Bug#687212: Processed: reassign 687212 to localechooser, found 687212 in 2.44
Hi,
I attached patch to one of the previous messages for detecting network-console
installs.
Regards,
Eugeniy Meshcheryakov
--- Оригінальне повідомлення ---
Quoting Debian Bug Tracking System (ow...@bugs.debian.org):
Processing commands for cont...@bugs.debian.org:
reassign 687212 localechooser
Bug #687212 [installation-reports] localechooser: should use higher
languagelevel with network-console Bug reassigned from package
'installation-reports' to 'localechooser'. No longer marked as found
in versions 2.44 and 2.46. Ignoring request to alter fixed versions of
bug #687212 to the same values previously set
found 687212 2.44
Bug #687212 [localechooser] localechooser: should use higher
languagelevel with network-console Marked as found in versions 2.44/.
The list of languages was very short because of this code:
# Determine the display level
language_display_level() {
local level
#log Frontend in use: $DEBIAN_FRONTEND
case $DEBIAN_FRONTEND in
gtk)
level=4 ;;
*)
# Keep only Latin1 languages if we don't have a framebuffer
if [ $TERM_FRAMEBUFFER ]; then
level=3
else
level=1
fi
# The hurd text-mode console has decent charset support
if [ $TERM = hurd ]; then
level=3
fi
# ASCII only if we are on serial console, dumb, or Mach terminal
# Both variables should already be set at init time
if [ $TERM_TYPE = serial ] || [ $TERM = dumb ] || [ $TERM =
mach-color ] ; then level=0
fi
;;
esac
#log Language display level is $level
echo $level
}
If, in the situation you describe, there is a way to recognize that
the terminal can display non Latin languages, then we can adapt that
code.
--
To UNSUBSCRIBE, email to debian-boot-requ...@lists.debian.org
with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org
Archive: http://lists.debian.org/1347352257.11721.2.camel@Nokia-N900
Bug#677931: installation-report: testing image does not boot on QNAP TS-212 (armel) / stable works
Well, those pure ascii languages still use UTF-8. Look at French for example. I was using UTF-8 locale and everything was looking correctly. I think many languages in the list will not be shown correctly if one uses ssh on a latin1 console. --- Оригінальне повідомлення --- Eugeniy Meshcheryakov, le Mon 10 Sep 2012 23:42:45 +0200, a écrit : Currently localechooser detects network-console as a terminal without framebuffer and uses languagelevel 1 for such installs. It should use higher level, because most modern ssh clients support unicode, and in any case d-i is running in UTF-8 locale (not latin1). Most modern ssh clients support unicode indeed, but the current ssh daemon does not take the LANG parameter from the ssh client, so d-i can not know whether it should emit utf-8 or ascii. That is why the list is restricted to pure ascii languages. Samuel -- To UNSUBSCRIBE, email to debian-boot-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/1347354239.12015.2.camel@Nokia-N900
Bug#677931: installation-report: testing image does not boot on QNAP TS-212 (armel) / stable works
--- Оригінальне повідомлення --- Eugeniy Meshcheryakov, le Tue 11 Sep 2012 11:03:59 +0200, a écrit : Well, those pure ascii languages still use UTF-8. Look at French for example. French is not a pure ascii language, precisely. I was using UTF-8 locale and everything was looking correctly. Because you happened to use a UTF-8 locale. If you connect from a latin1 terminal, accents will go wrong. Exactly my point. I think many languages in the list will not be shown correctly if one uses ssh on a latin1 console. That's precisely the problem and the reason why we can't assume non-ascii languages will render correctly. The state now is that d-i is using UTF-8 for all languages. So non-UTF clients already have broken display, probably even for English with all those pseudographical characters. I think it is safe to assume that UTF-8 ssh client is available in 2012. The UTF-8 transition was done 6 years ago. Another clean solution would be disable all languages except C. I do not think that this is acceptable. Regards, Eugeniy Meshcheryakov -- To UNSUBSCRIBE, email to debian-boot-requ...@lists.debian.org with a subject of unsubscribe. Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/1347361744.12071.2.camel@Nokia-N900
Bug#677931: installation-report: testing image does not boot on QNAP TS-212 (armel) / stable works
10 вересня 2012 о 14:08 +0100 Martin Michlmayr написав(-ла): Yes, it requires a complete installation. This is because a) we have to bring SSH up and b) network configuration is currently in one step, so we cannot configure a basic config and later ask for more details, such as DNS. How common is it for DHCP servers not to send gateway and DNS information? I don't know about that, but in my case I had a DHCP server that could not be configured to supply next-server option (modem). So I had to disconnect that and configure isc-dhcp-server. Minimal required configuration does not supply DNS info. Confusingly it was enough to flash the image, and d-i also obtained address after reboot. And there were no indication of any problem. But using rescue mode in not official way of installation, so if it is possible to use d-i first time without problems that's not so bad. Regards, Eugeniy Meshcheryakov signature.asc Description: Digital signature
Bug#677931: installation-report: testing image does not boot on QNAP TS-212 (armel) / stable works
clone 677931 -1
reassing -1 localechooser
found -1 2.44
retitle -1 localechooser: should use higher languagelevel with network-console
tag -1 + patch
thanks
Hi,
Currently localechooser detects network-console as a terminal without
framebuffer and uses languagelevel 1 for such installs. It should use
higher level, because most modern ssh clients support unicode, and
in any case d-i is running in UTF-8 locale (not latin1).
The attached patch changes languagelevel for network-console installs to
the same value as used for framebuffer based installs.
Regards,
Eugeniy Meshcheryakov
10 вересня 2012 о 14:08 +0100 Martin Michlmayr написав(-ла):
After I connected to QNAP via ssh and downloaded installer components, I
was presented with a list of languages to choose from. The list was
quite small (screenshot: http://people.debian.org/~eugen/di-languages.png ).
For example, Ukrainian was not in the list. There were no non-latin
languages
in the list. Some languages using Latin scripts were missing too. Why?
I cannot remember the reason for this; maybe someone else can comment.
From be7383de324551218354a9863f4bdf9f54d2cd17 Mon Sep 17 00:00:00 2001
From: Eugeniy Meshcheryakov eu...@debian.org
Date: Mon, 10 Sep 2012 23:29:59 +0200
Subject: [PATCH] Use language level 3 with network-console installs
---
localechooser |3 ++-
1 file changed, 2 insertions(+), 1 deletion(-)
diff --git a/localechooser b/localechooser
index 7f5eb60..0db279d 100644
--- a/localechooser
+++ b/localechooser
@@ -62,7 +62,8 @@ language_display_level() {
level=4 ;;
*)
# Keep only Latin1 languages if we don't have a framebuffer
- if [ $TERM_FRAMEBUFFER ]; then
+ # and not using network-console
+ if [ $TERM_FRAMEBUFFER -o $SSH_CLIENT ]; then
level=3
else
level=1
--
1.7.10.4
signature.asc
Description: Digital signature
Bug#677931: installation-report: testing image does not boot on QNAP TS-212 (armel) / stable works
10 вересня 2012 о 20:17 +0100 Martin Michlmayr написав(-ла): * Eugeniy Meshcheryakov eu...@debian.org [2012-09-10 21:07]: I don't know about that, but in my case I had a DHCP server that could not be configured to supply next-server option (modem). So I had to disconnect that and configure isc-dhcp-server. Minimal required configuration does not supply DNS info. Confusingly it was enough to flash the image, and d-i also obtained address after reboot. And there were no indication of any problem. The problem is that we don't really have any mechanism at this point to show that there is a problem. The QNAP doesn't have an LCD after all. If you have any ideas on how this could be improved, let me know. The fix was as easy as adding this preseed file and rebuilding the recovery image: d-i netcfg/no_default_route boolean true After this QNAP booted successfully and I was able to login and fix the network configuration from command line. D-I does not seem to require DNS setup. Is there any reason not to use this setting for all network-console installs? I didn't found yet how to change this setting not using pressed file. Regards, Eugeniy Meshcheryakov signature.asc Description: Digital signature
Bug#677931: installation-report: testing image does not boot on QNAP TS-212 (armel) / stable works
Hi, 8 вересня 2012 о 17:16 +0100 Martin Michlmayr написав(-ла): The testing installer boots fine on my TS-219. Can you try again on your TS-212? I tried to use testing d-i again, and it went better. First I set up a DHCP server and TFTP server on my laptop and disconnected the modem that also has built-in DHCP server from the network. With this configuration QNAP obtained an IP address, downloaded d-i image, rebooted into d-i, obtained the IP address again, and did nothing for several minutes, till I rebooted it. After reboot - the same problem, I can ping QNAP, but it does not start the ssh server. So I stopped the laptop's DHCP server and connected my modem to network. After I rebooted QNAP ssh worked fine, and it took around a minute to boot. Can it be that D-I requires full network configuration, with gateway, DNS, and so on via DHCP? That the only difference I can think about. After I connected to QNAP via ssh and downloaded installer components, I was presented with a list of languages to choose from. The list was quite small (screenshot: http://people.debian.org/~eugen/di-languages.png ). For example, Ukrainian was not in the list. There were no non-latin languages in the list. Some languages using Latin scripts were missing too. Why? I tested the installer only till I was able to mount the hard drive and reflash the kernel from the working system. There were no other problems. Regards, Eugeniy Meshcheryakov signature.asc Description: Digital signature
Bug#677931: installation-report: testing image does not boot on QNAP TS-212 (armel) / stable works
Package: installation-reports Version: 2.46 Severity: normal Boot method: flash Image version: testing/stable Date: night 16.06.2012-17.06.2012 Machine: QNAP TS-212 Partitions: df -Tl will do; the raw partition table is preferred Base System Installation Checklist: [O] = OK, [E] = Error (please elaborate below), [ ] = didn't try it Initial boot: [E] Detect network card:[ ] Configure network: [ ] Detect CD: [ ] Load installer modules: [ ] Detect hard drives: [ ] Partition hard drives: [ ] Install base system:[ ] Clock/timezone setup: [ ] User/password setup:[ ] Install tasks: [ ] Install boot loader:[ ] Overall install:[ ] Comments/Problems: I was trying to install Debian on TS-212 using instructions here: http://www.cyrius.com/debian/kirkwood/qnap/ts-219/ First I tried to use installer image from testing, but the device did not boot: both disk LEDs were constantly on, there was some disk activity, but absolutely no network activity. I tried several times wating 10-20 minutes each time. I also tried to remove disks but this did not help. After this I used stable installer image with testing repository. The installation was sucessfull, with small problems: had to remove RAID devices manually, partman did not align partition correctly. -- Please make sure that the hardware-summary log file, and any other installation logs that you think would be useful are attached to this report. Please compress large files using gzip. Once you have filled out this report, mail it to sub...@bugs.debian.org. == Installer lsb-release: == DISTRIB_ID=Debian DISTRIB_DESCRIPTION=Debian GNU/Linux installer DISTRIB_RELEASE=6.0 (squeeze) - installer build 20110106+squeeze4+b1 X_INSTALLATION_MEDIUM=network-console == Installer hardware-summary: == uname -a: Linux debian 2.6.32-5-kirkwood #1 Sun May 6 16:57:51 UTC 2012 armv5tel GNU/Linux usb-list: usb-list: Bus 01 Device 01: Marvell Orion EHCI [1d6b:0002] usb-list:Level 00 Parent 00 Port 00 Class 09(hub ) Subclass 00 Protocol 01 usb-list:Manufacturer: Linux 2.6.32-5-kirkwood ehci_hcd usb-list:Interface 00: Class 09(hub ) Subclass 00 Protocol 00 Driver hub usb-list: usb-list: Bus 01 Device 02: USB2.0 Hub [05e3:0608] usb-list:Level 01 Parent 01 Port 00 Class 09(hub ) Subclass 00 Protocol 01 usb-list:Interface 00: Class 09(hub ) Subclass 00 Protocol 00 Driver hub lsmod: Module Size Used by lsmod: raid1 19757 3 lsmod: dm_mod 56891 0 lsmod: md_mod 80506 3 raid1 lsmod: jfs 157227 0 lsmod: ext4 288593 1 lsmod: jbd2 64351 1 ext4 lsmod: ext3 111204 0 lsmod: jbd37738 1 ext3 lsmod: btrfs 539502 0 lsmod: crc32c 2562 1 lsmod: libcrc32c896 1 btrfs lsmod: vfat8152 0 lsmod: fat43613 1 vfat lsmod: ext2 55323 0 lsmod: mbcache 4860 3 ext4,ext3,ext2 lsmod: sd_mod 31344 8 lsmod: crc_t10dif 1106 1 sd_mod lsmod: sata_mv24398 6 lsmod: ehci_hcd 36525 0 lsmod: evdev 6582 0 lsmod: libata137914 1 sata_mv lsmod: usbcore 123103 2 ehci_hcd lsmod: scsi_mod 124484 2 sd_mod,libata lsmod: gpio_keys 3066 0 lsmod: mv643xx_eth22558 0 lsmod: nls_base5367 4 jfs,vfat,fat,usbcore lsmod: libphy 14836 1 mv643xx_eth lsmod: inet_lro5060 1 mv643xx_eth df: Filesystem 1K-blocks Used Available Use% Mounted on df: tmpfs 12783660127776 0% /dev df: /dev/md0 1031052504120474556 52% /target df: /dev/md0 1031052504120474556 52% /dev/.static/dev df: tmpfs 12783660127776 0% /target/dev free: total used free shared buffers free: Mem: 255676 18290472772017392 free: Swap: 524212 1124 523088 free: Total: 779888 184028 595860 /proc/cmdline: console=ttyS0,115200 root=/dev/ram initrd=0xa0,0x90 ramdisk=34816 /proc/cpuinfo: Processor: Feroceon 88FR131 rev 1 (v5l) /proc/cpuinfo: BogoMIPS : 1199.30 /proc/cpuinfo: Features : swp half thumb fastmult edsp /proc/cpuinfo: CPU implementer : 0x56 /proc/cpuinfo: CPU architecture: 5TE /proc/cpuinfo: CPU variant : 0x2 /proc/cpuinfo: CPU part : 0x131 /proc/cpuinfo: CPU revision : 1 /proc/cpuinfo: /proc/cpuinfo: Hardware : QNAP TS-119/TS-219
Bug#397678: main-menu: segfaults on amd64 after Set up users and passwords step
9 листопада 2006 о 16:29 +0100 Attilio Fiandrotti написав(-ла): Eugeny, were you installing textual (NEWT) or graphical (GTK) ? We experienced a similar crash (#373253 and #396520) with the g-i on AMD64 and it would b einteresting investigating if bugs are somehow related. I was installing with newt frontend. PS: CC-ing bug reporters is sometimes a good idea, I am not subscribed to debian-boot. -- Eugeniy Meshcheryakov signature.asc Description: Digital signature
Bug#397678: main-menu: segfaults on amd64 after Set up users and passwords step
Additional information: I cannot reproduce this bug with image from http://cdimage.debian.org/cdimage/daily-builds/daily/arch-latest/amd64/iso-cd/debian-testing-amd64-businesscard.iso built on 20061102 (at least it is written on F1 screen, I downloaded it today). 8 листопада 2006 о 22:06 +0100 Eugeniy Meshcheryakov написав(-ла): Package: main-menu Severity: grave After configuring users and passwords in d-i screen starts blinking and following messages appear at console (may contain typos): process: ###): INFO: kbd-mode: setting console mode to Unicode (UTF-8) On other console, following messages appear: init: starting pid ### console /dev/vc/1: '/sbin/debian-installer' debconf: setting debconf/language to uk kernel: main-menu[###]: segfault at rip error 4 I used image downloaded from http://cdimage.debian.org/cdimage/daily-builds/daily/arch-latest/amd64/iso-cd/debian-testing-amd64-businesscard.iso Image was built on 20061106. I also tried to run installation in English and got the same results (except that language in debconf's meesage was en). -- Eugeniy Meshcheryakov signature.asc Description: Digital signature
Bug#397678: main-menu: segfaults on amd64 after Set up users and passwords step
Package: main-menu Severity: grave After configuring users and passwords in d-i screen starts blinking and following messages appear at console (may contain typos): process: ###): INFO: kbd-mode: setting console mode to Unicode (UTF-8) On other console, following messages appear: init: starting pid ### console /dev/vc/1: '/sbin/debian-installer' debconf: setting debconf/language to uk kernel: main-menu[###]: segfault at rip error 4 I used image downloaded from http://cdimage.debian.org/cdimage/daily-builds/daily/arch-latest/amd64/iso-cd/debian-testing-amd64-businesscard.iso Image was built on 20061106. I also tried to run installation in English and got the same results (except that language in debconf's meesage was en). -- System Information: Debian Release: 4.0 APT prefers unstable APT policy: (500, 'unstable') Architecture: i386 (i686) Shell: /bin/sh linked to /bin/dash Kernel: Linux 2.6.19-rc5 Locale: LANG=uk_UA.UTF-8, LC_CTYPE=uk_UA.UTF-8 (charmap=UTF-8) -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: RFC: Possible (partial) solution for device persistence issue
15 жовтня 2006 о 23:45 +0200 Geert Stappers написав(-ла): Infact I don't understand why device naming does differ. AFAIK are the d-i kernel and the installed kernel from the same build. That d-i uses different tools (busybox, libc?) then installed kernel, will have it good reason. But it is no reason that devices naming differs ... Probably because drivers are loaded in different order, this can also happen on the installed system without any change to libraries or tools. And with current changes to the kernel (parallel probing of devices) that will happen more an more often (but it may be fixed in similar way as for network interfaces, see /etc/udev/persistent-net-generator.rules). -- Eugeniy Meshcheryakov signature.asc Description: Digital signature
Bug#388501: Installation Report
27 вересня 2006 о 11:33 -0400 Michael S. Peek написав(-ла): Did I do something wrong during the installation? No, you did not. I think it is bug in udev that loads modules for those devices in different order. The same problem was fixed for network devices some time ago, similar things problably should be done for disks too. Is there a better way to do this? Michael Peek -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED] -- Eugeniy Meshcheryakov signature.asc Description: Digital signature
Bug#389460: debian-installer: netinst image is broken
Package: debian-installer Severity: grave Current netinst package is broken. Installation stops just before Detecting network hardware ... step. There is no way to run it (d-i returns to main menu). Most of translations are missing in main menu. In syslog there are many string like INFO: Falling back to the package description. I used image http://cdimage.debian.org/cdimage/daily-builds/daily/arch-latest/i386/iso-cd/debian-testing-i386-netinst.iso md5sum: a889a861790172eb053804dcd171728f version: 20060924 -- System Information: Debian Release: testing/unstable APT prefers unstable APT policy: (500, 'unstable') Architecture: i386 (i686) Shell: /bin/sh linked to /bin/dash Kernel: Linux 2.6.18-me Locale: LANG=uk_UA.UTF-8, LC_CTYPE=uk_UA.UTF-8 (charmap=UTF-8) -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#389466: gdm: runs sessions with POSIX locale after installation
Package: gdm Severity: grave Tags: patch l10n d-i This is due to changes in location where information about locale is stored. Attached patch fixes this problem. This bug may not be grave for gdm itself, but IMHO it is release-critical for d-i. -- System Information: Debian Release: testing/unstable APT prefers unstable APT policy: (500, 'unstable') Architecture: i386 (i686) Shell: /bin/sh linked to /bin/dash Kernel: Linux 2.6.18-me Locale: LANG=uk_UA.UTF-8, LC_CTYPE=uk_UA.UTF-8 (charmap=UTF-8) diff -urN debian.orig/gdm-autologin.pam debian/gdm-autologin.pam --- debian.orig/gdm-autologin.pam 2006-09-25 23:17:22.0 +0200 +++ debian/gdm-autologin.pam2006-09-25 23:19:06.0 +0200 @@ -1,6 +1,6 @@ #%PAM-1.0 auth requisite pam_nologin.so -auth requiredpam_env.so +auth requiredpam_env.so readenv=1 envfile=/etc/default/locale auth requiredpam_permit.so @include common-account sessionrequiredpam_limits.so diff -urN debian.orig/gdm.pam debian/gdm.pam --- debian.orig/gdm.pam 2006-09-25 23:17:22.0 +0200 +++ debian/gdm.pam 2006-09-25 23:18:43.0 +0200 @@ -1,6 +1,6 @@ #%PAM-1.0 auth requisite pam_nologin.so -auth requiredpam_env.so +auth requiredpam_env.so readenv=1 envfile=/etc/default/locale @include common-auth @include common-account sessionrequiredpam_limits.so
Bug#388668: cdebconf-gtk-udeb: sometimes displays wrong mouse cursor
Package: cdebconf-gtk-udeb Severity: minor In graphical d-i mouse cursor sometimes replaced with text cursor even over fields that are not text entry fields (or at least do not look so). This bug is really easy to reproduce for me. Just do normal install and cursor will switch from text to normal pointer and back. -- System Information: Debian Release: testing/unstable APT prefers unstable APT policy: (500, 'unstable') Architecture: i386 (i686) Shell: /bin/sh linked to /bin/dash Kernel: Linux 2.6.18-rc7-me Locale: LANG=uk_UA.UTF-8, LC_CTYPE=uk_UA.UTF-8 (charmap=UTF-8) -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#343304: kbd-chooser: Ukrainian keyboard does not work (1st 2nd stages)
severity 343304 normal thanks 24 серпня 2006 о 01:07 +0200 eugen написав(-ла): Also, I tried to install system with mini.iso and did not get tasksel screen, and 'ukrainian' task was not installed, and there were no console-cyrillic in installed system. But if other images run tasksel that bug will not be so important (at least for Ukrainian) because console-cyrillic will setup keyboar correctly (hopefully, but I'll retest later). I retested d-i with latest image, and confirm that console-cyrillic is installed by default. So, that bug is not 'important'. -- Eugeniy Meshcheryakov signature.asc Description: Digital signature
Bug#388671: cdebconf-gtk-udeb: should support double click and pressing enter in partitioner menu
Package: cdebconf-gtk-udeb Severity: wishlist gtk frontend does not currently support double click or pressing enter in partitioner menu. I think it should, because it supports that in other menus. -- System Information: Debian Release: testing/unstable APT prefers unstable APT policy: (500, 'unstable') Architecture: i386 (i686) Shell: /bin/sh linked to /bin/dash Kernel: Linux 2.6.18-rc7-me Locale: LANG=uk_UA.UTF-8, LC_CTYPE=uk_UA.UTF-8 (charmap=UTF-8) -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#384481: debian-installer: current netboot and mini.iso images do not run tasksel
Package: debian-installer Severity: normal During installation with (at least) current netboot and mini.iso images tasksel is not run, so it is not possible to select any tasks, and default language tasks are not installed. -- System Information: Debian Release: testing/unstable APT prefers unstable APT policy: (500, 'unstable') Architecture: i386 (i686) Shell: /bin/sh linked to /bin/dash Kernel: Linux 2.6.18-rc4-me Locale: LANG=uk_UA.UTF-8, LC_CTYPE=uk_UA.UTF-8 (charmap=UTF-8) -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#384248: installation-report: Almost successful Etch installation with di-beta3
clone 384248 -1 reassing -1 localechooser retitle -1 localechooser: Should not overwrite console-cyrillic configuration thanks 23 серпня 2006 о 10:03 +0200 Christian Perrier написав(-ла): I selected Russian key-map during installation, but looks like Russian font was not installed (or properly configured). Therefore, switching to Russian in console via right-Ctrl - (this is another inconveniency -- I hate this way of switching key maps, because I use right-Ctrl for typing Ctrl+some_key) - gave me some fancy characters instead of Russian letters on the screen. Also I couldn't read Russian text in UTF-8 encoding on console which should be UTF-8 enabled AFAIK. did you also select Russian locale during installation? Was console-cyrillic configured during installation? It should ask about switch keys with right-Alt as default (but in installed system it will be configured with Ctrl-Shift, due to bug). I tried default installation in Russian (RF) in qemu with network off, and console-cyrillic was installed. There was also dialogs displayed by console-cyrillic after tasksel were I was able to select switch keys, but those choices was later overwritten by localechooser's finish-install. That was not a problem in earlier versions of d-i because console-cyrillic was configured in second stage, after finish-install (or how it was called). But now it is confusing. localechooser should preseed console-cyrillic or do not try to change it's configuration at all. This bug is probably reproducible with all languages that have cyr=... in languagelist (Russian, Belarusian). This is on the installed system, right? For Russian, console-cyrillic should be installed. Unfortunately, we no more have people active in testing Russian in the D-I team, with the required knowledge of specificities of the language rendering. CC'ing Eugenyi Mescheryakov.can you look at this, eugen? I have various USB devices connected to my PC. Logs about some of them always appear after bor login: prompt. Looks very unprofessional although I'm used to unexpected Linux kernel messages on the console. This has nothing to do with the installer, I'm afraid. BTW, I also don't like Linux kernel messages at the console. Probably something to set in syslog stuff, which I never succeeded to find. I agree this is pretty annoying. Besides that, backspace key stops working after invalid username/password combination, and it becomes possible to move the cursor over the whole screen. Nothing to do with the installer. Please investigate it, but this is indeed very probably due to non ASCII input in the login. All the remaining does not really belong to the installer and should be reported against the relevant packages. I heard about 'script' utility before, and wanted to create a log of my initial system set up, but 'script' killed the log when I ran it the second time. In other words 'script' should not overwrite default 'typescript' file, but rather create a backup of already present one. I understand though that this is not d-i deficiency. Configuring USB ADSL modem (via eciadsl, for the first time in Linux) was a real pain for me. I had to reboot back and forth between Windows with working ADSL Internet connection and fresh installed Debian Etch without Internet connection to dig for ADSL configuration information (which turned to be very vague from newcomer's point of view) and manually download required Debian packages and theirs dependencies. I believe average non-technical PC users had no chances to overcome this major obstacle of conecting Debian to the Internet via USB ADSL modem. /etc/apt/sources.list was not updated with any Debian mirror line, (presumably due to failed network configuration). So I had to search it in apt documentation examples, and guess correct domain name of the nearest mirror. I wish 'gpm' and 'less' packages were the part of netinst CD. I wish I could configure console beep sound during installation so that my wife didn't get angry about beeping through filesystem at 2am, and I needn't look for 2 year old notes with magic escape sequnces for the beep. -- Eugeniy Meshcheryakov signature.asc Description: Digital signature
Bug#343304: kbd-chooser: Ukrainian keyboard does not work (1st 2nd stages)
retitle 343304 kbd-chooser: Ukrainian keyboard does not work (installed system) thanks 14 грудня 2005 о 11:10 +0200 Eugeniy Meshcheryakov написав(-ла): When trying to install Debian in Ukrainian language and Ukrainian keyboard layout selected it is impossible to enter Cyrillic characters in 1st stage (it should be possible to switch to cyrillic layout by pressing right ctrl or alt key). In 2nd stage I get only white boxes instead of characters when entering from keyboard (but cyrillic characters are displayed correctly by programs like base-config). Keymap file /etc/console/boottime.kmap.gz looks completely incorrect (no charset declaration, some strange compose sequences, ...). Update: it is possible to enter Cyrillic characters in installer with kbd-chooser 1.30 (it needed keyboard in unicode mode to work). After installation /etc/console/boottime.kmap.gz contains junk, so nothing changed here. Also, I tried to install system with mini.iso and did not get tasksel screen, and 'ukrainian' task was not installed, and there were no console-cyrillic in installed system. But if other images run tasksel that bug will not be so important (at least for Ukrainian) because console-cyrillic will setup keyboar correctly (hopefully, but I'll retest later). -- Eugeniy Meshcheryakov signature.asc Description: Digital signature
Bug#379566: (forw) Bug#379566: Daily build install looking at Hebrew l10n issues
reassign 379566 newt severity 379566 grave thanks This is bug in newt, it can be reproduced with attached script (heb.sh). Latest version of libnewt does not contains code for bidi support, it can be checked by greppong 'fribidi' on libnewt. The code is not compiled because it needs symbols like HAVE_FRIBIDI_FRIBIDI_H defined (see newt.h). Package even have patch that adds needed chacks to configure.in, but configure is not updated. So to fix this bug it is needed to run autoconf, as attached patch does. Severity is set to grave because this bug makes d-i unuseable with some languages. 24 липня 2006 о 14:14 +0200 Christian Perrier написав(-ла): Could you guys have a look at this? Please followup to bug #379566. - Forwarded message from Baruch Even [EMAIL PROTECTED] - Subject: Bug#379566: Daily build install looking at Hebrew l10n issues Reply-To: Baruch Even [EMAIL PROTECTED], [EMAIL PROTECTED] Date: Mon, 24 Jul 2006 11:29:20 +0100 From: Baruch Even [EMAIL PROTECTED] To: [EMAIL PROTECTED] X-Mailing-List: debian-boot@lists.debian.org archive/latest/104663 X-CRM114-Status: Good ( pR: 26.4757 ) Package: installation-reports Boot method: cdrom in a qemu image Image version: daily build from 2006-07-23 Date: 2006-07-23 23:30 Machine: qemu i386 Processor: i386 Memory: 128MB Partitions: NA Output of lspci and lspci -n: NA Base System Installation Checklist: [O] = OK, [E] = Error (please elaborate below), [ ] = didn't try it Initial boot worked:[O] Configure network HW: [O] Config network: [O] Detect CD: [O] Load installer modules: [O] Detect hard drives: [O] Partition hard drives: [O] Create file systems:[O] Mount partitions: [O] Install base system:[O] Install boot loader:[O] Reboot: [O] Comments/Problems: This was all tested on the text install, on qemu it took several hours for the install to be finished. I'll try the gui installer at another time. Initial screen of choosing the language says Hebrew, but the word in Hebrew is reversed, this used to work properly in former versions. I expected getting IVRIT and got TIRVI, the bidi algorithm wasn't used maybe? It's not just the first screen, it's in all screens. The Hebrew wasn't bidi-fied and looks reversed. There are various Language Desktop but no Hebrew Desktop, what is done for them and how I can get Hebrew Desktop there? Message of x11 upgrade on fresh install??? it also delayed the install process from continuing until input was received. GDM login screen is left aligned instead of right aligned. Gnome desktop is left aligned instead of right aligned, we have user-he that can take care of that, how do we get it to run on system install of a bidi system? Keyboard setup for X11 doesn't include scroll-lock for noticing layout changes, this was provided in a patch to localization-config. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED] - End forwarded message - -- -- Eugeniy Meshcheryakov heb.sh Description: Bourne shell script diff -urN newt-0.52.2/debian/control newt.bidi/debian/control --- newt-0.52.2/debian/control 2006-07-24 16:43:15.0 +0200 +++ newt.bidi/debian/control2006-07-24 16:44:21.0 +0200 @@ -4,7 +4,8 @@ Maintainer: Alastair McKinstry [EMAIL PROTECTED] Standards-Version: 3.7.2 XS-Python-Version: all -Build-Depends: dbs, sgmltools-lite, libslang2-dev (=2.0.4-2), libpopt-dev, debhelper (=5.0.37.1), gettext, libfribidi-dev, tcl8.4-dev, python (= 2.3.5-7), python-central (= 0.4.10), python-all-dev +Build-Depends: dbs, sgmltools-lite, libslang2-dev (=2.0.4-2), libpopt-dev, debhelper (=5.0.37.1), gettext, libfribidi-dev, tcl8.4-dev, python (= 2.3.5-7), python-central (= 0.4.10), python-all-dev, autoconf +Build-Conflicts: autoconf2.13 Package: libnewt-dev Architecture: any diff -urN newt-0.52.2/debian/rules newt.bidi/debian/rules --- newt-0.52.2/debian/rules2006-07-24 16:43:15.0 +0200 +++ newt.bidi/debian/rules 2006-07-24 16:40:23.0 +0200 @@ -70,6 +70,7 @@ dh_testdir cp newt.spec $(BUILD_TREE) rm -f $(BUILD_TREE)/.depend + ( cd $(BUILD_TREE) autoconf --force -i ) ( cd $(BUILD_TREE) ./configure $(GPMSUPPORT) $(NOSTRIP) ) touch $@ signature.asc Description: Digital signature
Bug#262531: closed by David Härdeman [EMAIL PROTECTED] (Re: LVM LVs should be mounted as /dev/VG/LV)
23 червня 2006 о 15:18 -0700 Debian Bug Tracking System написав(-ла): grub and lilo currently support either way (/dev/VG/LV or /dev/mapper/VG-LV). However, initramfs-tools and the packages which use it (e.g. cryptsetup) only support /dev/mapper/VG-LV. I'm therefore closing this bug as the /dev/mapper syntax is supported by all programs while the other syntax is not. So maybe this bug should be reassigned to those packages instead? -- Eugeniy Meshcheryakov signature.asc Description: Digital signature
Re: [g-i] new screenshots available
Hi, 22 червня 2006 о 03:01 +0200 Davide Viti написав(-ла): It would be interesting to have some feedback from native speakers; It looks good for Ukrainian (except bad glyphs for б that hopefully will be fixed soon in DejaVu). I'd be particularly interested to know how is Vietnamese displayed using Dejavu (we don't use freefont anylonger to display it) and how non-latin languages look like. Most likely it is displayed incorrectly due to bug in some package that renders those glyphs (freetype, probably). See for example how ê and ô is displayed. It is now bug #374902. -- Eugeniy Meshcheryakov signature.asc Description: Digital signature
Bug#317165: Bug no longer applies to upcoming beta2
Hello, 21 лютого 2006 о 16:53 +0100 Frans Pop написав(-ла): Hi Eugeniy, On Tuesday 21 February 2006 15:49, Eugeniy Meshcheryakov wrote: But there is one new problem: I cannot enter any Cyrillic characters in d-i. I tryed different key combinations to switch to Cyrillic layout but failed. Maybe d-i now uses different Ukrainian keyboard layout? Maybe one that uses windows keys for swither? (I have no that keys). I tried to reproduce this bug with netboot image built on 20060330 and the bug is still there. I tried to run installer in Ukrainian, Belarusian and Russian languages with default keyboard for each of those languages. D-I was ran until it tries to configure network. Then option to ocnfigure network manually was selected. Then I rtied to enter some Cyrillic text. Results: 1. For Ukrainian and Belarusian I was not able to enter any Cyrillic text because I was not able to find switcher key (all keys Ctrl, Alt, Win, Menu were tried). 2. For Russian after pressing right Ctrl key I was able to enter somthing that is not Latin - that something was displayed as XX, where X - hexadecimal digit. It probably was some single-byte encoding (i guess it was KOI8-R). 3. Ukrainian was tested with install24 with the same result. 4. /var/log/syslog says that keymap ua was selected and contains a lot of warnings unknown keysym. I assume this is with a daily built CD-based image? If it was a netboot of floppy image the reason could be there. Nothing has really changed with regard to keyboards in d-i. If there have been changes, they must have been in console-data (where d-i takes its keyboard definitions from), but I see no real changes in its changelog. The only cause I can think of in d-i is maybe a difference between the 2.4 and 2.6 kernel. There have also been some changes in localechooser that could maybe cause this. Tracing this will probably take a bit of work (unless anybody else has a better idea of the probable cause). To narrow down the cause of the problem, could you see if Cyrillic characters work with the following: - daily built image booted with install - daily built image booted with install24 - Etch Beta1 image booted with install - Sarge official image booted with linux - Sarge official image booted with linux26 For each case, please check in the syslog what exact keyboard type and keymap were selected: 'grep kbd-chooser.[0-9]*.: /var/log/syslog'. Cheers, FJP -- Eugeniy Meshcheryakov signature.asc Description: Digital signature
Bug#317165: Bug no longer applies to upcoming beta2
Hello, 16 лютого 2006 о 19:02 +0100 Christian Perrier написав(-ла): I assume that this bug report about bad behaviour in 2nd stage when the user did choose a locale which differs from the default (for the BR, Eugenyi did choose uk_UA.UTF-8) is no longer relevant as we no more have any 2nd stage Well, I cannot reproduce this bug anymore because of console-cyrillic configured in d-i. So keyboard works fine after reboot. But there is one new problem: I cannot enter any Cyrillic characters in d-i. I tryed different key combinations to switch to Cyrillic layout but failed. Maybe d-i now uses different Ukrainian keyboard layout? Maybe one that uses windows keys for swither? (I have no that keys). So I hereby suggest closing it (I'm not doing so to give other, and Eugenyi, a chance to react). -- Eugeniy Meshcheryakov signature.asc Description: Digital signature
Bug#317165: Bug no longer applies to upcoming beta2
21 лютого 2006 о 16:53 +0100 Frans Pop написав(-ла): Hi Eugeniy, On Tuesday 21 February 2006 15:49, Eugeniy Meshcheryakov wrote: But there is one new problem: I cannot enter any Cyrillic characters in d-i. I tryed different key combinations to switch to Cyrillic layout but failed. Maybe d-i now uses different Ukrainian keyboard layout? Maybe one that uses windows keys for swither? (I have no that keys). I assume this is with a daily built CD-based image? If it was a netboot of floppy image the reason could be there. debian-testing-i386-netinst.iso downloaded today. Nothing has really changed with regard to keyboards in d-i. If there have been changes, they must have been in console-data (where d-i takes its keyboard definitions from), but I see no real changes in its changelog. At least one change that I can see is that directory /usr/share/keymaps/i386/qwerty now contains a lot of Ukrainian keymaps (IIRC there were only one when I looked at it last time, but that was long time ago). The only cause I can think of in d-i is maybe a difference between the 2.4 and 2.6 kernel. There have also been some changes in localechooser that could maybe cause this. Tracing this will probably take a bit of work (unless anybody else has a better idea of the probable cause). To narrow down the cause of the problem, could you see if Cyrillic characters work with the following: - daily built image booted with install - daily built image booted with install24 - Etch Beta1 image booted with install - Sarge official image booted with linux - Sarge official image booted with linux26 I'll do this later. For each case, please check in the syslog what exact keyboard type and keymap were selected: 'grep kbd-chooser.[0-9]*.: /var/log/syslog'. Cheers, FJP -- Eugeniy Meshcheryakov signature.asc Description: Digital signature
Bug#281556: s390 installation report, installed system does not boot
Package: installation-reports INSTALL REPORT Debian-installer-version: http://ftp.de.debian.org/debian/dists/unstable/main/ installer-s390/20041027/images/generic/ uname -a: Linux debian 2.4.27-1-s390 #1 SMP Tue Aug 17 12:18:33 CEST 2004 s390 GNU/Linux Date: 2004.11.15-2004.11.16 Method: ipl from tape, used ssh after Machine: hercules, version 2.17.1-2 Processor: Memory: Root Device: IDE? SCSI? Name of device? Root Size/partition table: Feel free to paste the full partition table, with notes on which partitions are mounted where. Output of lspci and lspci -n: Base System Installation Checklist: [O] = OK, [E] = Error (please elaborate below), [ ] = didn't try it Initial boot worked:[O] Configure network HW: [O] Config network: [O] Detect CD: [ ] Load installer modules: [O] Detect hard drives: [O] Partition hard drives: [E/-] Create file systems:[O] Mount partitions: [O] Install base system:[O] Install boot loader:[O] Reboot: [E] Comments/Problems: I was not able to format dasd using d-i (known problem, should be fixed now), so I formated dasd manually and formated it with cfdisk. After successfull firts stage I rebooted hercules but I was not able to boot using zipl defaul setup. Linux was not able to find root partition (zipl.conf attached). After using google.com I found that I need to pass dasd=0120 to zipl in order to boot. That worked. It is posible that dasd parameter is missing because I did not use d-i for formating dasd and partitioning. [defaultboot] defaultmenu = menu :menu target = /boot 1 = debian 2 = old default = 1 prompt = 1 timeout = 10 [debian] target = /boot image = /boot/vmlinuz parameters = ro vmpoff=LOGOFF root=/dev/dasda1 [old] target = /boot image = /boot/vmlinuz.old parameters = ro vmpoff=LOGOFF root=/dev/dasda1 optional = 1
Bug#278953: Package: installation-reports
30.10.2004 18:48 +0200 Davi Leal (-): It would be cool offer the posibility to choose LILO besides GRUB as RedHat does. You can do this with d-i. You can install lilo from the main menu. To do this press back button in grub-installer or install in 'expert' mode. -- Eugeniy Meshcheryakov Kyiv National Taras Shevchenko University Information and Computing Centre http://icc.univ.kiev.ua signature.asc Description: Digital signature
Bug#277949: base-installer: progress bar goes backward
Package: base-installer Severity: minor Tags: d-i During base-system install step of d-i it displays Preparing to install the Debian base system... and fills progress bar to about 1/10. Then progress bar is set to 0. -- System Information: Debian Release: 3.1 APT prefers unstable APT policy: (500, 'unstable') Architecture: i386 (i686) Kernel: Linux 2.4.27 Locale: LANG=uk_UA.KOI8-U, LC_CTYPE=uk_UA.KOI8-U (charmap=KOI8-U) -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#277524: Incorrect keyboard in X
20.10.2004 18:55 +0200 xerces8 : Package: installation-reports Debian-installer-version: 20040930 ( rc2 from web ) rc2 is not released AFAIK. Comments/Problems: After install and setup, the keymap in X11 is US ( or similar ), not slovene as I selected during installation ( it is correctly slovene in the /dev/tty[1-6] terminals (VT1-VT5). I also selected slovene ( 'sl' I think ) during config phase of X installation ( whet is asks about keyboard type, number of keys and layout ). Please try latest sid-di. It should install localization-config that should setup X keyboard automatically (in locale sl_SI, keyaboard sl,us). -- Eugeniy Meshcheryakov Kyiv National Taras Shevchenko University Information and Computing Centre http://icc.univ.kiev.ua signature.asc Description: Digital signature
Bug#276752: Urgent patch for kbd-chooser
19.10.2004 06:41 +0200 Christian Perrier (-): Quoting Denis Barbier ([EMAIL PROTECTED]): There may be problems with Ukrainian, I do not remember if they use console-tools or console-cyrillic. IMO adding charset iso-8859-5 at the beginning of ua-*.kmap.gz files could help (untested). I remember now: they use console-cyrillic. No. console-cyrillic is only used after reboot after second stage. It is installed by tasksel. I am not sure that charset lines are needed in keymap files - there is no symbolic names in files. And keymaps are in koi8-u and utf-8 not in iso-8859-5! -- Eugeniy Meshcheryakov Kyiv National Taras Shevchenko University Information and Computing Centre http://icc.univ.kiev.ua signature.asc Description: Digital signature
Bug#276752: (no subject)
19.10.2004 12:28 +0200 Denis Barbier : Eugeniy Meshcheryakov wrote: No. console-cyrillic is only used after reboot after second stage. It is installed by tasksel. I am not sure that charset lines are needed in keymap files - there is no symbolic names in files. And keymaps are in koi8-u and utf-8 not in iso-8859-5! This charset line (with the right charset, ie. koi8-u) is needed if you want to be able to load the same keymap both with koi8-u and UTF-8 encodings. If so, lines 'charset koi8-u' should be added to ua.kmap, ua-ws.kmap and uaw.kmap. Other Ukrainian keymaps are in UTF-8. The charset declaration tells console-tools how to perform conversions between Unicode codepoints, symbolic names and legacy encodings values. This is convenient on an installed system because keymaps does not have to be duplicated, but a charset line has to be declared. Recai spotted another problem in this bugreport, do you know if your keymap works fine after first reboot? It worked when I last time installed Debian, but it was netboot with udebs from testing. I will try with unstable. -- Eugeniy Meshcheryakov Kyiv National Taras Shevchenko University Information and Computing Centre http://icc.univ.kiev.ua signature.asc Description: Digital signature
Bug#276752: (no subject)
The charset declaration tells console-tools how to perform conversions between Unicode codepoints, symbolic names and legacy encodings values. This is convenient on an installed system because keymaps does not have to be duplicated, but a charset line has to be declared. Recai spotted another problem in this bugreport, do you know if your keymap works fine after first reboot? It worked when I last time installed Debian, but it was netboot with udebs from testing. I will try with unstable. I tested it with dayly build with unstable. There was no problems. -- Eugeniy Meshcheryakov Kyiv National Taras Shevchenko University Information and Computing Centre http://icc.univ.kiev.ua signature.asc Description: Digital signature
Bug#276071: languagechooser: Change Hebrew default locale to UTF-8
11.10.2004 23:19 +0200 Lior Kaplan (-): Package: languagechooser Severity: normal Tags: patch l10n Please change the Hebrew default locale to he_IL.UTF-8. This is almost as in bug #273718. I hope this will solve another bug related to the 2nd stage. It cannot be the same as bug #273718. This bug is about the ISO-8859-6 encoding that does not contain Arabic representation forms. ISO-8859-8 AFAIK contain all needed characters. The problem may be that languagechooser contains incorrect kbd record for Hebrew and that iso08.f16 font does not contain line-drawing characters but termwrap use jfbterm for 2nd stage of installation... -- Eugeniy Meshcheryakov Kyiv National Taras Shevchenko University Information and Computing Centre http://icc.univ.kiev.ua signature.asc Description: Digital signature
Re: LVM on software RAID - LILO
08.10.2004 12:43 +0100 Dan Peachey (-): Hi there, I've had a go at installing a system using the pre-rc2 netinst image. With the install I have 2 x HDD which I have setup a software RAID for in partman (/dev/md/0). I have then fired up a console and used pvcreate and vgcreate to create LVM groups. Going back to partman, it recognises the LVM VG and allows me to setup logical volumes. All goes smooth and I have an LV partition with /root mounted on it and a number of other LV's. No complaints from the installer, but when I get to installing the boot loader (LILO) it bombs out and will not let me install to /dev/md/0. I do not think that lilo can be booted from md device, try to use one of the physical volumes of this device (or disk device). Does lilo-installer display any disk devices (hda, sda...) in the menu? I was expecting to see GRUB as the boot loader as this has been in previous releases? But I guess GRUB might have failed too. So I'm a bit stuck how to get this to work. Any help greatly appreciated. GRUB probably cannot boot operating system from LV, lilo can. TIA, Dan -- Eugeniy Meshcheryakov Kyiv National Taras Shevchenko University Information and Computing Centre http://icc.univ.kiev.ua signature.asc Description: Digital signature
Re: d-i pre-rc2 has support for russian installation broken
08.10.2004 15:52 +0400 Nikita V. Youshchenko : On Fri, Oct 08, 2004 at 10:36:09AM +0200, Christian Perrier wrote: There have been discussions a few weeks ago about changes between Evguenyi, Anton and Nikolaithings have been commited and it seems that they weren't tested enough. I've got it. As far as I see that, the necessary changes for russian have been in both base-config and languagechooser. The changes in languagechooser has been half-deactivated, but not the changes in base-config. Therefore, the tools installation in languagechooser is deactivated (my change), but the language line in languagelist is commented out (old value is active). Therefore, stage2 tries to use console-cyrillic, which is not installed. Can languagechooser get an urgent update to the languagelist file? If at all possible, I think this upload should revert change, making it install console-cyrillic again. As a quick look shows, code currently in base-config expects console-cyrillic, and that's thr Right Thing. It may be Right Thing only for machines with AT keyboards. On machines with other keyboards console-cyrillic will make keyboard unusable (bug #271308). The fact is that console-cyrillic is the only tool currently in debian that can properly set up cyrillic VGA console (not framebuffer but VGA console). The problem goes deep into VGA: koi8 uses character codes 0xC0 - oxFF. which do interfer with some VGA control settings. This causes vertical lines (in debconf dialogs, in mc, etc) to be non-continious. To overcome that, a complex mapping has to be defined, such that KOI codes don't actually go to VGA console. Console-cyrillic does that. Colsole-tools does not. I tried installing Debian in Ukrainian using font ruscii_8x16 used by termwrap for Russian and Bulgarian (version in unstable, there is a patch for Ukrainian but it was not applied, I changed fort manually). I see no described deffects. Also, console-cyrillic gives a rich set of settings for cyrillic console. It's almoist a must-have on russian system. It will be instaled during second stage for all supported Cyrillic languages. -- Eugeniy Meshcheryakov Kyiv National Taras Shevchenko University Information and Computing Centre http://icc.univ.kiev.ua signature.asc Description: Digital signature
Re: d-i pre-rc2 has support for russian installation broken
08.10.2004 18:51 +0400 Nikita V. Youshchenko : Also, console-cyrillic gives a rich set of settings for cyrillic console. It's almoist a must-have on russian system. It will be instaled during second stage for all supported Cyrillic languages. ... and left unconfigured (i.e. not setting up console at system boot) in most cases - corresponding debconf question is of priority medium, and defaults no no :(. The default is yes for be, bg, mk, mn, ru, sr, uk but this question is really asked with medium priority and this will broke systems with non-AT keyboard... -- Eugeniy Meshcheryakov Kyiv National Taras Shevchenko University Information and Computing Centre http://icc.univ.kiev.ua signature.asc Description: Digital signature
Re: d-i pre-rc2 has support for russian installation broken
08.10.2004 13:36 -0400 Joey Hess (-): Nikita V. Youshchenko wrote: In fact, languagechooser currently is set up do use console-cyrillic, AND base-config expects console-cyrillic. What version of base-config? There are two and they do two different things for Russian, because people have been making changes to termwrap without consulting me; and unless I understand a change, I cannot get it into sarge at this time. As I've said before, repeatedly. Both sid and sarge versions of base-config use console-cyrillic if it is installed. -- Eugeniy Meshcheryakov Kyiv National Taras Shevchenko University Information and Computing Centre http://icc.univ.kiev.ua signature.asc Description: Digital signature
Bug#250789: patch: do not reset console font
I propose attached patch to fix this problem. It removes console resetting code from termwrap. I am sure that the only effect from using this patch is that console after termwrap will be with settings used during second stage (state suitable for used language). Patch does not change anything for languages that require terminal emulators. -- Eugeniy Meshcheryakov Kyiv National Taras Shevchenko University Information and Computing Centre http://icc.univ.kiev.ua Index: termwrap === --- termwrap(revision 1285) +++ termwrap(working copy) @@ -396,49 +396,3 @@ $@ -# The console font loaded before termwrap started. No idea how to -# detect it, so we assume it was ISO-8859-1. -ORIG_ENCODING=ISO-8859-1 -ORIG_ACM=iso01 -ORIG_FONT=lat0-sun16 - -case $ENCODING in -ISO-8859-1) - # Nothing to do, using default config - ;; -ISO-8859-2) - # Unload ISO-8859-2 charset mapping, by loading ISO-8859-1 instead - try_load_charset $ORIG_ENCODING $ORIG_ACM $ORIG_FONT - ;; -ISO-8859-7) - # Unload ISO-8859-7 charset mapping, by loading ISO-8859-1 instead - try_load_charset $ORIG_ENCODING $ORIG_ACM $ORIG_FONT - ;; -ISO-8859-9) - # Unload ISO-8859-9 charset mapping, by loading ISO-8859-1 instead - try_load_charset $ORIG_ENCODING $ORIG_ACM $ORIG_FONT -;; -ISO-8859-13) - # Unload ISO-8859-13 charset mapping, by loading ISO-8859-1 instead - try_load_charset $ORIG_ENCODING $ORIG_ACM $ORIG_FONT - ;; -ISO-8859-15) - # Unload ISO-8859-15 charset mapping, by loading ISO-8859-1 instead - try_load_charset $ORIG_ENCODING $ORIG_ACM $ORIG_FONT - ;; -KOI8-R|KOI8-U) - # Unload KOI8-R or KOI8-U charset mapping, by loading ISO-8859-1 instead - try_load_charset $ORIG_ENCODING $ORIG_ACM $ORIG_FONT - ;; -UTF-8) - # Disable UTF-8 in console - #if [ $TERM = console ] ; then - # echo -ne \033%@ - #else - # : # Nothing to do - #fi - ;; -*) - # Nothing to do - ;; -esac signature.asc Description: Digital signature
Re: Bug#250789: patch: do not reset console font
28.09.2004 17:25 +0300 Konstantinos Margaritis : On 28 2004 16:45, Eugeniy Meshcheryakov wrote: I propose attached patch to fix this problem. It removes console resetting code from termwrap. I am sure that the only effect from using this patch is that console after termwrap will be with settings used during second stage (state suitable for used language). Patch does not change anything for languages that require terminal emulators. This it almost definately NOT correct. Assuming that it was ISO 8859-1 is wrong, and if it was already UTF-8, disabling it is plain unacceptable. And console is in UTF-8 in other cases except Greek. I am looking into this myself, but I don't want to make it impossible for others to use the console. Assumptions like the ones you made are not good. Have you really seen this patch? It is *removes* code that sets ISO-8859-1 font. Konstantinos -- Eugeniy Meshcheryakov Kyiv National Taras Shevchenko University Information and Computing Centre http://icc.univ.kiev.ua signature.asc Description: Digital signature
Please fix bug #250789 in base-config
Hello, Please fix bug #250789 (termwrap unloads the charset mapping in second-stage installation). Today I installed Debian (sid) and see that it is not fixed and termwrap resets console font after installation. If it is not possible to just remove all code that resets console font after installation, then you can add command line option (say --no-reset) and run termwrap with this option in second stage. Thanks, -- Eugeniy Meshcheryakov Kyiv National Taras Shevchenko University Information and Computing Centre http://icc.univ.kiev.ua signature.asc Description: Digital signature
Re: /etc/init.d/console-screen.sh does not run on 1st boot
Konstantinos Margaritis : Hi, Joey and I tried testing greek support here in Oldenburg, and since console-* packages seem to support greek quite well right now, I suggested we try dumping use of jfbterm for 2nd stage. Unfortunately, it does not work, it does not set the font at 1st boot (just after d-i reboots the system). Apparently, /etc/init.d/console-screen.sh does not run to set the console font, or if it does run, it does not succeed in setting the font to greek. The strange thing is that it works from second boot. Are you sure that it is console-tools issue? Maybe it is termwrap that resets console font? It will do this if it did not found wrapper (jfbterm). Alastair, do you have any idea as to what might stop the script from setting the font on 1st boot, but not on 2nd boot? Thanks Konstantinos -- Eugeniy Meshcheryakov Kyiv National Taras Shevchenko University Information and Computing Centre http://icc.univ.kiev.ua signature.asc Description: OpenPGP digital signature
Bug#272136: [URGENT] Debian Installer: Bug#272136: countrychooser: db_go fails when choosing Arabic as language
Steve Langasek : On Mon, Sep 20, 2004 at 06:58:43AM +0200, Christian Perrier wrote: Steve, Alastair, Could you have a look at 272136 ? This bug has a very long story. It started by Debian installer crashes on Arabic menu for countrychooser. Now, after a lot of investigation, Eugeniy reduced it to something crashes when trying to display Arabic with whiptail. If I understand properly what is mentioned by Evgueniy, this even happens without libfribidi installed. libfribidi is only responsible for the bi-directional text, not the shaping; shaping was patched into slang directly. But slang does not execute shaping code when libfribidi is not loaded... -- Eugeniy Meshcheryakov Kyiv National Taras Shevchenko University Information and Computing Centre http://icc.univ.kiev.ua signature.asc Description: OpenPGP digital signature
Bug#272136: More progress about the Arabic crash in Debian Installer
Christian Perrier : Another info : Steve tried without libfribidi0 and the crash didn't happen Strange, when I tried without fribidi it crashed too... -- Eugeniy Meshcheryakov Kyiv National Taras Shevchenko University Information and Computing Centre http://icc.univ.kiev.ua signature.asc Description: OpenPGP digital signature
Bug#272136: More progress about the Arabic crash in Debian Installer
Petter Reinholdtsen : [Christian Perrier] Now really attached Running the program in valgrind gave this error report. Most of them look like bugs in glibc to me. Not sure if these bugs are the ones trigging the crash. I have different output on my system (from from gdb, without fribidi): Program received signal SIGSEGV, Segmentation fault. 0x401344dd in malloc_consolidate (av=0x401f1fe0) at malloc.c:4382 4382malloc.c: No such file or directory. in malloc.c (gdb) bt #0 0x401344dd in malloc_consolidate (av=0x401f1fe0) at malloc.c:4382 #1 0x40133f7a in _int_malloc (av=0x401f1fe0, bytes=1604) at malloc.c:4103 #2 0x40132ef3 in __libc_malloc (bytes=52) at malloc.c:3296 #3 0x40054cf9 in SLsmg_write_nchars () from /lib/libslang.so.1-UTF8 #4 0x4005431e in SLsmg_write_string () from /lib/libslang.so.1-UTF8 #5 0x4001dc1a in buttonDrawIt (co=0x80597b8, active=0, pushed=16) at button.c:113 #6 0x4001db6d in buttonDraw (co=0x644) at button.c:96 #7 0x4001ec76 in newtDrawForm (co=0x80586c0) at form.c:609 #8 0x4001f53f in newtFormRun (co=0x80586c0, es=0xb5a0) at form.c:911 #9 0x4001f29c in newtRunForm (co=0x80586c0) at form.c:816 #10 0x0804bafb in listBox (text=0x644 Address 0x644 out of bounds, height=18, width=31, optCon=0x804e770, flags=0, default_item=0x0, result=0x644) at dialogboxes.c:330 #11 0x0804ab13 in main (argc=1604, argv=0x644) at whiptail.c:540 -- Eugeniy Meshcheryakov Kyiv National Taras Shevchenko University Information and Computing Centre http://icc.univ.kiev.ua signature.asc Description: OpenPGP digital signature
Bug#272136: [URGENT] Debian Installer: Bug#272136: countrychooser: db_go fails when choosing Arabic as language
Christian Perrier : Looks like the problem is coming from diacritics (harakat). I don't see stand-alone harakat in your first attachment. Also, the only thing that is common between the country names mentioned and the po file you have attached, it seems, is that they include the 'shadda' diacritic. Others, feel free to correct me if I am wrong. Hmmm, diacritics indeed modify another character, right? This means that combining a character and a diacritic would give a different character. The combination is done by the shaping code and thus the final character is indeed not present in the original file, right? It is not correct. For Arabic diacrits console should render two characters in one position. As I see from bterm sources it supports such type of combining characters. If that is true, this means that the needed characters (the rsult of the combination of harakat and the original character) should be in the needed-characters/ar file as well probably as the standalone harakat. harakats should be in needed-characters. Could you complete the ar file I originally sent so that it includes harakat as well as combined characters (if I'm not wrong above) ? -- Eugeniy Meshcheryakov Kyiv National Taras Shevchenko University Information and Computing Centre http://icc.univ.kiev.ua signature.asc Description: OpenPGP digital signature
Bug#272136: [URGENT] Debian Installer: Bug#272136: countrychooser: db_go fails when choosing Arabic as language
Christian Perrier : Do you by chance have an idea of what is causing this problem. I suspect a missing character in the attached file, this character being indeed used after shaping of the translations for Syria and Saudi Arabia. As I can see this is not a bug in debian-installer, I can reproduce it with attached script (UTF8-encoded). Probably, it is not a bug in shaping/bidi code, I can reproduce it without /usr/lib/libfribidi.so.0. So this is bug in newt or slang. -- Eugeniy Meshcheryakov Kyiv National Taras Shevchenko University Information and Computing Centre http://icc.univ.kiev.ua #! /bin/sh /usr/bin/whiptail --menu menu 20 33 12 -- ألمانيّة سويسريّة signature.asc Description: OpenPGP digital signature
Bug#260227: patch
Ognyan Kulev : Christian Perrier wrote: If the same is needed for Bulgarian, then I leave the decision to Anton and Ognyan. Fine by me. Updated patch is attached. I commited this patch to svn, but I do not close this bug because console-cyrillic still used for Serbian and Russian and I do not sure whether Russian and Serbian translators want font to be changed to ruscii_8x16 or other. -- Eugeniy Meshcheryakov Kyiv National Taras Shevchenko University Information and Computing Centre http://icc.univ.kiev.ua signature.asc Description: OpenPGP digital signature
Bug#260227: patch
Christian Perrier : For Serbian, go ahead..the translation is still in very early stages (indeed, the translator did not answer much of my mails). I do not know what font to use for Sebian. ruscii_8x16 dose not contain all letters needed for Serbian. -- Eugeniy Meshcheryakov Kyiv National Taras Shevchenko University Information and Computing Centre http://icc.univ.kiev.ua signature.asc Description: OpenPGP digital signature
Bug#260227: patch
Nikolai Prokoschenko : On Mon, Sep 13, 2004 at 11:54:06AM +0200, Christian Perrier wrote: For Russian, we, for sure, need some input by Nikolai NIKOLA, DAVA!! :-) If it's fine for Ukrainian, it should be good enough for Russian. Committing patch. Hope it works, I'm not a console user... :( If so you should also remove code that installs console-cyrillic and console-terminus from languagechooser and make termwrap use ruscii_8x16 font instead unavailable console-cyrillic. -- Eugeniy Meshcheryakov Kyiv National Taras Shevchenko University Information and Computing Centre http://icc.univ.kiev.ua signature.asc Description: OpenPGP digital signature
Bug#271147: termwrap doesn't work correctly with KOI8-U and serial console
tags 271147 + patch thanks Anton Zinoviev : This bug is not that termwrap doesn't work for KOI8-U on serial console (look at #271145). It is that it should not invoke /usr/bin/cyr on serial console. Anton Zinoviev I propose to not use console-cyrillic for console setup for koi8-u, patch attached. -- Eugeniy Meshcheryakov Kyiv National Taras Shevchenko University Information and Computing Centre http://icc.univ.kiev.ua Index: termwrap === --- termwrap(revision 1257) +++ termwrap(working copy) @@ -302,11 +302,7 @@ ;; KOI8-U) # Load KOU8-U charset mapping into console - if [ -x /usr/bin/cyr ]; then - /usr/bin/cyr -s uni -f 16 -e koi8-u ua_ms ctrl_shift_toggle - else - try_load_charset $ENCODING koi8u koi8u_8x16 - fi + try_load_charset $ENCODING koi8u ruscii_8x16 ;; CP1251) # Load CP1251 charset mapping into console signature.asc Description: OpenPGP digital signature
Bug#261682: reassign
reassign 261682 grub-installer thnaks This bug still present. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#260227: patch
tags 260227 + patch thanks I propose to not use console-cyrillic for Ukrainian in order to fix this bug. User will be able to configure console-cyrillic in second stage. Patch also allows user to use non-cyrillic console. Patch if for Ukrainian only but the same font may be used for Russian (acm should be koi8-r). Index: languagelist === --- languagelist(revision 21466) +++ languagelist(working copy) @@ -49,6 +49,6 @@ # @euro because of Finland Swedish;sv;[EMAIL PROTECTED];sv;SV;sv_SV:sv:en_GB:en;kbd=lat0-sun16(iso15) Turkish;tr_TR;tr_TR;tr;TR;tr_TR:tr:en_GB:en;kbd=ter-916f(iso09) -Ukrainian;uk_UA;uk_UA;uk;UA;uk_UA:uk:en_GB:en;cyr=uni,16,koi8-u,ua_ms(ctrl_shift_toggle) +Ukrainian;uk_UA;uk_UA;uk;UA;uk_UA:uk:en_GB:en;kbd=ruscii_8x16(koi8u) Vietnamese;vi_VN;vi_VN;vi;VN;vi_VN:vi:en_GB:en Welsh;cy_GB.UTF-8;cy_GB.UTF-8;cy;GB;cy_GB:en_GB:en;kbd=iso14.f16(utf8)
Bug#271147: termwrap doesn't work correctly with KOI8-U and serial console
Ognyan Kulev : Package: base-config Version: 2.44 Tags: d-i When termwrap is invoked for KOI8-U environment and terminal is serial console, /usr/bin/cyr is executed. However, this command has no sense for serial console. Does it work for other charsets? Regards, ogi -- Eugeniy Meshcheryakov Kyiv National Taras Shevchenko University Information and Computing Centre http://icc.univ.kiev.ua -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#261476: languagechooser entry for bulgarian
Anton Zinoviev : By the way languagechooser apt-installs console-cyrillic and xfonts-terminus for ru but not for uk. languagechooser installs console-cyrillic for Ukrainian, see prebaseconfig. Is it possible to split fonts from console-cyrillic to use them as for other languages (kbd=...)? I did not found good font for Ukrainian in console-data. -- Eugeniy Meshcheryakov Kyiv National Taras Shevchenko University Information and Computing Centre http://icc.univ.kiev.ua -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#269972: languagechooser: should not display languages unavailable on floppy
Package: languagechooser Severity: normal Tags: d-i Languagechooser currently displays all available languages after booting from floppies, so if user choose language that is not available on root floppy (Ukrainian on root.img) installation will proceed in English, that is very confusing. -- System Information: Debian Release: 3.1 APT prefers unstable APT policy: (500, 'unstable') Architecture: i386 (i686) Kernel: Linux 2.4.27 Locale: LANG=uk_UA.KOI8-U, LC_CTYPE=uk_UA.KOI8-U -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#269971: languagechooser: first screen is untranslated after booting from cyrillic-root floppy
Package: languagechooser Severity: normal Tags: l10n d-i After booting from boot+cyrillic-root floppies first srceen (Choose Language) appears untranslated (only English language names). -- System Information: Debian Release: 3.1 APT prefers unstable APT policy: (500, 'unstable') Architecture: i386 (i686) Kernel: Linux 2.4.27 Locale: LANG=uk_UA.KOI8-U, LC_CTYPE=uk_UA.KOI8-U -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: hindi xlation
Steve Langasek : Hindi uses an alphabetic writing system, so this shouldn't be too problematic. The unifont package may even include the characters already. Hindi uses combining characters, this is main problem. I am not sure whether bterm and jfbterm support combining characters and I am not sure whether all applications properly support them. -- Eugeniy Meshcheryakov Kyiv National Taras Shevchenko University Information and Computing Centre http://icc.univ.kiev.ua -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#262714: (no subject)
Keith Hellman : I went through the repartitioning by first removing all the partitions, then adding back in a root, swap, and two LVM physical volumes. When it came time to declare a volume group however, the name system failed. The error messgages on the third console suggest the issue being that the system booted and detected the original 'system', and refused to creat another A volume group called 'system' alread exists. I'm not sure if I would have experienced this error had I *first* deleted the volume group and *then* repartitioned the drive. Yes, now you must delete all volume groups and maybe run pvremove for all PVs that you does not want to use. I'm afraid it is too late to try that now since the drive is already repartitioned. I worked around this be simply calling my Volume Group 'sys'. This is my third use of the installer for LVM and things keep getting smoother. I'm not sure if you've fixed things in the LVM logic or perhaps I'm just becoming more experienced with it - but its PROGRESS! Thanks for all your efforts. -- Eugeniy Meshcheryakov Kyiv National Taras Shevchenko University Information and Computing Centre http://icc.univ.kiev.ua -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#262531: installation-reports: failed, root on LVM
Package: installation-reports Debian-installer-version: built from svn, today uname -a: i386 Date: Date and time of the install Method: netboot Machine: Description of machine (eg, IBM Thinkpad R32) Processor: Memory: Root Device: LVM LV Root Size/partition table: Feel free to paste the full partition table, with notes on which partitions are mounted where. Output of lspci: Base System Installation Checklist: Initial boot worked:[O] Configure network HW: [O] Config network: [O] Detect CD: [ ] Load installer modules: [O] Detect hard drives: [O] Partition hard drives: [O] Create file systems:[O] Mount partitions: [O] see comments Install base system:[O] Install boot loader:[E] see comments Reboot: [O] [O] = OK, [E] = Error (please elaborate below), [ ] = didn't try it Comments/Problems: I tried to install Debian (testing) with root device on LVM. Installation failed on boot loader installation step. First time d-i offered to install grub. I returned to main menu to install lilo. lilo installation failed. Problem is that root partition (VG vg01, LV newroot) was mounted as /dev/mapper/vg01-newroot. LVM tools does not understand logical volumes given in that format. Partition name should be /dev/vg01/newroot (this file is really symlink to /dev/mapper/vg01-newroot). d-i must not resolve such symlinks (resolved name was also writen to /target/etc/fstab!). After that I tryed to install d-i with low debconf priority. Before bootloader installation step I unmounted /target and mounted it again with command: mount -t ext3 /dev/vg01/newroot /target Then I choosed to install lilo (into MBR). There was no messages about failed installation. Screen just blinked. After that main menu appeared with header in English and menu entries in Ukrainian. Messages in other dialogs was in English (all messages are translated in Ukrainian in normal install!). After reboot my old grub bootloader appeared, so lilo was not installed. Next try... After lilo install i ran: chroot /target /sbin/lilo (this code is in lilo-installer too) Following messages appeared: Warning: '/proc/partitions' does not exists, disc scan bypassed Fatal: Sorry, don't know how to handle device 0xfe00 I did not seen such messages on any console before! I mounted /target/proc and rerun this command. Some warings appeared, no error messages. After reboot lilo loaded new Debian system. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#262531: reassign to partman-target
reassign 262531 partman-target retitle 262531 LVM LVs should be mounted as /dev/VG/LV thanks LVM logical volumes should be named /dev/VG/LV for lvm tools. Now fstab entries for LVM LVs are written as /dev/mapper/VG-LV this brokes lilo-installer with root on LVM. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Why isn't console-cyrillic part of console-data?
Anton Zinoviev : 27 2004 Eugeniy Meshcheryakov : Is it possible to configure console font automatically and ask for keymap in second stage? My idea was automatically configure console font in d-i and ask wheter user want to use console-cyrillic to setup keyboard in second stage (as fix for bug #260227). Today I uploaded a new version of console-cyrillic where I changed the Debconf part in the following way. When the package is not installed for first time, for example when it is upgraded or when it was removed (but not purged) then it asks the questions the same way as now. But when the package is installed for first time (probably by the debian-installer) then it uses the LANG environment variable in order to compute better default values for keyboard, font and encoding and asks the corresponding questions with smaller priority. For example when console-cyrillic is installed with LANG=uk_UA there will be only two questions - about the keys to use as Cyr/Lat toggler and switcher. (The terminology comes from XFree: Shift switches, Caps Lock toggles.) Anton Zinoviev -- Eugeniy Meshcheryakov Kyiv National Taras Shevchenko University Information and Computing Centre http://icc.univ.kiev.ua -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Why isn't console-cyrillic part of console-data?
Anton Zinoviev : 26 2004 Eugeniy Meshcheryakov : Anton Zinoviev : Another difference (that made console-cyrillyc the best choice for the Cyrillic languages) are the debconf questions during the installation. Oops, now there is no such question for Serbian, Russian and Ukrainian :-) Oops! I decided to try the installer with Russian language. Which one of the following scenarios you like more: For machines with AT keyboard - first. RUSSIAN - console-cyrillic is configured with default settings during the first stage - it is activated during the base-config No it is activated during boot. Termwrap also contains code that should be remove after release (#250789) - it is deactivated afterwards Same bug in termwrap. - after reboot it is activated permanently but not in the preferred for the user way (I troubled which key to use as Cyr/Lat switcher) As I can see from termwrap it is preferred way (ctrl_shift_toggle). BULGARIAN - console-cyrillic is not configured during the first stage Yes. - during base-config jbterm is active (although this might be console-tools like the Latin languages and Greek) Termwrap has no settings for CP1251. See also #250789. - during base-config the Cyrillic task is automatically selected and console-cyrillic installed cyrillic task contains console-cyrillic - the user configures the keyboard as he or she wishes one more step (or maybe 4?) in install process ;-). And this can be not so intuitive for new users. - after base-config jbterm and Cyrillic are deactivated bug in termwrap - after reboot the user has customized Cyrillic on the console By the way, what do you think about bug #260227? The quick fix can be: 1. d-i doesn't feed console-cyrillic with default configuration 2. console-cyrillic allows the user to choose don't touch the keyboard Is it possible to configure console font automatically and ask for keymap in second stage? The best fix is to add support for non-i386 in console-cyrillic. I can do this if there are testers. This may be a problem for all supported architectures. By the way I observed the following bugs during second stage of the installer with Russian language: There was a message before the configuration of the base-system: cyr: Unknown option: -t cyr: Try `cyr --help for more information However for some unknown reason the Cyrillic setup worked. ;-) This is known bug. Fixed in base-config. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#261671: some entries in languagechooser.l10n still untranslated
Package: languagechooser Severity: normal Tags: l10n Some messages in languagelist.l10n sill untranslated after changing format of this file: Albanian: Albanian Vietnamese: Vietnamese Following translations seems to be incorrect (wrong case): Slovenian: Slovenini (Slovenina?) Lithuanian: Lietuvikai - not sure about this -- System Information: Debian Release: 3.1 APT prefers unstable APT policy: (500, 'unstable') Architecture: i386 (i686) Kernel: Linux 2.4.26 Locale: LANG=uk_UA.KOI8-U, LC_CTYPE=uk_UA.KOI8-U -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#261682: installation-report: OS detection problems
Package: installation-reports INSTALL REPORT Debian-installer-version: build from SVN, 2004/07/27 uname -a: Linux intel4 2.4.26 #1 26 19:53:34 EEST 2004 i686 GNU/Linux Date: 2004/07/27 Method: using netboot initrd Machine: Description of machine (eg, IBM Thinkpad R32) Processor: Memory: Root Device: IDE? SCSI? Name of device? Root Size/partition table: part1 1364829302528+ 5 Extended part236493953 2449912+ 4f QNX4.x 3rd part (QNX Momentix) part339544870 7365802+ 83 Linux (new Debian system) part5 1 37 297139+ 82 Linux swap (swap for new and old Debian systems) part6 38 81 353398+ 83 Linux (root of old Debian system) part7 82364828651896 8e Linux LVM (old Debian System) Output of lspci and lspci -n: Base System Installation Checklist: [O] = OK, [E] = Error (please elaborate below), [ ] = didn't try it Initial boot worked:[O] Configure network HW: [O] Config network: [O] Detect CD: [O] Load installer modules: [O] Detect hard drives: [O] Partition hard drives: [O] Create file systems:[O] Mount partitions: [O] Install base system:[O] Install boot loader:[O] Reboot: [O] Comments/Problems: I installed unstable. There was no error in first stage. grub-installer found my old Debian system (it was listed as Debian GNU/Linux ). It did not found QNX, but I do not think that it is a problem. After reboot grub appeared with following menu: Debian GNU/Linux, kernel 2.4.26-1-686 Debian GNU/Linux, kernel 2.4.26-1-686 (recovery mode) QNX Momentix (on /dev/hda6) There was no other Debian GNU/Linux listed in menu and /dev/hda6 device is really root partition of old Debian system. That not only problem. Commands for loading QNX Momentix...: root(hd0,5) kernel /boot/vmlinuz-2.6.3 ro root=/dev/hda6 initrd /boot/initrd.img-2.4.26-1-686 savedefault boot Kernel 2.6.3 was not default in old system. I am not sure that this kernel will boot with that initrd ;-) My old grub.conf attached. Booted new Debian system. During boot some program detected that I have nvidia card and loaded framebuffer module. After that I had 3 cursors instead of one :-) (one big graphical cursor, and too small cursors on top of sreen with blue line). I do not think that loading framebuffer module is really necessery. During passwd configuration I was asked about my full name. I entered my name in Ukrainian. After that passwd suggested me my first name as default value for login - will it work? :-) timeout 5 default 0 splashimage=(hd0,5)/boot/grub/splash.xpm.gz title Debian GNU/Linux 3.0 (2.4.26) root (hd0,5) kernel /boot/vmlinuz-2.4.26 ro root=/dev/hda6 vga=771 video=vesa:[EMAIL PROTECTED] title Debian GNU/Linux 3.0 (2.4.26-1-686) root (hd0,5) kernel /boot/vmlinuz-2.4.26-1-686 ro root=/dev/hda6 initrd /boot/initrd.img-2.4.26-1-686 title Debian GNU/Linux 3.0 (2.6.3) root (hd0,5) kernel /boot/vmlinuz-2.6.3 ro root=/dev/hda6 title QNX Momentix root (hd0,2) chainloader +1 boot
Bug#261476: languagechooser entry for bulgarian
Ognyan Kulev (): Package: languagechooser Version: 1.28 Tags: d-i /* Bcc: Recai and Anton. */ Recai Oktas wrote: I've noticed that Bulgarian does not use a console-cyrillic entry in languagechooser: Bulgarian;bg_BG;bg_BG;bg;BG;bg_BG:bg:en_GB:en; Instead it fallbacks to jfbterm since nothing has been defined in the last field. It also doesn't use the console font Terminus. Just been curious whether these are intentional since those mentioned packages (console-cyrillic and console-terminus) are maintained, AFAIK, by a well-known Bulgarian developer Anton Zinoviev. You can use folowing line instead: Bulgarian;bg_BG;bg_BG;bg;BG;bg_BG:bg:en_GB:en;cyr=uni,16,cp1251,bg_bds(ctrl_shift_toggle) You seem to be right. But I don't know how it's supposed to be fixed. So I file a bug. All this should be delayed for after RC1. Regards, ogi -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Why isn't console-cyrillic part of console-data?
Petter Reinholdtsen : It would be easier to get the console setup right during installation if all the console configuration stuff worked the same for all. console-cyrillic seem to be an exception. Why isn't the content and behaviour of console-cyrillic merged into console-data/console-tools? My question is why console-data is not part of console-cyrillic? ;-) Console-cyrillic have more features and simpler interface. But I agree that one package for all console fonts, keyboards, etc will be better. The main problem for now is bad cyrillic fonts in console-data, and console-cyrillic is much simper: only one command with simple interface for setting up cyrillic console with any font, encoding, keyboard. But there is one problem: console cyrillic contains only AT keymaps... -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#261476: languagechooser entry for bulgarian
Ognyan Kulev : Eugeniy Meshcheryakov wrote: You can use folowing line instead: Bulgarian;bg_BG;bg_BG;bg;BG;bg_BG:bg:en_GB:en;cyr=uni,16,cp1251,bg_bds(ctrl_shift_toggle) Does this use console-cyrillic? yes Does it use console-terminus font? Terminus Unicode has problems, Terminus Slavic is the best choice. I am not sure about what font your are talking. With this parameters cyr will be running as follows: cyr -s uni -f 16 -e cp1251 bg_bds ctrl_shift_toggle Used font is UniCyr_8x16 from console-cyrillic package? Does this font have problems? Are there other locations that need to be changed, like termwrap? Without any changes termwrap will continue to run jfbterm but console will be correctly set up after reboot. Some changes probably will be made after release, see #250789 for more info. Regards, ogi -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Why isn't console-cyrillic part of console-data?
Anton Zinoviev : Another difference (that made console-cyrillyc the best choice for the Cyrillic languages) are the debconf questions during the installation. Oops, now there is no such question for Serbian, Russian and Ukrainian :-) By the way, what do you think about bug #260227? -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#250789: Please fix this bug!
Recai Oktas : * Denis Barbier [2004-07-24 22:51:00+0200] [...] If termwrap is needed outside of d-i, I suggest to have a minimal termwrap-di to deal with Asian locales, and nothing more. Eugeniy has once said that only those languages without a declared debian-installer/consoletype needed termwrap. I've prepared proof of concept by using the idea. Attached is the minimal implementation. It doesn't behave according to the ENCODING, instead it first evaluates the CONSOLETYPE from debian-installer/consoletype and only then it does something based on the ENCODING. Following explains this idea: WRAPPER= WRAPPER_OPTION= CONSOLETYPE=$(get_db debian-installer/consoletype) ENCODING=$(locale charmap) case $CONSOLETYPE in kbd|cyr) Should be kbd*|cyr*) here. Consoletype is kbd=font(acm) or cyr=font,size,encoding,keymap(modifiers). # Nothing to do. ;; *) case $ENCODING in eucJP|EUC-JP) [...] Please note that this is only a draft, far from complete and untested. Now, could it be a base for further discussions? -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#250789: Please fix this bug!
Recai Oktas : Hi, I'm really annoyed this bug still lives despite my (and Eugeniy's) numerous attempts to explain the problem. Here, I've been receiving complaints from users saying that the Turkish characters didn't display properly after the installation finished. Please apply the patch attached and fix this issue. We haven't even give it a try. I can safely state that I am not expecting problems for latin and cyrillic languages. I agree that this bug must be fixed. It is very bad to see that just after installing the system I cannot use any console application with Ukrainian interface because I cannot see *any* Cyrillic letter. IMO, console font setup should also be removed from termwrap, languagechooser handle it. Termwrap only needed for running jfbterm for some languages. But this change is not so important. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#250789: Please fix this bug!
Petter Reinholdtsen : [Recai Oktas] Please apply the patch attached and fix this issue. Your patch is not correct. I suspect a real fix will involve detecting the initialy loaded font/acm/encoding, and undo the termwrap font loading by loading the initialy loaded font at the end. Why do this? This does not needed for installer. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#250789: Please fix this bug!
Recai Oktas : * Petter Reinholdtsen [2004-07-24 18:20:54+0200] I suggest we patch languagechooser to update /etc/locale.gen and /etc/environment to set the default locale, and remove the locale generating code from termwrap. I also suggest we teach termwarp to look in the data file for console-tool (is this /etc/console-tool/config? The variables seem to be SCREEN_FONT and SCREEN_FONT_MAP), and tell it to use these values if available as the default instead of ISO-8859-1. This way the termwrap tool is still a generic wrapper to run programs with correct console font and locale, while the locale generating is moved to a more sensible place. No, instead of SCREEN_FONT_MAP, we should use APP_CHARSET_MAP (hence acm) which has already been changed in languagechooser. For cyrillic languages settings does not stored in /etc/console-tool/config, in /etc/console-cyrillic instead, format of this file is completely different. Care to make a patch for termwrap? I already have a patch for languagechooser available. What about making program for installer only? It will only contain checking for debian-installer/consoledisplay debconf value and loading jfbterm if this value is not set? I am definitely willing to work on it. But what about the release schedule? I think such a gross change needs to have a serious test which takes a while. Thanks for your explanation. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Console-common basque tanslation strange situation
Christian Perrier : Quoting ^pi^ ([EMAIL PROTECTED]): Hi: I saw in tanslation-status page the basque translation to console-common have a fuzzy string, but when I downloade the console-common svn (svn co svn://svn.debian.org/console-common console-common) in the file isnt any fuzzy string :? If I look the web file (http://people.debian.org/~seppy/d-i/2nd-stage/files/eu/console-common_po_eu.po) the third string was fuzzy but no in the svn. I dont update the file time ago, the last bug (#244675) was closed in 11 May 2004 Can any help me ? A msgmerge is probably needed. Doing it. Also install-keymap.pot contains strings that I cannot find in any file: Warning: cannot access console; deferring until console is accessible. and I do not know how to regenerate it. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#260227: countrychooser: console-cyrillic always installs AT keymaps
Package: countrychooser, console-cyrillic Version: SVN Severity: important Languagechooser currently uses console-cyrillic package to set up Cyrillic fonts and keymaps for several languages. The problem is that this package contains only AT keymaps and will always install them, even on architectures with different keyboards. It will make installation on some archs (PPC for example) problematic (but cannot test that). -- System Information: Debian Release: testing/unstable APT prefers unstable APT policy: (500, 'unstable') Architecture: i386 (i686) Kernel: Linux 2.4.26 Locale: LANG=uk_UA, LC_CTYPE=uk_UA -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#260233: tasksel: incorrect description of th japanese-desktop task
Package: tasksel Version: 2.06 Severity: normal Description of the japanese-desktop is Japanese environment, should be Japanese desktop. -- System Information: Debian Release: testing/unstable APT prefers unstable APT policy: (500, 'unstable') Architecture: i386 (i686) Kernel: Linux 2.4.26 Locale: LANG=uk_UA, LC_CTYPE=uk_UA Versions of packages tasksel depends on: ii aptitude 0.2.15.2-1 curses-based apt frontend ii debconf [debconf-2.0] 1.4.29 Debian configuration management sy ii liblocale-gettext-perl1.01-17Using libc functions for internati -- debconf information excluded -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#260227: countrychooser: console-cyrillic always installs AT keymaps
Christian Perrier : Quoting Eugeniy Meshcheryakov ([EMAIL PROTECTED]): Package: countrychooser, console-cyrillic Version: SVN Severity: important Languagechooser currently uses console-cyrillic package to set up Uh, why assign this to countrychooser, then ? :-) Countrychooser contains code for console setup... Except if there's a reason I'm missing, I think this needs to be reassigned to languagechooser, console-cyrillic and probably more to console-cyrillic with Please also provide non-AT keymaps Maybe. But providing keymaps for all architectures and than selecting right keymap... I think possibility to use console-cyrillic without setting up keymaps (only fonts) will be useful though. Keymap setup may be done by using console-tools+console-data. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#260258: languagechooser: jfbterm is not required for Ukrainian
Package: languagechooser Severity: minor Tags: patch jfbterm and unifont are not required for installation in Ukrainian but installed by languagechooser. -- System Information: Debian Release: testing/unstable APT prefers unstable APT policy: (500, 'unstable') Architecture: i386 (i686) Kernel: Linux 2.4.26 Locale: LANG=uk_UA.KOI8-U, LC_CTYPE=uk_UA.KOI8-U Index: languagechooser === --- languagechooser (revision 17985) +++ languagechooser (working copy) @@ -70,7 +70,7 @@ fi case $LANGUAGE in -ja|ko|ko_KR|el|zh|zh_CN|zh_TW|bg|ar|he|uk|uk_UA) +ja|ko|ko_KR|el|zh|zh_CN|zh_TW|bg|ar|he) # Japanese, Korean, Greek, Chinese, Bulgarian, Arabic, Hebrew, # Ukrainian apt-install jfbterm || true
Bug#258981: network-console: partman does not show partitions
Frans Pop : On Tuesday 13 July 2004 12:37, Bastian Blank wrote: This may be a charset problem. Try to use an UTF8 capable terminal. I tried the following: - konsole with 'unicode font' selected = partitions not shown - xterm = partitions not shown - uxterm = partitions not shown - vt1 = partitions are shown, but borders, scrollbar and non-ascii chars are not shown correctly It does seem to have to do with non-ascii chars. If I select Greek as language, almost the whole main menu disappears on the remote console :-( If I choose Chech, lines are cut off at the first non-ascii character. (My remark on IRC was incorrect: the language on the remote is identical to the one selected on the laptop.) Any other ideas? Configuration changes or packages I could install on my Woody box to get UTF8 support? Try to use following command line to connect to d-i: TERM=linux ssh [EMAIL PROTECTED] It should be executed terminal that work in UTF-8 mode, like LANG=ll_CC.UTF-8 konsole -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: [Proposal] The locale generation at the prebaseconfig
Recai Oktas : * By moving the locale generation code to prebaseconfig, one of the central duties of termwrap will be transfered to this package, so we will be able to simplify the termwrap. According to my tests, I can safely state that, termwrap would not be needed at all for languages having a charset of ISO-8859-[1-15]. But it would have to be used mainly for Asian and Cyrillic consoles (not sure for the latter). Termwrap will not be needed for all languages that have kbd or cyr entries in languagelist. Languages that does not have such entries are: Bulgarian, Chinese, Japanese, Korean and Vietnamese (Bulgarian can be handled by console-cyrillic but there is no cyr entry in languagelist). Other languages does not need any console setup in termwrap. As a side note, please notice that we wouldn't need to touch the termwrap in an effort to simplify it, till the proposed patch matures enough. So, have you any comments or objections? Regards, [1] The current code base of 'loadkeys' is rather messy, which makes such modifications difficult. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: 20040708 i386 sid_d-i netinst 2.4 OK (sv_SE) and NOT OK (ru_RU)
Christian Perrier : Environment: VmWare Platform : i386 Kernel : 2.4 Locale : sv_SE Image : netinst sid_d-i 20040707 Priority : high Result : No special problem a ru_RU install however fails on horrible display in 2nd stage Eugeniy, Recai, can you have a look and try to understand why Russian has a so bad display on 2nd stage? ru_RU locale created and used in second stage (encoding is ISO-8859-5). But console is set up to use KOI8-R encoding (or ISO-8859-1 by termwrap?). languagelist contains proper value for Russian locale (ru_RU.KOI8-R) but encoding part is lost somewhere. I guess Ukrainian has the same problem. Do not think so, but will check. -- Eugeniy Meshcheryakov Kyiv National Taras Shevchenko University Information and Computing Centre http://icc.univ.kiev.ua -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: 20040708 i386 sid_d-i netinst 2.4 OK (sv_SE) and NOT OK (ru_RU)
Eugeniy Meshcheryakov : Christian Perrier : Environment: VmWare Platform : i386 Kernel : 2.4 Locale : sv_SE Image : netinst sid_d-i 20040707 Priority : high Result : No special problem a ru_RU install however fails on horrible display in 2nd stage Eugeniy, Recai, can you have a look and try to understand why Russian has a so bad display on 2nd stage? ru_RU locale created and used in second stage (encoding is ISO-8859-5). But console is set up to use KOI8-R encoding (or ISO-8859-1 by termwrap?). languagelist contains proper value for Russian locale (ru_RU.KOI8-R) but encoding part is lost somewhere. This is a bug in countrychooser. It determines part of locale after . or @ by using following code: EXTRA_LANGUAGECHOOSER=`echo $DEFAULTLOCALE | sed -e 's/[EMAIL PROTECTED]//'` But DEFAULTLOCALE is usually ll or ll_CC. It should be: EXTRA_LANGUAGECHOOSER=`echo $FALLBACKLOCALE | sed -e 's/[EMAIL PROTECTED]//'` -- Eugeniy Meshcheryakov Kyiv National Taras Shevchenko University Information and Computing Centre http://icc.univ.kiev.ua Index: debian/postinst === --- debian/postinst (revision 17695) +++ debian/postinst (working copy) @@ -124,7 +124,7 @@ DEFAULTLOCALE=C fi # If present, keep track of charset or modifier we got from languagechooser -EXTRA_LANGUAGECHOOSER=`echo $DEFAULTLOCALE | sed -e 's/[EMAIL PROTECTED]//'` +EXTRA_LANGUAGECHOOSER=`echo $FALLBACKLOCALE | sed -e 's/[EMAIL PROTECTED]//'` FIRST_LANG=$(echo $LANGUAGELIST | sed -e 's/:.*$//')
Re: Bug#257426: marked as done (console-tools: incorrect koi8 to unicode mapping)
Debian Bug Tracking System : This bug was fixed in the latest upload, 0.58. Regards Alastair Are you sure? SVN version of kbd-chooser still contains incorrect Unicode values for koi8u symbols. -- Eugeniy Meshcheryakov Kyiv National Taras Shevchenko University Information and Computing Centre http://icc.univ.kiev.ua -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Modified patch for plus symbols problem
Recai Oktas :
* Recai Oktas [2004-07-05 10:16:30+0300]
[...]
Index: packages/kbd-chooser/getfd.c
===
--- packages/kbd-chooser/getfd.c (revision 17443)
+++ packages/kbd-chooser/getfd.c (working copy)
@@ -45,7 +45,9 @@
}
int getfd() {
-int fd;
+static int fd = -1;
+if (fd = 0)
+ return fd;
fd = open_a_console(/dev/tty);
if (fd = 0)
Ok, This should be the way to go. I'll test it and modify the patch.
Thanks for your time and comments.
I prepared a new patch against kbd-chooser. First of all, I applied
Denis's patch for plus symbols. I also modified my previous patch in
accordance with Denis's suggestion quoted above. By reverting the
changes applied to some functions, now kbd-chooser codebase doesn't
divert from the console-tools. The problem regarding the getfd calls
was solved by modifying only the kbd-chooser spesific getfd.c file.
Here is the changelog:
* Denis Barbier
- Make kbd-chooser work with keymaps containing plus symbols.
* Recai Oktas
- Preserve compatibility with the console-tools by modifying
the previous patch.
I tested it for French, Ukrainian and Turkish keyboards and didn't see
any problem. It should also work with the 'ua.kmap' containing plus
symbols. Eugeniy, could it be possible to confirm this? [1]
This patch works. Thanks.
Regards,
[1] http://l10n-turkish.alioth.debian.org/kbd/kbd-chooser_0.58_i386.udeb
--
Eugeniy Meshcheryakov
Kyiv National Taras Shevchenko University
Information and Computing Centre
http://icc.univ.kiev.ua
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#257663: partman-lvm: / on lv does fails at kernel install
Karl Hegbloom : package: partman-lvm severity: normal There should be a check that / is not a logical volume since the mkinitrd does not support that. initrd-tools should support root on lvm. What version of installer do you use? Grub probably doesn't either, right? Grub does not but lilo does. It would be useful to have recipes for LVM installs. -- Eugeniy Meshcheryakov Kyiv National Taras Shevchenko University Information and Computing Centre http://icc.univ.kiev.ua -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: Patches for Bug#251550
Recai Oktas 03.07.2004 14:38: Hi, I've created two patch sets: http://l10n-turkish.alioth.debian.org/kbd/kbd-chooser-1.patch http://l10n-turkish.alioth.debian.org/kbd/kbd-chooser-2.patch * kbd-chooser-1.patch: - Activate unicode mode in d-i console. This resolves the keyboard freeze issue (#251550). * kbd-chooser-2.patch: - Apply Denis's patch. - Use a global 'file_descriptor' variable instead of multiple getfd() calls. This resolves the problem of Turkish unicode keymap. I don't know whether it could also resolve the problem of Ukrainian keymap. I've also built a new kbd-chooser udeb with the patches above applied. http://l10n-turkish.alioth.debian.org/kbd/kbd-chooser_0.56_i386.udeb Could you review and test them? That works in d-i for Ukrainian keymap. Regards, -- Eugeniy Meshcheryakov Kyiv National Taras Shevchenko University Information and Computing Centre http://icc.univ.kiev.ua -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: (forw) Bug#251550 acknowledged by developer (Bug#251231: fixed in console-tools 1:0.2.3dbs-53)
Christian Perrier 30.06.2004 09:14: 251550 was originally a d-i bug which after lot of analysis turned out to be a console-tools bug. This is the d-i freezes when entenring high ASCII characters bug. Denis Barbier spotted the problem, provided a patch, and Alastair integrated it in console-tools. IIRC, the same patch has to be applied to kbd-chooser's loadkeys. I guess Alastair will make the change in the currently unrelease 0.55 version. My opinion is that this should make its way into testing as soon as a few days sid_d-i tests will have proven it does not break the general keyboard handling. This breaks keyboard handling: unicode keymaps (ua-utf, consoel-cyrillic unicode keymaps) does not work (I can enter only ASCII symbols). -- Eugeniy Meshcheryakov Kyiv National Taras Shevchenko University Information and Computing Centre http://icc.univ.kiev.ua -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#256371: discover1: [INTL:uk] Ukrainian debconf templates translation update
Package: discover1 Severity: wishlist Tags: patch l10n d-i Please update Ukrainian debconf templates translation (po file attached). Thanks -- System Information: Debian Release: testing/unstable APT prefers unstable APT policy: (500, 'unstable') Architecture: i386 (i686) Kernel: Linux 2.4.26 Locale: LANG=uk_UA, LC_CTYPE=uk_UA # translation of discover1-templates.po to Ukrainian # translation of discover-templates.po to Ukrainian # #Translators, if you are not familiar with the PO format, gettext #documentation is worth reading, especially sections dedicated to #this format, e.g. by running: # info -n '(gettext)PO Files' # info -n '(gettext)Header Entry' #Some information specific to po-debconf are available at #/usr/share/doc/po-debconf/README-trans # or http://www.debian.org/intl/l10n/po-debconf/README-trans# #Developers do not need to manually edit POT or PO files. # Eugeniy Meshcheryakov [EMAIL PROTECTED], 2004. # msgid msgstr Project-Id-Version: discover1-templates_uk\n Report-Msgid-Bugs-To: \n POT-Creation-Date: 2004-05-04 14:13+0200\n PO-Revision-Date: 2004-06-26 18:29+0300\n Last-Translator: Eugeniy Meshcheryakov [EMAIL PROTECTED]\n Language-Team: Ukrainian\n MIME-Version: 1.0\n Content-Type: text/plain; charset=UTF-8\n Content-Transfer-Encoding: 8bit\n X-Generator: KBabel 1.3.1\n Plural-Forms: nplurals=3; plural=(n%10==1 n%100!=11 ? 0 : n%10=2 n%10=4 (n%10010 || n%100=20) ? 1 : 2);\n #. Type: boolean #. Description #: ../discover1.templates:4 msgid Manage CD-ROM devices and mount points with discover? msgstr CD-ROM discover? #. Type: boolean #. Description #: ../discover1.templates:4 msgid The \discover\ program can automatically manage the /dev/cdrom* symbolic links and device mount points, if you desire. The available CD-ROM (and compatible) devices are scanned at boot-time and these links and mount points can be automatically updated. msgstr , \discover\ /dev/cdrom* . CD- ROM ( ) , . #. Type: boolean #. Description #: ../discover1.templates:4 msgid If you enable this option, discover will link /dev/cdrom to /dev/cdrom0 for convenience. msgstr , discover /dev/ cdrom /dev/cdrom0. #. Type: boolean #. Description #: ../discover1.templates:4 msgid If you don't want discover to manage the device symbolic links and mount points for you, decline this option. msgstr , discover , . #. Type: string #. Description #: ../discover1.templates:19 msgid Directory for CD-ROM device mount points: msgstr CD-ROM: #. Type: string #. Description #: ../discover1.templates:19 msgid Mounted filesystems accessible via CD-ROM (or compatible) devices can be placed just about anywhere in the filesystem hierarchy, but the standard says they should be placed in /media, so that they are available as, e.g., \/media/cdrom0\, \/media/cdrom1\, and so forth. msgstr , CD-ROM ( ), - , , /media, , , \/media/cdrom0\, \/media/cdrom1\ . . #. Type: string #. Description #: ../discover1.templates:19 msgid Note that discover will link the \cdrom\ mount point to the \cdrom0\ mount point for convenience. msgstr , discover \cdrom\ \cdrom0\. #. Type: string #. Description #: ../discover1.templates:19 msgid To stay compatible with earlier versions of discover the path entered here should end with a / (eg. /media/). msgstr discover, , , / (: /media/). #. Type: note #. Description #: ../discover1.templates:33 msgid Invalid path for mount point msgstr #. Type: note #. Description #: ../discover1.templates:33 msgid You entered an invalid path as CD-ROM devices mount point. msgstr , CD-ROM. #. Type: note #. Description #: ../discover1.templates:33 msgid The base filesystem mount point for CD-ROM drives must be an absolute path (begin with \/\) and must not be null. If the path does not exist, it will be created when needed. msgstr CD-ROM ( \/\) . , . #~ msgid #~ Unless you have a need or preference for placing the CD-ROM mount points #~ elsewhere, the root directory (\/\) is a good choice. #~ msgstr #~ #~ CD-ROM , (\/\) #~ .
Re: Call For Help: Accessibility support in d-i
The problem is that syslinux incorrectly computes amount of data it should read to load a kernel. Attached patch fixes this. With this patch applied I can boot from access floppy. Also tested it with netboot (PXE) and d-i boot.img. I does not understand however why does syslinux works with other kernels without patch. -- Eugeniy Meshcheryakov Kyiv National Taras Shevchenko University Information and Computing Centre http://icc.univ.kiev.ua diff -urN syslinux-2.04/runkernel.inc syslinux-mod/runkernel.inc --- syslinux-2.04/runkernel.inc 2003-04-11 08:44:56.0 +0300 +++ syslinux-mod/runkernel.inc 2004-06-23 10:05:37.0 +0300 @@ -258,8 +258,8 @@ inc si ; plus 1 boot sector shl si,9 ; Convert to bytes mov ecx,8000h ; 32K - sub ecx,esi ; Number of bytes to copy push ecx + sub ecx,esi ; Number of bytes to copy add esi,(real_mode_seg 4); Pointer to source mov edi,10h ; Copy to address 10h
Bug#254630: LVM names
Christian Perrier wrote: Currently unusable in cdebconf (seems to be a whiptail bug in Unicode environments). Just try to enter any non ASCII character in a dialog box..:-( #251550 I can enter cyrillic characters (that are not ASCII) used in Ukrainian in d-i. This looks more like problem with keymap files. -- Eugeniy Meshcheryakov Kyiv National Taras Shevchenko University Information and Computing Centre http://icc.univ.kiev.ua -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#251550: cdebconf: Appears to hang when typing high-ascii characters in string templates
Christian Perrier wrote: Package: cdebconf Severity: normal This happens at least in debian-installerAs soon as one enters a character such as french , no other characters may be inputted and the screen seems hanged. Indeed it is not and TAB works as well as switching to another console. Hitting Ctrl-Q helps getting the keyboard back to work again This may be d-i related or not, no idea.It may be reproduced with a brasilian keyboard with the key also... It looks like you use non-UTF8 keymaps. There is a problem in whiptail when it works with UTF-8, it cannot handle incorect input. -- Eugeniy Meshcheryakov Kyiv National Taras Shevchenko University Information and Computing Centre http://icc.univ.kiev.ua -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#250789: Fix
Recai Oktas wrote: Hi, Knowing that the termwrap is used only in fresh installations, I assume this bug as Turkish spesific. Please find the patch attached. It is not Turkish specific. It loads ISO-8859-1 at the end for most of encodings. It should not do this. It also sets charmaps at the beginning. But this is also done by console-tools/console-cyrillic with parameters from languagechooser's languagelist file. May be termwrap should only setup console for languages without kbd or cyr entries in languagelist (bg - why?, zh, jp, ko, vi - maybe use console-tools for this?)? Regards, -- Eugeniy Meshcheryakov Kyiv National Taras Shevchenko University Information and Computing Centre http://icc.univ.kiev.ua -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Re: status
Nikolai Prokoschenko wrote: In the first dialog, russian entry is still cut after about the half into it - if a better resolution has been chosen (e.g. vga=791), it is not cut anymore. This is result of bug #245982 in cdebconf, not in languagechooser. -- Eugeniy Meshcheryakov Kyiv National Taras Shevchenko University Information and Computing Centre http://icc.univ.kiev.ua -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#249330: lvm root
tags 249330 + patch
thanks
Martin Michlmayr wrote:
lilo-installer's postinst contains the following:
prefix=$(echo $bootfs_devfs | sed 's/\(\/dev\/[a-z]\+\).*/\1/')
disc_offered_devfs=$(find $prefix -name disc | head -n 1)
disc_offered=$(mapdevfs $disc_offered_devfs)
I created root on /dev/vg00/root. So:
prefix=/dev/vg
other assigments will fail.
Thanks. Maybe you can come up with a patch.
Attached patch fixes this problem, I hope.
--
Eugeniy Meshcheryakov
Kyiv National Taras Shevchenko University
Information and Computing Centre
http://icc.univ.kiev.ua
Index: postinst
===
--- postinst (revision 15554)
+++ postinst (working copy)
@@ -20,11 +20,24 @@
df /target$1 | sed '1d;s/ .*//'
}
+lvm_vg () {
+ lvdisplay -c $1| sed 's/^[^:]*:\([^:]*\):.*$/\1/'
+}
+
+first_pv () {
+ pvdisplay -c|grep ^[^:]*:$1: | sed 's/^ *\([^:]*\):.*$/\1/' | head -n 1
+}
+
# detect the partitions /target and /target/boot
rootfs_devfs=$(findfs /)
bootfs_devfs=$(findfs /boot)
rootfs=$(mapdevfs $rootfs_devfs)
+
+bootvg=$(lvm_vg $bootfs_devfs)
+if [ $bootvg ]; then
+ bootfs_devfs=$(first_pv $bootvg)
+fi
bootfs=$(mapdevfs $bootfs_devfs)
prefix=$(echo $bootfs_devfs | sed 's/\(\/dev\/[a-z]\+\).*/\1/')
Bug#249311: (no subject)
Package: installation-reports Debian-installer-version: daily netboot initrd uname -a: The result of running uname -a on a shell prompt Date: 16.05.2004 Method: network install Machine: Description of machine (eg, IBM Thinkpad R32) Processor: Memory: Root Device: IDE? SCSI? Name of device? Root Size/partition table: Feel free to paste the full partition table, with notes on which partitions are mounted where. Output of lspci: Base System Installation Checklist: Initial boot worked:[O] Configure network HW: [O] Config network: [O] Detect CD: [ ] Load installer modules: [O] Detect hard drives: [O] Partition hard drives: [O] Create file systems:[O] Mount partitions: [O] Install base system:[E] Install boot loader:[E] Reboot: [E] [O] = OK, [E] = Error (please elaborate below), [ ] = didn't try it Comments/Problems: I wnated to install Debian with root on LVM. On partitioning step first time i did not found LVM configuration, I expected to see it after all partitions, then I found it on top of the menu. Then I created 1 partition (PV) and VG on it. Creating LV on this VG with default size (all free space) failed. Relevant output of vgdisplay follows: VG Size: 57.27 GB PE Size: 4.00 MB Free PE / Size: 14660 / 57.27 GB There free space = 14660 * 4.00 MB = 58640 MB = 57.265625 GB 57.27 Then I installed base system. This step failed on kernel installation. Output on 3rd console: .../initrd: device /dev/mapper/vg00/root not a block device LVM was not installed in /target Then I reformated discs and reinstalled base system. Before kernel installation i did the following: apt-install lvm2 chroot /target mount /proc /proc -tproc vgmknodes Then installed kernel. Installation of LILO failed too. I manualy edited lilo-installer.postinst and set disc_offered_devfs to /dev/ide/host0/bus0/target0/lun0/disc disc_offered to /dev/hda Then I rerun lilo-installer and choosen /dev/hda from list. After system boot failed with the following messages on console: No volume groups found ... /scripts/ext3-add-journal.sh: 32: arith: syntax error: 0x ... /sbin/init: 417: arith: syntax error: 0x Then kernel panic... -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#249311: (no subject)
Martin Michlmayr wrote: reassign 249311 lvmcfg retitle 249311 lvm on root problems, please run vgmknodes thanks * Eugeniy Meshcheryakov [EMAIL PROTECTED] [2004-05-16 19:58]: Installation of LILO failed too. I manualy edited lilo-installer.postinst and set disc_offered_devfs to /dev/ide/host0/bus0/target0/lun0/disc disc_offered to /dev/hda Then I rerun lilo-installer and choosen /dev/hda from list. Hmm, this doesn't seem LVM related at all. Do you also get this when root is not on LVM? Code that set this variables tries to detect disc device with given partition device. I did not get this when root is not on LVM. After system boot failed with the following messages on console: No volume groups found /scripts/ext3-add-journal.sh: 32: arith: syntax error: 0x /sbin/init: 417: arith: syntax error: 0x Then kernel panic... This looks like #246562 which is apparently fixed in initrd-tools (0.1.68). Can you do an installation of unstable rather than testing to see if it's fixed? I installed unstable and have initrd-tools_0.1.68_all.deb in apt-proxy cache. -- Eugeniy Meshcheryakov Kyiv National Taras Shevchenko University Information and Computing Centre http://icc.univ.kiev.ua -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#249311: (no subject)
Martin Michlmayr wrote: * Eugeniy Meshcheryakov [EMAIL PROTECTED] [2004-05-16 20:32]: Hmm, this doesn't seem LVM related at all. Do you also get this when root is not on LVM? Code that set this variables tries to detect disc device with given partition device. I did not get this when root is not on LVM. *sigh* Do you know if that's a bug in lilo or lilo-installer? This is a bug in lilo-installer. It should not find for 'disc' device if partition is LVM LV. It should not also try to install lilo to LVM LV but, only to one of the disc devices or partition (not LVM LV). Do you know if grub works? As I know grub cannot be installed on LVM LV. -- Eugeniy Meshcheryakov Kyiv National Taras Shevchenko University Information and Computing Centre http://icc.univ.kiev.ua -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#249311: lvm root
Martin Michlmayr wrote: * Eugeniy Meshcheryakov [EMAIL PROTECTED] [2004-05-16 21:11]: This is a bug in lilo-installer. It should not find for 'disc' device if partition is LVM LV. It should not also try to install lilo to LVM LV but, only to one of the disc devices or partition (not LVM LV). Hold on, what was disc_offered_devfs and disc_offered before you modified it? lilo-installer's postinst contains the following: prefix=$(echo $bootfs_devfs | sed 's/\(\/dev\/[a-z]\+\).*/\1/') disc_offered_devfs=$(find $prefix -name disc | head -n 1) disc_offered=$(mapdevfs $disc_offered_devfs) I created root on /dev/vg00/root. So: prefix=/dev/vg other assigments will fail. -- Eugeniy Meshcheryakov Kyiv National Taras Shevchenko University Information and Computing Centre http://icc.univ.kiev.ua -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
Bug#248951: en is not a valid locale, but is added to locale.gen
Nikolai Prokoschenko wrote: On Fri, May 14, 2004 at 01:26:38PM +0200, Frans Pop wrote: Well it has the check now, but no fallback. Isn't LANG=C the proper fallback? Don't think so, at least it shouldn't be the first fallback. I think the best would be to check for the default language locale - if ru_DE is not available, take whichever ru_* is the default - i.e. ru_RU. There is other problem for Russian: ru_RU is ISO-8859-5 locale, I do not think it is that most Russians expect. So it shold be ru_RU.KOI8-R. If not, one should change charset in console configuration (but other locale - ru_UA - have KOI8-U charset). -- Eugeniy Meshcheryakov Kyiv National Taras Shevchenko University Information and Computing Centre http://icc.univ.kiev.ua -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of unsubscribe. Trouble? Contact [EMAIL PROTECTED]
