - It didn't succeed in loading all the modules it wanted to; the culprit appears to be ide-cd. - There was no way to specify where it downloaded the debian installer modules from, or to tell it to use a proxy server. - It kept wanting to find a CD drive; there wasn't one, so it couldn't. Selecting No when it asked whether you wanted to configure a CD did nothing. Selecting Yes and then No when it asked about non-IDE non-SCSI CDs took me back to the menu. Surely a network installation should be able to work properly without a CD? - After choosing to partition the drive, I got the modules error again. Other than that, partitioning with cfdisk seemed to work. Mounting filesystems likewise seemed to work. I wasn't told exactly which partitions were going to be mkfs'ed in the final warning - it would be nice to have that final confirmation list the partitions it was going to wipe. - The keyboard navigation for Go Back/Yes/No choices was a bit weird - it seemed as though the installer thought that Go Back was on the right hand side, not the left. Cheers, Cameron. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: d-i calender
Hi Am Mit, den 10.12.2003 schrieb Sven Luther um 11:18: On Mon, Dec 08, 2003 at 06:36:53PM -0500, Joey Hess wrote: 12/08 cvs reopens 12/14 string freeze begins 12/21 string freeze ends 12/28 beta 2 release At least that's my current thinking. Err, i just hope that somewhen along those lines, the new powerpc packages get frozen out of the NEW queue. [Sven talks of the new powerpc kernel packages here] I understand that there is little chance of this happening now, since the ftp-master are busy with the intrusion or whatever, but it has now been over a month that i have not have any feedback on this from the ftp-masters, apart from the indirect feedback i got from you. As the dust around the intrusion has settled a bit, I ask all of you to clear this issue. What I observe from the outside is Sven periodically ranting about his packages not being accepted by the ftpmasters and not being given any feedback and that the packages are not in the archive by now. I don't know what's the real reason for this. The non-acceptance of these packages holds up the d-i development for several powerpc subarches (namely oldworld macintosh and pegasos). Also the integration of the powerpc kernels into the linux-kernel-di framework is stalled. I'm using svens new kernels on my machine for several weeks now without any bigger problems. I know they are not perfect, but now package is perfect on the first upload after a major change. I ask you to provid some feedback on this issue or to accept these new kernel packages into the archive as soon as is reasonably possible. Thanks Gaudenz -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: discover-udeb: Depends: libdiscover1 but it is not installable
On Tue, Dec 09, 2003 at 07:39:07PM -0500, Joey Hess wrote: Geert Stappers wrote: libdiscover1 is installed on that build computer. Doing `grep -R libdiscover1 *` didn't bring up a usefull hint. What other things do I have to hunt down that dependency. This must be a consequence of Goswin's grody hack in get-packages: [ get-packages fragment that modifies APT status file ] So even though you have it installed, apt is coaxed to not know it is installed. Goswin and I discussed this for hours, I think it is 100% useless to dummy up a status file, and I never got an answer that was satesfactory to me about why it's needed. I missed that discussion. So I missed also the _why_ Try either adding libdiscover1 to the list, or changing the APT_GET to not use the dummy status file. I've done the latter in my own tree, for weeks. Besides adding libdiscover1, also discover-data was needed. Removing the status file tweaking stuff did also work and much cleaner. Please apply attached getpackages.patch The udeb pkg-list was incomplete, see baselist.patch. -- see shy jo Geert Stappers Index: build/get-packages === RCS file: /cvs/debian-boot/debian-installer/build/get-packages,v retrieving revision 1.6 diff -u -r1.6 get-packages --- build/get-packages 17 Nov 2003 12:20:42 - 1.6 +++ build/get-packages 10 Dec 2003 12:20:21 - @@ -59,23 +59,12 @@ APT_GET=apt-get --assume-yes \ -o Dir::Etc::sourcelist=`pwd`/$LIST \ -o Dir::State=`pwd`/$APTDIR/state \ - -o Dir::State::Status=`pwd`/$APTDIR/state/status \ -o Debug::NoLocking=true \ -o Dir::Cache=`pwd`/$APTDIR/cache # Prepare APTDIR mkdir -p $APTDIR/state/lists/partial mkdir -p $APTDIR/cache/archives/partial -# Prime status file with system libraries - -echo -n $APTDIR/state/status -if [ $TYPE = udeb ]; then - # Some archs have libc6 others have libc6.1 - dpkg -s libc6 $APTDIR/state/status || dpkg -s libc6.1 $APTDIR/state/status - for i in libnewt0.51 libdebconfclient0 libdebian-installer4 \ - libdb1-compat slang1a-utf8; do - dpkg -s $i $APTDIR/state/status - done -fi # Update package lists and autoclean cache if [ $ONLINE = y ]; then Index: build/pkg-lists/base === RCS file: /cvs/debian-boot/debian-installer/build/pkg-lists/base,v retrieving revision 1.18 diff -u -r1.18 base --- build/pkg-lists/base22 Oct 2003 00:44:49 - 1.18 +++ build/pkg-lists/base10 Dec 2003 12:21:30 - @@ -7,3 +7,7 @@ anna di-utils-shell di-utils-reboot +# GSt dropped these here +archdetect +libc-udeb +slang1a-utf8-udeb
Re: discover-udeb: Depends: libdiscover1 but it is not installable
[EMAIL PROTECTED] (Geert Stappers) writes: On Tue, Dec 09, 2003 at 07:39:07PM -0500, Joey Hess wrote: Geert Stappers wrote: libdiscover1 is installed on that build computer. Doing `grep -R libdiscover1 *` didn't bring up a usefull hint. What other things do I have to hunt down that dependency. This must be a consequence of Goswin's grody hack in get-packages: [ get-packages fragment that modifies APT status file ] So even though you have it installed, apt is coaxed to not know it is installed. Goswin and I discussed this for hours, I think it is 100% useless to dummy up a status file, and I never got an answer that was satesfactory to me about why it's needed. I missed that discussion. So I missed also the _why_ Either way breaks for some. You have to little packages installed, I have to many. Turning the status file off will only work for udebs and then only for most cases. And its only udebs with broken depends that make the problems with the empty status file. The problem you see is secondary: The right[tm] way for mklibs is to not use system libraries in which case all libs must be provided by udebs or debs that get installed into the tmp/tree driectories. As a favour to joey I didn't use the debs (so he doesn't have to download the huge Packages.gz file over modem on every update) but pulled in the needed library infos from the system. Shouldn't have done that seeing all the trouble its causing. One thing that will completly break down without an empty status file is building d-i images with an older libc than your systems (and what some essential packages have a minimum version). Installing the older libc6-pic file into the ramdisk for mklibs will cause apt to remove essential packages (not realy since they aren't installed in the initrd) and break the build. But even with the right glibc the downloading of debs for the d-i cdrom images (not debian-cd build ones) will break down if any installed deb conflicts with the base debs, i.e. i you don't have the default MTA installed (I still have exim3 and not exim4 for example). Try either adding libdiscover1 to the list, or changing the APT_GET to not use the dummy status file. I've done the latter in my own tree, for weeks. Besides adding libdiscover1, also discover-data was needed. There is no dicover-data. The right package is discover-data-udeb as it stands. If the archve scripts could get fixed to allow debs and udebs of equal names that could be avoided. There is also no libdiscover1 udeb. Nothing may depend on libdiscover1 strictly speaking. Doing so is just plain broken, sorry. The discover-udeb won't be installable over the net for this reason. It will only work in the initrd itself. You need an libdiscover1 udeb. period. Removing the status file tweaking stuff did also work and much cleaner. Please apply attached getpackages.patch Don't. Fix the broken udebs, fix the archive scripts to allow equaly named packages for debs and udebs to get all the shlibs bugs fixed. Package up the missing libraries as udebs or convince Joey to allow me using debs. The udeb pkg-list was incomplete, see baselist.patch. -- see shy jo Geert Stappers MfG Goswin -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Why builds linux-kernel-di for multiple architectures?
Thiemo Seufer wrote: dpkg-buildpackage ran gen-control, made a huge control file with packages of all architectures mentioned in it. Yes, the control file must mention all binaries or the archive maintenance scripts will get confused. It eventually tried to build a i386 package on my mips machine (and failed of course, because there is no i386 kernel installed). I have never seen anything like this happen. debian/rules uses the -s flag to debhelper to make it only build packages in the same architecture as the host system. Show me a build log.. -- see shy jo signature.asc Description: Digital signature
Re: debian-installer/kernel/linux-kernel-di/debian changelog,1.40,1.41
Thiemo Seufer wrote: Joey Hess wrote: Please be careful to add a new changelog entry if the current entry is released. I've fixed this. Sorry, I didn't knew that. Does this apply for every changelog which hasn't the UNRELEASED distribution tag? If yes, I've probably broken about of a dozen changelogs in CVS. Yes, the whole tree. This is the only one I noticed, most of the rest of your commits happened near to when I was releasing pending versions of stuff, and went properly into those releases. -- see shy jo signature.asc Description: Digital signature
Re: Installation report on CDROM-less machine
Cameron Patrick wrote: - Throughout the installer, the line-drawing characters looked a bit messed up. (using ^ for horizontal lines, for corner bits) Lack of proper frame buffer, bterm, and fonts stuff on netboot. This has been fixed in the dailys. Although the underlying issue of why things look that way w/o a frame buffer has still not been resolved. - The keyboard layout selection is confusing. It first asked me for a keymap for a USB keyboard - but I wasn't using a USB keyboard, and all of the options it gave me were for the Mac, and for some reason it defaulted to British English. It then took me back to the menu. After selecting Select a Keyboard Layout /again/, it asked me whether I was using a PC or USB keyboard, defaulting to USB; I chose PC. Why was it necessary to ask this, and why was I prompted to choose a keyboard layout twice? Finally, it gave me a selection for a PC-style keyboard layouts. This is at least partly fixed. - It didn't succeed in loading all the modules it wanted to; the culprit appears to be ide-cd. Not really a problem, but in the dailys it is less annoying about this sort of error. - There was no way to specify where it downloaded the debian installer modules from, or to tell it to use a proxy server. This is a new problem to me, AFAIK the Choose a mirror menu item should do that. - It kept wanting to find a CD drive; there wasn't one, so it couldn't. Selecting No when it asked whether you wanted to configure a CD did nothing. Selecting Yes and then No when it asked about non-IDE non-SCSI CDs took me back to the menu. Surely a network installation should be able to work properly without a CD? Yeah, this is annoying for all non-CD installs, though you can work around it. - After choosing to partition the drive, I got the modules error again. Other than that, partitioning with cfdisk seemed to work. Mounting filesystems likewise seemed to work. I wasn't told exactly which partitions were going to be mkfs'ed in the final warning - it would be nice to have that final confirmation list the partitions it was going to wipe. Good idea.. - The keyboard navigation for Go Back/Yes/No choices was a bit weird - it seemed as though the installer thought that Go Back was on the right hand side, not the left. I think this has just been fixed, and will be in the next set of daily builds. -- see shy jo signature.asc Description: Digital signature
Re: discover-udeb: Depends: libdiscover1 but it is not installable
Goswin von Brederlow wrote: One thing that will completly break down without an empty status file is building d-i images with an older libc than your systems (and what some essential packages have a minimum version). Installing the older libc6-pic file into the ramdisk for mklibs will cause apt to remove essential packages (not realy since they aren't installed in the initrd) and break the build. But even with the right glibc the downloading of debs for the d-i cdrom images (not debian-cd build ones) will break down if any installed deb conflicts with the base debs, i.e. i you don't have the default MTA installed (I still have exim3 and not exim4 for example). Please keep debs and udebs separate when discussing this. I have never objected to using a dummy status file when it is operating on debs. The libc version mismatch stuff always causes the build to fail in a way we're well familiar with, and is easy to correct when it happens. Don't. Fix the broken udebs, fix the archive scripts to allow equaly named packages for debs and udebs to get all the shlibs bugs fixed. Package up the missing libraries as udebs or convince Joey to allow me using debs. See, this is a lot of work, and we want to have a releasable installer in two weeks.. -- see shy jo signature.asc Description: Digital signature
Re: discover-udeb: Depends: libdiscover1 but it is not installable
Geert Stappers wrote: Removing the status file tweaking stuff did also work and much cleaner. Please apply attached getpackages.patch That patch is not quite correct, it should use a dummy status file for debs only. Index: build/pkg-lists/base === RCS file: /cvs/debian-boot/debian-installer/build/pkg-lists/base,v retrieving revision 1.18 diff -u -r1.18 base --- build/pkg-lists/base 22 Oct 2003 00:44:49 - 1.18 +++ build/pkg-lists/base 10 Dec 2003 12:21:30 - @@ -7,3 +7,7 @@ anna di-utils-shell di-utils-reboot +# GSt dropped these here +archdetect +libc-udeb +slang1a-utf8-udeb We surely do not need archdetect on all arches, and the library udebs, when available, will be pulled in by the other udebs that need them. -- see shy jo signature.asc Description: Digital signature
Re: debian-installer/libdebian-installer/debian changelog,1.58,1.59 rules,1.15,1.16
Geert Stappers wrote:
On Wed, Dec 10, 2003 at 06:50:52AM +0100, Bastian Blank wrote:
On Tue, Dec 09, 2003 at 11:26:15PM +0100, Thiemo Seufer wrote:
It's not. debian/rules is a makefile, and makefile variables
aren't handed down to the environment of a subshell.
it is, if not either your shell or your make is horrible broken.
Actually, my test of the shell/make behaviour was broken. I'll revert
the other instances of this change. Thanks for the heads-up.
There is no harm in having
env CFLAGS=${CFLAGS} ./configure
it does not break any ( if it does, please elaborate )
It doesn't break anything, but it's actually not needed.
If the porter has a horrible broken shell or make,
he still wants to port the debian-installer.
Please make that possible.
Thanks for your support. :-)
Thiemo
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Why builds linux-kernel-di for multiple architectures?
Joey Hess wrote: Thiemo Seufer wrote: dpkg-buildpackage ran gen-control, made a huge control file with packages of all architectures mentioned in it. Yes, the control file must mention all binaries or the archive maintenance scripts will get confused. It eventually tried to build a i386 package on my mips machine (and failed of course, because there is no i386 kernel installed). I have never seen anything like this happen. debian/rules uses the -s flag to debhelper to make it only build packages in the same architecture as the host system. Show me a build log.. It failed again, after I uncommented the mips/mipsel entries in kernel-versions. Thiemo dpkg-buildpackage: source package is linux-kernel-di dpkg-buildpackage: source version is 0.18 dpkg-buildpackage: source maintainer is Joey Hess [EMAIL PROTECTED] fakeroot debian/rules clean ./make-links ./gen-control debian/control dh_testdir dh_clean `find modules -type l` dpkg-source -ICVS -I.svn -b linux-kernel-di dpkg-source: warning: missing information for output field Standards-Version dpkg-source: warning: source directory `./linux-kernel-di' is not sourcepackage-upstreamversion `linux-kernel-di-0.18' dpkg-source: building linux-kernel-di in linux-kernel-di_0.18.tar.gz dpkg-source: building linux-kernel-di in linux-kernel-di_0.18.dsc dpkg-genchanges -S dpkg-genchanges: including full source code in upload fakeroot debian/rules clean dh_testdir dh_clean `find modules -type l` dpkg-buildpackage: source only upload: Debian-native package ** Warning: no SOURCEDIR for arch i386 dpkg-buildpackage: source package is linux-kernel-di dpkg-buildpackage: source version is 0.18 dpkg-buildpackage: source maintainer is Joey Hess [EMAIL PROTECTED] dpkg-buildpackage: host architecture is i386 fakeroot debian/rules clean make: Entering directory `/home/ica2_ts/devel/debian-installer/kernel/linux-kernel-di' dh_testdir dh_clean `find modules -type l` make: Leaving directory `/home/ica2_ts/devel/debian-installer/kernel/linux-kernel-di' debian/rules build make: Entering directory `/home/ica2_ts/devel/debian-installer/kernel/linux-kernel-di' ./make-links dh_testdir make: Leaving directory `/home/ica2_ts/devel/debian-installer/kernel/linux-kernel-di' fakeroot debian/rules binary make: Entering directory `/home/ica2_ts/devel/debian-installer/kernel/linux-kernel-di' ./make-links dh_testdir dh_testdir dh_clean -k ./install-files install -D -m 644 /boot/vmlinuz-2.4.22-1-386 debian/kernel-image-2.4.22-1-386-di/boot/vmlinuz install: cannot stat `/boot/vmlinuz-2.4.22-1-386': No such file or directory command exited with status 1 make: *** [binary-arch] Error 9 make: Leaving directory `/home/ica2_ts/devel/debian-installer/kernel/linux-kernel-di' -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Timeline for restoration of test install images?
Hello, I realize that the cd images are offline and will be until the recovery from the break-in incident is complete, but can anyone either: a) give me a vague guess of how long that will be? OR b) give me a URL where I can find a powerpc image for the installer? I have a new system here that I would like to install Debian on, and I figure that I might as well test out the new installer while I do it. It isn't critical, so I can afford to wait, but I can't wait too long. Thank you for your help, Evan Jones -- Evan Jones: http://www.eng.uwaterloo.ca/~ejones/ Computers are useless. They can only give answers - Pablo Picasso -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Timeline for restoration of test install images?
* Evan Jones [EMAIL PROTECTED] [2003-12-10 19:17]: | Hello, | | I realize that the cd images are offline and will be until the recovery | from the break-in incident is complete, but can anyone either: | | a) give me a vague guess of how long that will be? well, as for the powerpc images, I'm waiting for one of the debian powerpc machines to restart the daily-image generating process. | b) give me a URL where I can find a powerpc image for the installer? http://www.soziologie.ch/~steinlin/d-i/powerpc/ | I have a new system here that I would like to install Debian on, and I | figure that I might as well test out the new installer while I do it. | It isn't critical, so I can afford to wait, but I can't wait too long. which architecture you are needing? Maybe we can create such a cdrom by hand? Bye Thorsten signature.asc Description: Digital signature
Hungarian l10n of d-i
Hi! Attached is a full complement of Hungarian localizations of d-i PO files, based on a debian-installer.tar.bz2 snapshot someone floated on this list a few days ago. AIUI, this is not the full set of files necessary to enjoy a finished d-i in all its native glory, other udebs must also me l10n-ed (specifically, those mentioned in the debian-installer/doc/translations.txt Second stage section). Will someone please commit this to CVS and set me straight as to where to obtain those additional PO files external to d-i. I have no familiarity with d-i code at all, and am no DD, so verbose replies are welcome. Please advise. Cheers, Istvan PS: I read the list via the web archive, which seems to be down right now, so CC me on the replies. d-i-hu20031210.tar.gz Description: Binary data
beta install i386
Does anyone happen to have a url where I could download the i386 beta installer? I tried the below website with i386, 386 or intel, but nothing. | b) give me a URL where I can find a powerpc image for the installer? http://www.soziologie.ch/~steinlin/d-i/powerpc/ Thanks, Keith. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
RE: beta install i386
Does anyone happen to have a url where I could download the i386 beta installer? I tried the below website with i386, 386 or intel, but nothing. | b) give me a URL where I can find a powerpc image for the installer? http://www.soziologie.ch/~steinlin/d-i/powerpc/ Found on: http://www.mmweg.rwth-aachen.de/~sebastian.ley/d-i/ -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: debian-installer/kernel/linux-kernel-di/debian changelog,1.40,1.41
Joey Hess wrote: Thiemo Seufer wrote: Joey Hess wrote: Please be careful to add a new changelog entry if the current entry is released. I've fixed this. Sorry, I didn't knew that. Does this apply for every changelog which hasn't the UNRELEASED distribution tag? If yes, I've probably broken about of a dozen changelogs in CVS. Yes, the whole tree. This is the only one I noticed, most of the rest of your commits happened near to when I was releasing pending versions of stuff, and went properly into those releases. I've fixed the remaining ones now. Thiemo -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Hungarian l10n of d-i
On Wed, 2003-12-10 at 17:58, VEROK Istvan wrote: Hi! Will someone please commit this to CVS and set me straight as to where to obtain those additional PO files external to d-i. I have no familiarity with d-i code at all, and am no DD, so verbose replies are welcome. Committed to CVS. Thank you for your contribution to Debian. - Alastair McKinstry -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Hungarian l10n of d-i
Quoting VEROK Istvan ([EMAIL PROTECTED]): Hi! Attached is a full complement of Hungarian localizations of d-i PO files, based on a debian-installer.tar.bz2 snapshot someone floated on this list a few days ago. AIUI, this is not the full set of files necessary to enjoy a finished d-i in all its native glory, other udebs must also me l10n-ed (specifically, those mentioned in the debian-installer/doc/translations.txt Second stage section). A few changes are likely to occur to these ones as some bugs are still opened for debconf polishing. Will someone please commit this to CVS and set me straight as to where to obtain those additional PO files external to d-i. I have no familiarity with d-i code at all, and am no DD, so verbose replies are welcome. When all servers will be up again : http://www.debian.org/intl/l10n/po-debconf/hu Then get the templates.pot file for : base-config shadow (changes pending) exim4-config console-data discover tasksel Seems that base-config is already partially translated as you will see on the above page. As people.debian.org is still not OK, the only way you have is : Put appropriate deb-src entries in your /etc/apt/sources.list file : deb-src http://ftp.debian.org/debian/ sarge main non-free contrib Then, after apt-get update : $ apt-get source package Then go into package-x.y/debian/po get templates.pot rename it to hu.po edit and send it back to the maintainer as a bug report (wishlist type) For base-config, you should find an existing debian/po/hu.po file which you need to update -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: debian-installer/main-menu/debian/po ar.po,NONE,1.1
On Wed, Dec 10, 2003 at 10:18:30PM +, Alastair McKinstry wrote: [...] Project-Id-Version: PACKAGE VERSION\n Report-Msgid-Bugs-To: \n POT-Creation-Date: 2003-11-18 19:22+0100\n PO-Revision-Date: YEAR-MO-DA HO:MI+ZONE\n Last-Translator: FULL NAME [EMAIL PROTECTED]\n Language-Team: LANGUAGE [EMAIL PROTECTED]\n MIME-Version: 1.0\n Content-Type: text/plain; charset=CHARSET\n Content-Transfer-Encoding: 8bit\n Alastair, I set charset to UTF-8, but could you please ask translator to fill up at least Last-Translator and Language-Team (if there is one) fields? Denis -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Why builds linux-kernel-di for multiple architectures?
Joey Hess wrote: Thiemo Seufer wrote: ./install-files install -D -m 644 /boot/vmlinuz-2.4.22-1-386 debian/kernel-image-2.4.22-1-386-di/boot/vmlinuz install: cannot stat `/boot/vmlinuz-2.4.22-1-386': No such file or directory command exited with status 1 make: *** [binary-arch] Error 9 make: Leaving directory `/home/ica2_ts/devel/debian-installer/kernel/linux-kernel-di' install-files looks up the host architecture with dpkg-architecture -qDEB_HOST_ARCH, reads kernel-versions and skips all other architectures, and copys kernel images for the host architecture. I cannot imagine how the above could happen unless dpkg-architecture returned i386. Found. It was a bogosity in my local build script. Nevertheless, I improved install-files and find-dups a bit (probably not the real perl way, but it works). Thiemo -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Installation report on CDROM-less machine
On Wed, Dec 10, 2003 at 08:00:26PM +0800, Cameron Patrick wrote: Hi, I've just attempted a debian-installer installation on an i386 system with no CDROM drive, using Etherboot. I have tried to do that but was unsuccessful. Could you describe how you got it working (what files you used to boot with, if you had to make a custom initrd, and the relevant section of dhcpd.conf). I have got the old boot-floppies to boot over the network with etherboot, but I could not actually install anything with it since once booted, the kernel had not access to the modules for the nic :-( The documentation for netbooting has always been poor in Debian. If I actually learn how to do it I might try to improve the documentation. -- Hans Ekbrand (http://sociologi.cjb.net) [EMAIL PROTECTED] pgp0.pgp Description: PGP signature
Re: Why builds linux-kernel-di for multiple architectures?
Thiemo Seufer wrote: Found. It was a bogosity in my local build script. Nevertheless, I improved install-files and find-dups a bit (probably not the real perl way, but it works). Good.. where can I find the debs for the 4 mips kernels so I can build this? Perhaps they are under a source package name I didn't think to look in, but I haven't found them in the pool. -- see shy jo signature.asc Description: Digital signature
Bug#223624: Installer Report - CD boot problem, IDE_SCSI module prevents IDE CDROM mounting, and inflexibility in options - but good hardware detection and working sound.
Package:installation-reports Debian-installer-version: download 21 November - http://gluck.debian.org/ cdimage/testing/netinst/i386/beta-1/sarge-i386-netinst.iso 100MB uname -a: Linux localhost 2.4.22-1-386 #9 Sat Oct 4 14:30:39 EST 2003 i686 GNU/Linux Date: 25 or 26 Nov. Method: Tried to boot CD (see comments), failed, and then used floppy image to boot the CD. Machine: Self-build. EPOX 8KHA+ Mobo (VIA KT266A, VT8366A and VT8233 onboard sound) Processor: AMD XP 1600+ Memory: 256MB DDR Root Device: IDE (IBM ATA100 HDD) : /dev/hda13 Root Size/partition table: A 5GB single partition on an existing extended partition Output of lspci: 00:00.0 Host bridge: VIA Technologies, Inc. VT8366/A/7 [Apollo KT266/A/333] 00:01.0 PCI bridge: VIA Technologies, Inc. VT8366/A/7 [Apollo KT266/A/333 AGP] 00:0d.0 Ethernet controller: 3Com Corporation 3cSOHO100-TX Hurricane (rev 30) 00:11.0 ISA bridge: VIA Technologies, Inc. VT8233 PCI to ISA Bridge 00:11.1 IDE interface: VIA Technologies, Inc. VT82C586A/B/VT82C686/A/B/VT8233/ A/C/VT8235 PIPC Bus Master IDE (rev 06) 00:11.2 USB Controller: VIA Technologies, Inc. USB (rev 1b) 00:11.3 USB Controller: VIA Technologies, Inc. USB (rev 1b) 00:11.4 USB Controller: VIA Technologies, Inc. USB (rev 1b) 00:11.5 Multimedia audio controller: VIA Technologies, Inc. VT8233/A/8235 AC97 Audio Controller (rev 30) 01:00.0 VGA compatible controller: Matrox Graphics, Inc. MGA G400 AGP (rev 04) Base System Installation Checklist: Initial boot worked: [E] Configure network HW: [O] Config network: [E] Detect CD: [E] Load installer modules: [O] Detect hard drives: [O] Partition hard drives: [ ] Create file systems: [O] Mount partitions: [O] Install base system: [O] Install boot loader: [E] Reboot: [O] [O] = OK, [E] = Error (please elaborate below), [ ] = didn't try it Comments/Problems: 1) This is the first CD that has ever failed to boot on my system. It booted a laptop OK, but my system failed completely. I've installed a range of distros on this box - Potato, Woody r0, Woody r1, Slackware 8, 9, 9.1, Mandrake 8.1, 8.2, 9 without any problems booting. I had the first few Sarge unofficial CDs from the .hu mirror and the first of these booted fine. The boot device selector floppy image on the installer CD solved the problem and I was able to start the process from the CD after booting the floppy. 2) I suffered the IDE-SCSI module problem (which I hadn't seen at the time). As I had got a few of the unofficial Sarge CDs I intended to register them as apt sources during the install, but this failed. Following the reboot it was clear that this was down to the IDE CD not being mounted. I got round it by simply removing the IDE-SCSI module and accessing the CD drive was was then fine. 3) The network configure with dhcp wouldn't let me leave it unconfigured. I have a separate Linux firewall connected by ethernet so my system is a static network with NAT on the firewall. I got around the installer throwing me back to the start of the dhcp config section by skipping ahead to network configuration without dhcp and after this it was ok. 4) I tried to avoid LILO but it refused to let me. I didn't want to re-write my mbr so put /dev/fd0 as the LILO target. It didn't like this for some reason. I then skipped ahead to the GRUB section and set this to a floppy which was OK. 5) My onboard VIA8233 sound chip has given quite a few Debian users a problem (as a quick google search will show), and the only real solution has been installing ALSA and recompiling the kernel - no big deal, but inconvenient compared to other distros which use ALSA and manage to get sound configured in the initial install. It's encouraging that the new Debian Installer managed to get sound working, although it uses the VIA 82cxxx module which isn't correct (Knoppix will happily boot my machine and also gets sound running on the VIA 82cxxx module) I haven't had a chance yet to really test the sound capabilities, but it does seem fine - a great result for a Debian installer. Conclusion : I had several problems, and can't say I particularly liked the installer. The hardware detection is certainly improved and is a very positive aspect, but the way that a couple of specific steps tried to prevent me leaving them was a negative. I doubt that a newbie would have a happy time with the installer on my box. Suggestions : I'm surprised the network config is split into 2 sections, dhcp and static. I would have thought that this would be 2 options in a single section, maybe with dial-up as a further option as a lot of Debian users will have dial-up as their only network connection. I would also make a floppy a prominent option for LILO or GRUB - if you already have a working system I think it makes more sense to use a floppy for the first couple of boots rather than overwriting the mbr during a new install. -- To UNSUBSCRIBE, email to [EMAIL
Re: Installation report on CDROM-less machine
On Wed, Dec 10, 2003 at 11:01:46PM +0100, Hans Ekbrand wrote: On Wed, Dec 10, 2003 at 08:00:26PM +0800, Cameron Patrick wrote: Hi, I've just attempted a debian-installer installation on an i386 system with no CDROM drive, using Etherboot. I have tried to do that but was unsuccessful. Could you describe how you got it working (what files you used to boot with, if you had to make a custom initrd, and the relevant section of dhcpd.conf). http://wiki.debian.net/index.cgi?DebianInstallerMknbi I have got the old boot-floppies to boot over the network with etherboot, but I could not actually install anything with it since once booted, the kernel had not access to the modules for the nic :-( The documentation for netbooting has always been poor in Debian. If I actually learn how to do it I might try to improve the documentation. That would be great. -- Hans Ekbrand (http://sociologi.cjb.net) [EMAIL PROTECTED] Geert Stappers -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Processing of choose-mirror_0.023_i386.changes
choose-mirror_0.023_i386.changes uploaded successfully to localhost along with the files: choose-mirror_0.023.dsc choose-mirror_0.023.tar.gz choose-mirror_0.023_i386.udeb Greetings, Your Debian queue daemon -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
choose-mirror_0.023_i386.changes ACCEPTED
Accepted: choose-mirror_0.023.dsc to pool/main/c/choose-mirror/choose-mirror_0.023.dsc choose-mirror_0.023.tar.gz to pool/main/c/choose-mirror/choose-mirror_0.023.tar.gz choose-mirror_0.023_i386.udeb to pool/main/c/choose-mirror/choose-mirror_0.023_i386.udeb Announcing to [EMAIL PROTECTED] Thank you for your contribution to Debian. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Why builds linux-kernel-di for multiple architectures?
Joey Hess wrote: Thiemo Seufer wrote: Found. It was a bogosity in my local build script. Nevertheless, I improved install-files and find-dups a bit (probably not the real perl way, but it works). Good.. where can I find the debs for the 4 mips kernels so I can build this? Perhaps they are under a source package name I didn't think to look in, but I haven't found them in the pool. http://ftp.debian.org/pool/main/k/kernel-patch-2.4.22-mips/ Thiemo -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Installation report on CDROM-less machine
On Wed, Dec 10, 2003 at 11:01:46PM +0100, Hans Ekbrand wrote:
| I have tried to do that but was unsuccessful. Could you describe how
| you got it working (what files you used to boot with, if you had to
| make a custom initrd, and the relevant section of dhcpd.conf).
I used the net.gz initrd from the iso I downloaded and created a netboot
image with the command below:
mknbi-linux --output=sarge-installer.img /mnt/iso/install/vmlinuz
/mnt/iso/install/net.gz --append='vga=normal ramdisk_size=8192 root=/dev/rd/0
init=/linuxrc devfs=mount,dall' --rootmode=rw
The --append= came from the command in boot.bat on the CD - I wasn't
aware of the page on the wiki.
In dhcpd.conf:
host merlin {
hardware ethernet 00:E0:4C:AC:00:79;
fixed-address merlin.home;
filename /deb/sarge-installer.img;
}
Cheers,
Cameron.
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Installation report on CDROM-less machine
On Wed, Dec 10, 2003 at 10:45:30AM -0500, Joey Hess wrote: | Cameron Patrick wrote: | - Throughout the installer, the line-drawing characters looked a bit |messed up. (using ^ for horizontal lines, for corner bits) | | Lack of proper frame buffer, bterm, and fonts stuff on netboot. This has | been fixed in the dailys. Okay. I was actually kind of glad to see that it wasn't using the frame buffer, it runs incredibly slowly on that machine (reminiscent of 9600 baud serial connections)... Where can I find these daily builds? I found the d-i ports page, but all the links under i386 were broken - the ones on gluck gave 404 Not Found errors, and the ones on people.d.o have the infinite redirect loop problem. | - It didn't succeed in loading all the modules it wanted to; the culprit |appears to be ide-cd. | | Not really a problem, but in the dailys it is less annoying about this | sort of error. You mean you've found a more descriptive error message than Something must have gone wrong? :-) | - There was no way to specify where it downloaded the debian installer |modules from, or to tell it to use a proxy server. | | This is a new problem to me, AFAIK the Choose a mirror menu item | should do that. For some reason, I got the choose a mirror menu /after/ it had downloaded the installer, so I could choose the mirror it used for downloading the 'real' .debs, but not for the installer itself. | - It kept wanting to find a CD drive; there wasn't one, so it couldn't. |Selecting No when it asked whether you wanted to configure a CD did |nothing. Selecting Yes and then No when it asked about non-IDE |non-SCSI CDs took me back to the menu. Surely a network installation |should be able to work properly without a CD? | | Yeah, this is annoying for all non-CD installs, though you can work | around it. Shouldn't it remember that you don't want to use a CD after you hit No for the first time? Cameron. P.S. Thanks for your detailed reply. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: architectures status
Hi Sven, thanks for correcting my comments. * Sven Luther [EMAIL PROTECTED] [2003-12-10 21:37]: | - NewWorld PowerPC's are well supported (Apple) | - OldWorld PowerPC's are _not_ supported (because we need a floppy boot |method here) | | There is already a oldworld kernel in the NEW queue. It is called | powerpc-small, and should fit with miboot on a floppy. a cool. this means 2.4 kernels are now working on OldWorld machines? Why we have then this stuipid d-i doesn't support oldworld discussion? | - Pegasos is supported, if the new kernels escape from NEW (Sven?) | | Yep, altough i will have to redo a build because of some RTC stuff i | have doubts about. cool. | - IBM RS/6000 machines (Chrep): should be supported, I'm working on a |special kernel build here at the moment. (direkt cdrom is maybe also |possible) | | Huh ? The normal kernel build should work, you just need the | powerpc-chrp-rs6k flavor of the kernel. There is also a non-rs6k chrp | flavor (used by the Pegasos, but also other chrps out there, i think). well, I have tested the kernels, which are currently in the archive (normal or d-i) and both of those doesn't work. The machine will *not* boot at all. (I have used the chrp-rs6k kernels). Have you tested those kernels on such machines? | - PReP machines are completly unsupported | | There is the powerpc-prep kernel, which should work. that's why I have written unsupported. Noone here has really tested the images on PReP machines. The d-i cdrom's are *not* bootable on those machines. I guess we should say here _unsupported. | For chrp, chrp-rs6k and prep, i don't really know what support is needed | apart from the kernel. The kernel with the same powerpc config should | work, but you need to adjust the boot-loader wrapper. I'm playing here with the chrp-rs6k part. | Ah, i forgot, you need to embedd the initrd for some of those, but i | have code and a package (also sitting in the NEW queue) for taking an | initrd and embedding it in the kernel. | | I am not sure of the interest of embeded initrds for pmac, but they can | be used on prep, chrp and chrp-rs6k, when not doing netboots (which can | load the initrd from the net, if i am not wrong). could you please create such a kernel, which an embedded d-i initrd for rs6-chrp machines? I would like to test it, because the only way to load the ramdisk on those machines is currently by using a second floppy disk. But this can maybe replaced by such a embedded initrd file. | (unsupported means here, you can simply download one of the d-i images | and install a system) | | Nope, this does not work, you need the kernel sitting in the NEW queue | since over a month for that to happen. Can you put theses deb's online somewhere other, so I can download it and testing the kernels on the rs6k? Bye Thorsten signature.asc Description: Digital signature
Re: Extensible partitioner based on libparted and cdebconf
Just looked at the screen shots. In http://lml.bas.bg/~anton/partman/screenshots/2.png there's an English localization error: Choose is an irregular verb, so the heading should say Please choose what to do with the chosen partition (not 'choosed', which isn't a word). While you're at it, Copy here data from another partition doesn't parse in English -- it should be Copy data to here from from another partition. I'm sure the strings could be improved in other ways, but those are the only ones I saw which actually *need* to be changed. Thanks for your impressive work! -- Nathanael Nerode neroden at gcc.gnu.org http://home.twcny.rr.com/nerode/neroden/fdl.html -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
