Re: [Pkg-kbd-devel] Switching d-i to kbd for squeeze?
On 28 Jan 2009, at 13:34, Michael Schutte wrote: d-iers, On Wed, Jan 28, 2009 at 11:52:42AM +, Colin Watson wrote: Can we switch from console-tools to kbd for squeeze? These days kbd is better maintained (I know they've gone back and forth in the past), and its setfont program supports a wider range of fonts than does consolechars. Doing this would let us get rid of some grottiness in localechooser's finish-install script where we make automatic modifications to a conffile (!). This is largely independent of the work on switching to console- setup, which still needs a set of underlying console utilities. If there's consensus, I can commit the relevant fixes. No objections from me as a kbd maintainer. Strong agreement from me as console-tools maintainer. -- Michael Schutte ___ Pkg-kbd-devel mailing list pkg-kbd-de...@lists.alioth.debian.org http://lists.alioth.debian.org/mailman/listinfo/pkg-kbd-devel Regards, Alastair -- Alastair McKinstry , http://blog.sceal.ie Anyone who believes exponential growth can go on forever in a finite world is either a madman or an economist - Kenneth Boulter, Economist. -- To UNSUBSCRIBE, email to debian-boot-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
Bug#388464: debian-installer: ~2.6.17-2 amd64 kernel appears to be missing lvm-mod
Frans Pop wrote: > reassign 388464 debian-installer > tags 388464 + unreproducible > thanks > > On Wednesday 20 September 2006 16:34, Alastair McKinstry wrote: > >> Installing a Dell 1950 Server (x86_64), LVM support is broken as there >> is no LVM support in the kernel (-8 or -9, I believe), and no LVM >> modules available. >> >> This worked with the i386 installer of the same build (20060920) >> > > LVM works fine for me using current daily AMD64 image using kernel udebs > based on -9. > > Please be more specific as to what is failing. > Is it just that you do not see the LVM option in the main partman menu? If > so, that is because an interface change. Please see the installation > guide on alioth. > Sorry about the delay in replying; the servers are at work, so I had to wait until today. This was the d-i netinst amd64 image for 20060920, on Dell PowerEdge 1950 hardware; there are two SATA disks configured with softraid (RAID1) across the two disks, and then LVM on these, so the path was: - netinst boot. - from partitioner, select Configure RAID , then "Finish", so that RAID configuration is spotted. Then - "Configure LVM" to detect the LVM configuration. This fails with "LVM could not be configured; you may need lvm-mod. " In a seperate console, lsmod shows no lsm modules. The syslog does show: "dm_mod: Unknown symbol idr_replace" Now, this morning I repeated this with the same CD and with the weekly image amd64 binary-1.iso : the bug is no longer present on the 20060918 weekly image. Todays daily netinst image for amd64 failed to spot the CD, and so I could not test this far; I am investigating this further before submitting a bug report. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Bug#388464: debian-installer: ~2.6.17-2 amd64 kernel appears to be missing lvm-mod
Package: debian-installer Version: Daily build 20060920 Severity: normal Installing a Dell 1950 Server (x86_64), LVM support is broken as there is no LVM support in the kernel (-8 or -9, I believe), and no LVM modules available. This worked with the i386 installer of the same build (20060920) - Alastair -- System Information: Debian Release: testing/unstable APT prefers testing APT policy: (500, 'testing') Architecture: powerpc (ppc) Shell: /bin/sh linked to /bin/bash Kernel: Linux 2.6.8-powerpc Locale: LANG=C, LC_CTYPE=C (charmap=ANSI_X3.4-1968) -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Bug#379938: Crashes with libnewt0.52 (0.52.2-5.1) if libfribidi0 not installed
Christian Perrier wrote: > Quoting Andree Leidenfrost ([EMAIL PROTECTED]): > >> Package: libnewt0.52 >> Version: 0.52.2-5.1 >> Severity: important >> >> Dear Alistair, >> >> In version 0.52.2-5.1, the crashes in case libfribidi0 is not installed >> have returned. I am attaching a full backtrace of a non-stripped program >> (mondoarchive) using a non-stripped libnewt0.52 (0.52.2-5.1). >> >> I believe this is really the same as #351367 and #290722. >> >> 0.52.2-5 is fine, the problem only starts with 0.52.2-5.1. Also as >> before, if libfribidi0 is installed, the problem goes away. >> >> I presume the underlying reason is (from the changelog): >> >> [...] >> * Apply patch from Eugeniy Meshcheryako to regenerate configure during build >> so that bidi support is enabled again. Closes: #379566 >> [...] >> >> Please let me know if you need any other information. >> > > > Please note that this latter change was needed by D-I, so any solution > reverting that change is not acceptable. > > I guess that a short-term solution would be making libnewt0.52 depend > on libfribidi0but that'd virtually make the latter part of the > base system. > > > The NMU patch puts the fribidi code back into Newt; unfortunately thats not where it belongs: it somehow got dropped from libslang2. It was moved from newt to slang (the underlying library on which newt is based) in 2004; I need to do some work to figure out whats happened. Working on this as quickly as possible, Alastair -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Debian mactel linux support?
Frans Pop wrote: > On Thursday 29 June 2006 17:57, Junichi Uekawa wrote: > >> Debian-installer: >> elilo-installer is not built for ia32 >> (http://lists.debian.org/debian-boot/2006/06/msg00185.html) >> >> Anyone already working on these stuff? >> > > There has been some work done on the installer (including uploading elilo > installer for i386 to the archive...). > There does not seem to be a real drive behind this though. Owners of such > machines are welcome to jump in and contact us on the debian-boot list. > > Cheers, > FJP > Well, I have an intel MacMini, and a copy of Bootcamp to play with. I've been waiting for DSL to work at home (which it now does, yay!) before joining in again. So I'll give it a try. Regards Alastair McKinstry -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
libnewt transition from 0.51 to 0.52
Hi, You are receiving this mail because you maintain a package that depends on libnewt0.51. This has now moved to 0.52 in upstream, and the versioned symbols patch from Debian was applied with the 0.52 tag, requiring that all packages depending on libnewt0.51 be rebuilt to depend on libnewt0.52 instead. It would be appreciated if you could rebuild your package accordingly. Maintainers with packages depending on libnewt0.51, according to whodepends: John Lines <[EMAIL PROTECTED]> (plptools) Alastair McKinstry <[EMAIL PROTECTED]> (libnewt-dev libnewt0.52 newt-tcl python-newt whiptail) Debian Install System Team (cdebconf) Andree Leidenfrost <[EMAIL PROTECTED]> (mindi-partimagehack mondo) Michael Biebl <[EMAIL PROTECTED]> (partimage partimage-server) Chris Lawrence <[EMAIL PROTECTED]> (lokkit) Noèl Köthe <[EMAIL PROTECTED]> (kudzu) Marek Habersack <[EMAIL PROTECTED]> (pike7.2-pexts-newt pike7.4-pexts-newt pike7.6-pexts-newt) Debian QA Group <[EMAIL PROTECTED]> (libnewt-perl) Debian VoIP Team <[EMAIL PROTECTED]> (asterisk-bristuff asterisk-classic zaptel) Regards Alastair McKinstry -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Bug#284455: debian-installer: linux26 fails to detect CD-ROM on Dell PowerEdge 750
Package: debian-installer Severity: normal On a Dell PowerEdge 750, normal d-i installs fine, but 2.6 install via 'linus26' fails to detect the CD-ROM. CD-ROM works fine when 2.6 installed later; modules lists and PCI output attached below. Regards Alastair McKinstry -- System Information: Debian Release: 3.1 APT prefers testing APT policy: (500, 'testing') Architecture: i386 (i686) Kernel: Linux 2.6.8-1-686 Locale: [EMAIL PROTECTED], [EMAIL PROTECTED] (charmap=ISO-8859-15) :00:00.0 Host bridge: Intel Corp. 82875P Memory Controller Hub (rev 02) :00:03.0 PCI bridge: Intel Corp. 82875P Processor to PCI to CSA Bridge (rev 02) :00:1c.0 PCI bridge: Intel Corp. 6300ESB 64-bit PCI-X Bridge (rev 02) :00:1d.0 USB Controller: Intel Corp. 6300ESB USB Universal Host Controller (rev 02) :00:1d.1 USB Controller: Intel Corp. 6300ESB USB Universal Host Controller (rev 02) :00:1d.4 System peripheral: Intel Corp. 6300ESB Watchdog Timer (rev 02) :00:1d.5 PIC: Intel Corp. 6300ESB I/O Advanced Programmable Interrupt Controller (rev 02) :00:1d.7 USB Controller: Intel Corp. 6300ESB USB2 Enhanced Host Controller (rev 02) :00:1e.0 PCI bridge: Intel Corp. 82801 PCI Bridge (rev 0a) :00:1f.0 ISA bridge: Intel Corp. 6300ESB LPC Interface Controller (rev 02) :00:1f.2 IDE interface: Intel Corp. 6300ESB SATA Storage Controller (rev 02) :00:1f.3 SMBus: Intel Corp. 6300ESB SMBus Controller (rev 02) :01:01.0 Ethernet controller: Intel Corp. 82547GI Gigabit Ethernet Controller :03:02.0 Ethernet controller: Intel Corp. 82541GI/PI Gigabit Ethernet Controller :03:0e.0 VGA compatible controller: ATI Technologies Inc Rage XL (rev 27) Module Size Used by ipip 10660 0 af_packet 22600 4 ipv6 264612 48 ipt_REJECT 7008 3 ipt_REDIRECT2208 1 ipt_MASQUERADE 3968 1 iptable_nat25228 3 ipt_REDIRECT,ipt_MASQUERADE ip_conntrack 35368 3 ipt_REDIRECT,ipt_MASQUERADE,iptable_nat iptable_filter 2880 1 ip_tables 18464 5 ipt_REJECT,ipt_REDIRECT,ipt_MASQUERADE,iptable_nat,iptable_filter usbhid 32192 0 uhci_hcd 33136 0 shpchp101996 0 pciehp 99084 0 pci_hotplug34640 2 shpchp,pciehp intel_agp 22816 1 intel_mch_agp 10608 0 agpgart34696 2 intel_agp,intel_mch_agp pcspkr 3592 0 tsdev 7392 0 mousedev 10476 0 psmouse20360 0 floppy 61200 0 evdev 9600 0 ehci_hcd 32004 0 usbcore 119012 5 usbhid,uhci_hcd,ehci_hcd e1000 85860 0 capability 4520 0 commoncap 7232 1 capability ide_cd 42656 0 ide_core 139908 1 ide_cd cdrom 40732 1 ide_cd rtc12760 0 isofs 37240 0 ext2 71848 1 ext3 127240 4 jbd62616 1 ext3 mbcache 9348 2 ext2,ext3 sd_mod 21696 6 ata_piix8164 5 libata 41700 1 ata_piix scsi_mod 125196 2 sd_mod,libata unix 28692 209 font8320 0 vesafb 6656 0 cfbcopyarea 3840 1 vesafb cfbimgblt 3040 1 vesafb cfbfillrect 3776 1 vesafb
Bug#284454: partman loops on "Go back" with preseed file
Package: partman Severity: normal Given a pre-seed file (see attached), partman presented a list of filesystems to create, which I wanted to edit. So instead of creating the filesystems, I selected "Go Back" or "No". Both threw loops: selecting "No" throws you back to the same "Write changes" screen, while "Go back" continually restarts partman. -- System Information: Debian Release: 3.1 APT prefers testing APT policy: (500, 'testing') Architecture: i386 (i686) Kernel: Linux 2.6.8-1-686 Locale: [EMAIL PROTECTED], [EMAIL PROTECTED] (charmap=ISO-8859-15) # Verify password slapd slapd/password2 passwordadmin # Root password: passwd passwd/root-passwordpassword # encrypted admin password slapd slapd/internal/adminpw passworduWbgieD4M.RXs # Admin password slapd slapd/password1 passwordadmin # Re-enter password to verify: passwd passwd/root-password-again password # Re-enter password to verify: passwd passwd/user-password-again password # Enter a password for the new user: passwd passwd/user-passwordpassword # Do you want system wide readable home directories? adduser adduser/homedir-permission boolean true # Are you in the Europe/Dublin time zone? base-config tzconfig/choose_country_zone_single boolean true # What would you like to do about ? ucf ucf/changepromptselect keep your currently-installed version # Which specific keymap do you want ? console-dataconsole-data/keymap/qwertz/german/standard/keymap select # Do you want to handle the daemon configuration with debconf? clamav-daemon clamav-daemon/debconf boolean true # Which specific keymap do you want ? console-dataconsole-data/keymap/qwertz/swiss/german/keymap select # Which specific keymap do you want ? console-dataconsole-data/keymap/qwertz/german/apple_usb/keymap select # Which specific keymap do you want ? console-dataconsole-data/keymap/qwerty/estonian/standard/keymap select # Which variant do you have ? console-dataconsole-data/keymap/qwerty/swedish/variant select # Correct master.cf for upgrade? postfix postfix/master_upgrade_warning boolean # Which specific keymap do you want ? console-dataconsole-data/keymap/qwertz/serbian/standard/keymap select # Prepare /etc/resolv.conf for dynamic updates? resolvconf resolvconf/linkify-resolvconf boolean true # Which variant do you have ? console-dataconsole-data/keymap/qwertz/serbian/variant select # What is the keys layout of your keyboard ? console-common console-data/keymap/template/layout select console-dataconsole-data/keymap/template/layout select # Select a city or time zone: base-config tzconfig/select_zoneselect # What is the keys layout of your keyboard ? console-dataconsole-data/keymap/qwerty/layout select # Fix LDAP directory on upgrade slapd slapd/fix_directory boolean true # What is the keys layout of your keyboard ? console-dataconsole-data/keymap/azerty/layout select # Organisational Unit Name ssl-certmake-ssl-cert/ounamestring One Organization Unit # Retry configuration slapd slapd/invalid_configboolean true # Is this information correct? base-config tzconfig/verify_choices boolean true # Host Name ssl-certmake-ssl-cert/hostname string localhost # Which specific keymap do you want ? console-dataconsole-data/keymap/qwerty/lithuanian/standard/keymap select # Virus database update method: clamav-freshclamclamav-freshclam/autoupdate_freshclam select cron # Which variant do you have ? console-dataconsole-data/keymap/azerty/french/variant select # Enter the name of your organization slapd shared/organization string Ildana # Select your time zone: base-config tzconfig/choose_country_zone/BR select East # On what network interfaces should the DHCP server listen? dhcp3-serverdhcp3-server/interfaces string eth0 # Which variant do you have ? console-dataconsole-data/keymap/qwerty/finnish/variant select # Overwrite existing /etc/aliases? exim4-configexim4/dc_noalias_regenerate boolean false # Run MRTG with its own user? mrtgmrtg/own_user boolean true # Fix permissions of cache_dir? squid squid/fix_cachedir_permsboolean false # Which variant do you have ? console-dataconsole-data/keymap/qwertz/czech/variantselect # Should tcpd setup paranoid hosts.allow and hosts.access? tcpdtcpd/paranoid-mode boolean false # Do you want to ignore PCI display devices? hotplug hotplug/ignore_pci_class_displayboolean true # Which specific keymap do you want ? console-dataconsole-data/keymap/qwerty/portugese/standard/keymapselect # Should Postfix upgrade hash and btree maps? postfix postfix/db_upgrade_warning boolean true # Which specific keymap do you want ? console-dataconsole-data/keymap/fggiod/tu
Proposed changes to console-data
Hi, The following changes are in console-data SVN repository: console-data (2002.12.04dbs-45sarge2) UNRELEASED; urgency=low * Translations: - Added Macedonian (mk) by Georgi Stanojevski - Corrected encoding error in French. Closes: #277569 * Select br-abnt2 as the default Brazilian keymap. Closes: #275273. * Added Lithuanian keymap to console-keymaps-at, console-keymaps-acorn for debian-installer. Closes: #275087. * Fixed dangling symlink in console-keymaps-at; thanks to Recai Oktas. Closes: #262693. -- Alastair McKinstry <[EMAIL PROTECTED]> Thu, 21 Oct 2004 20:34:31 +0100 In addition, I propose to add the following patches for new keymaps which are in the BTS: #267380 : Persian/Iranian keymap #267451 : Arabic #273461 : Greek UTF-8 To avoid problems with increasing d-i size, and because there isn't time to translate them, they would go into the 'extended' console-data set only, not into console-keymaps-at for d-i. console-data can be built quickly (its arch:any) but it impacts d-i, so the release managers have the final say -- is it too late to make these changes? Regards Alastair McKinstry -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Bug#274364: #274364: please confirm that this bug is fixed
Hi, I am clearing old bugs, and it appears that one you submitted, #274364, about USB keyboard not working, has been fixed. Can you please confirm this? Regards Alastair McKinstry -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Bug#276752: Urgent patch for kbd-chooser
-Original Message- >From: "Christian Perrier"<[EMAIL PROTECTED]> >Sent: 18/10/04 22:40:17 >To: "Alastair McKinstry"<[EMAIL PROTECTED]> >Cc: "Recai Oktas"<[EMAIL PROTECTED]>, "[EMAIL PROTECTED]"<[EMAIL PROTECTED]>, "Denis Barbier"<[EMAIL PROTECTED]>, "[EMAIL PROTECTED]"<[EMAIL PROTECTED]> >Subject: Re: Bug#276752: Urgent patch for kbd-chooser >Quoting Alastair McKinstry ([EMAIL PROTECTED]): >> On Luan, 2004-10-18 at 23:41 +0300, Recai Oktas wrote: >> > Small rearrangement of the patch. Tested successfully for Turkish and >> > this time Latvian under Vmware. Sorry for my aesthetics obsessions :-) >> > >> > Regards, >> > >> >> Patch looks good to me; Should work with Lithuanian, which was broke >> due to #275087, just fixed in console-data SVN. >> >> Committing to kbd-chooser SVN; will leave the decision to include to >> upload to Joey. > > >It's commited already. > >I have used the very last version of Recai patch, at the time I write >this. So I discovered ;-) Thanks, Alastair > -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Bug#276752: Urgent patch for kbd-chooser
-Original Message- >From: "Christian Perrier"<[EMAIL PROTECTED]> >Sent: 18/10/04 22:40:17 >To: "Alastair McKinstry"<[EMAIL PROTECTED]> >Cc: "Recai Oktas"<[EMAIL PROTECTED]>, "[EMAIL PROTECTED]"<[EMAIL PROTECTED]>, "Denis Barbier"<[EMAIL PROTECTED]>, "[EMAIL PROTECTED]"<[EMAIL PROTECTED]> >Subject: Re: Bug#276752: Urgent patch for kbd-chooser >Quoting Alastair McKinstry ([EMAIL PROTECTED]): >> On Luan, 2004-10-18 at 23:41 +0300, Recai Oktas wrote: >> > Small rearrangement of the patch. Tested successfully for Turkish and >> > this time Latvian under Vmware. Sorry for my aesthetics obsessions :-) >> > >> > Regards, >> > >> >> Patch looks good to me; Should work with Lithuanian, which was broke >> due to #275087, just fixed in console-data SVN. >> >> Committing to kbd-chooser SVN; will leave the decision to include to >> upload to Joey. > > >It's commited already. > >I have used the very last version of Recai patch, at the time I write >this. So I discovered ;-) Thanks, Alastair > -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Bug#276752: Urgent patch for kbd-chooser
On Luan, 2004-10-18 at 23:41 +0300, Recai Oktas wrote: > Small rearrangement of the patch. Tested successfully for Turkish and > this time Latvian under Vmware. Sorry for my aesthetics obsessions :-) > > Regards, > Patch looks good to me; Should work with Lithuanian, which was broke due to #275087, just fixed in console-data SVN. Committing to kbd-chooser SVN; will leave the decision to include to upload to Joey. Thanks to all involved; I've been busy recently starting a new job, and off the net more than I'd planned. Clearing the backlog of bugs now. Regards Alastair -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: newt and slang issues
>> I will upload the slang change to testing-proposed-updates >> as soon as the sid package is accepted into the archive; >> is the newt translation significant enough for t-p-u ? > >Please upload to unstable only; the release team can push the package to >testing from there once it's built on all archs. Ok, I won't upload then. Should the bugs be re-opened then and tagged RC and sarge to track this? The current console-tools, console-data, console-common in sid fix RC issues and should be pushed into sarge. >Thanks, >-- >Steve Langasek >postmodern programmer > Regards Alastair -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
newt and slang issues
Hi, I've uploaded to sid packages to fix the slang crashes (patch thanks to Bastian) and the newt Bulgarian translation. (don't know what your FTBFS issue was - I couldn't reproduce). I will upload the slang change to testing-proposed-updates as soon as the sid package is accepted into the archive; is the newt translation significant enough for t-p-u ? Incidentally, I've a new bug in iso-codes asking for the common name of F.Y.R. Macedonia to be changed to "Macedonia". I think I'll leave that to post-sarge ... Regards Alastair -- 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?
>On Tue, Jul 27, 2004 at 09:55:39AM +0100, Alastair McKinstry wrote: >> >> I think some of the stuff in console-cyrillic should be merged >> into console-common post-sarge; console-common is due for a >> rewrite: mostly to debconf' the font, etc. settings, >> but also make it more robust. >I see that many of the fonts in console-data (not only the Cyrillic) >have multiple variants for different encodings. Is this intentional or >due to the history of the package? All fonts in console-cyrillic have >different faces and use some custom artificial encogings. The real >encoding is chosen by ACM. The collection of fonts and keymaps in console-data is mostly historical rather than planned. >> Conversely, some of the cyrillic fonts/keymaps in >> console-data may be better off in console-data; > >It is possible to optimize the fonts and keymaps in console-data. The >problem is that all maintainers of this package added stuff to it but >only on request and after negotiations. As a result some of the files >in console-data are rarely used and some languages and keyboards are not >supported at all >I think we should not base the debconf questions on the existing data >but develop them in some generic way. Then the files that are necessary >to support the configurations handled via debconf should be in >console-data. All other files are obviously rarely used and should be >moved in other packages. > >> the idea should be console-common as standard, with >> console-tools | kbd choice, and console-data | console-cyrillic. > >At present console-cyrillic is an alternative of console-data. In >sarge+1 I'd like to see this package not as an alternative but as a >complement - all important stuff is in console-data and console-cyrillic >is one of the packages containing the less used data. > >Another variant is this: two packages console-data-base and >console-data. The first contains the important data and the second more >rarely used files. In this way console-cyrillic can be obsoleted. I favour minimising the amount of console data in base, but console-data should always be present: think of a blade system without console, but someone hotplugging a monitor and keyboard in; for most of its life we don't need all the data currently in console-data, but some of it we'd need in a hurry to debug issues. There is also the Unicode data in /usr/share/unidata; thats a 1.4 MiB chunk that we rarely need .. I think console data should be optimised into console-data, console-data-optional, console-cyrillic. >Ofcourse in both of these packages the data should be well organized. If >you want I can help with this. I am not happy with the current state of >console-data. Help appreciated, especially on the fonts end. I had not planned on much work for fonts for sarge+1; If I get console-common / keyboards sorted out in a 6 month release cycle, I'll be very happy. >> loadkeys should work with use X keymaps. > >This is great! Do you know who will work for this and how it will be >implemented? I'd volunteered for this. Apparently the X code is now modular enough that I can dynamically link in code to parse the X keymaps. I intend to do this: - separate out the keymap parsing code from loadkeys / kbd-chooser. Move it into libconsole. One code base. - Create a wrapper to load the X keymap parsing code in the same way, so loadkeys can load X keymaps. - "Merge" the console-data and X keymaps, so that we only need one set eventually. With loadkeys / dumpkeys we can dump the keymap (to /etc/console/boottime.kmap.gz) so that we don't need X keymaps or X to boot.. but we only need one keymap selection (X's) and to set the keymap once (possibly speeding up booting into X). >Anton Zinoviev > Regards Alastair -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: New d-i devcamp?
>also sprach Petter Reinholdtsen <[EMAIL PROTECTED]> [2004.07.26.2352 +0200]: >> OK. 5 people confirmed their interest in a d-i devcamp. > >Please add me to the list of interested parties. Ditto. Regards Alastair McKinstry. -- 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?
Hi,
Sorry I'm a bit late to the thread;
I think a bit of reorganisation of console-* would be a
good idea, post sarge. (Note, with a freeze planned for July 31,
I'm working towards fixing bugs on console-* packages before
then; no disruptive stuff right now).
I've posted some notes on plans I had made at
http://people.debian.org/~mckinstry/console-plans.html
Two main bits:
As there appears to be little interest in console-{tools,data}
development upstream, I'm working towards merging stuff back
to kbd.
I think some of the stuff in console-cyrillic should be merged
into console-common post-sarge; console-common is due for a
rewrite: mostly to debconf' the font, etc. settings,
but also make it more robust.
Conversely, some of the cyrillic fonts/keymaps in
console-data may be better off in console-data; the idea should
be console-common as standard, with
console-tools | kbd choice, and console-data | console-cyrillic.
Also, note that 'fonty' contains non-latin fonts that
people may use instead of console-data.
I intend to make kbd-chooser mainstream, as well as d-i;
the keyboard selection should be hotplug, triggered on
keyboard detection. This should be merged with X developments
in this area: loadkeys should work with use X keymaps.
Comments welcome.
Regards
Alastair
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: keyboard autodetection
> > >Hi, > >> > > * Is there a way to autodetect (based on hardware IDs) the keyboard t>ype >> > > (eg., when you install Debian, you have to select keyboard layout and >> > > country/ type). >> > Hmm. What you're asking, I think, is whether the hardware ID implies a >> > layout. (It's not really autodetection; it's a lookup.) I don't know.> Its hard to do keyboard detection based on hardware ID because of the cheap way keyboard manufacturers build keyboards: they do not want to build key layout into the keyboard electronics, but rather just build 'local' keyboards by sticking the keycaps on at the last minute. E.g. it is easy for a manufacturer to assign different USB IDs to different models, so that keymap type could be autodetected in software. But if they eg. make boards for 1000 French azerty keyboards, with a French ID in the 'firmware', they can't just make these into German 'qwertz' keyboards (if they get a demand for German keyboards), by putting different keycaps on them before packaging, which is what they like to do right now. This is irritating as it means kbd-chooser, etc. can't completely autoconfigure the keyboard. For non-AT keyboards, e.g. Sparc, Apple, etc. the prospects are better; see code in kbd-chooser source for reading keyboard type from these from the input layer in 2.6 kernels. (Hint, hint: please move to 2.6 kernels where possible...) >I've thought about this extensively at this year's DebConf, however I've >been unable to implement a real solution. I had to install Debian on one >box in Brazil, which of course had a Brazilian keyboard. As there are >two different variants of "Brazilian" keyboards available, I could not >decide for myself which one was appropriate. > >The idea I have in mind is to generate a "decision tree" from the >available keymap files (flat tree being optimal) and have the user press >different keys in turn to see what keycode they give. The greatest >hinderance is the fact that I'm not allowed to read data from the >keyboard when using DebConf, which is why I have no real idea how to >implement it in a compatible way for d-i, short of adding a new template >type "single-keycode" (which would probably work on the console, but >require additional work if it is supposed to work together with bterm, >which the installer currently uses, or X, which the installer might use >in the future). Agreed; a much better widget, specifically for X use, should be designed for sarge+1. > Simon > - Alastair -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Bug#256405: crash displaying main menu in arabic
DÃ Domh, 2004-06-27 ag 11:29 -0500, scrÃobh Steve Langasek: > On Sat, Jun 26, 2004 at 04:35:42PM -0400, Joey Hess wrote: > > Package: debian-installer > > Severity: normal > > Tags: d-i > > > I booted today's sid_d-i build from a usb stick. At the language > > chooser, I chose the nice new Arabic choice, and then at countrychooser, > > it was at the full list of contries, and I realized I din't speak > > Arabic, and selected the back button. > > Some experimentation and it looks like the problem is just with showing > > the current d-i main menu, localised to Arabic. I can reproduce the same > > problem by booting in expert mode, selecting Arabic, and pressing enter; > > when it tries to draw the main menu (same as it does when you go back > > from countrychooser), it must not like something, and it crashes. I suspect that the arabic shaping code is re-writing the strings and using characters that were not present in the original string. iso-scan saves space by removing all glyphs from unifont.bgf that were not present in the original string; if the code assumes all glyphs are present, it could crash ... - Alastair -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Bug#244978: refile under hotplug
refile 244978 hotplug thanks Refiling this bug under hotplug, as it appears to be the package still in need of work. This bug no longer involves usb-disover thanks Alastair -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Bug#244886: kbd-chooser: can't select keymap on sparc64 netboot 20040411 image
Hi, Can you please confirm whether or not you still see this problem? I do not have access to a sparc machine and cannot debug this without help Regards Alastair -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Bug#244978: is usbutils ever needed?
Hi, I'm working on Debian bug #244978, which asks if usbutils is needed for anything, or will hotplug suffice. In /etc/hotplug/hotplug.functions, there is a note: " usbmodules used for (a) multi-interface devices, (b) coldplug" Do you know if usbutils is needed for multi-interface devices, or will hotplug suffice? Regards Alastair McKinstry -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Review of arcboot-installer template wanted
Hi,
I'm planning to include a patch to arcboot-installer,
for #241178, which posts a reminder note about prom variables.
(Assuming joeyh says ok).
Before I go stressing out all the translators, can people please
review the English version below?
Template: arcboot/prom_variables
Type: note
Description: Setting PROM variables for Arcboot
If this is your first Linux installation on this machine, or if you have
repartitioned your hard drives, you will need to set certain variables
in the PROM before the system will boot normally.
.
At the end of this installation phase, you will reboot your system. When
you do, select "Stop for Maintenance", then press "5" to enter the command
monitor. There, type the following commands:
.
setenv SystemPartition
scsi(${abootbus})disk(${abootid})rdisk(${abootlun})partition(8)
setenv OSLoadPartition
scsi(${arootbus})disk(${arootid})rdisk(${arootlun})partition(${arootpart})
setenv OSLoader arcboot
setenv OSLoadFilename Linux
.
You will only need to do this once. Afterwards, type "boot" or reboot the
system to proceed.
Regards,
Alastair McKinstry
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Bug#250548: Default keyboard for Netherlands should be 'US American'
DÃ Domh, 2004-05-23 ag 20:47 +0200, scrÃobh Frans Pop: > Package: kbd-chooser > Version: 0.50 > > Please change the default keyboard for nl_NL installations from 'Dutch' to 'US > American'? > > Rationale: > Most PC's (at a guess: over 90%) in the Netherlands are equipped with a US > keyboard. Dutch keyboards are available from some suppliers, but have to be > specially ordered in most cases. > > I've also seen this mentioned in one or two installation reports. > Example: #246725 > It has also been discussed on debian-l10n-dutch. > > TIA Ok, This patch is already in the console-data repository . This is expected to be uploaded tomorrow, as the current version in console-data goes into testing. Regards, Alastair > > -- > To UNSUBSCRIBE, email to [EMAIL PROTECTED] > with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED] >
Bug#249164: installation report on DECstation (mipsel)
DÃ Domh, 2004-05-16 ag 16:24 +0100, scrÃobh Martin Michlmayr:
> * Karsten Merker <[EMAIL PROTECTED]> [2004-05-15 16:17]:
> > Keyboard selection: [E] Kbd-Chooser only presents PC keymaps by default.
> > Pressing ESC brings one back to the main menu.
> > On reselection of "choose keyboard" I get a menu
> > asking to choose between DEC, USB and AT
> > keyboards. Selecting "DEC" gives an error. The
> > log says: choose_keymap: keymap = US English
> > ERROR: cannot open file US English
> > I have not yet had time to look into this, but the
> > DEC US keymap is called "LK201-US", so there
> > might be a wrong string passed.
>
> Let's copy Alastair to this.
DEC is the only arch. that has only one keyboard (atm) so it has a
default entry in dec-kbd.c. Found a bug in the deflt keyboard handling.
Try the following fix to kbd-chooser.c and report if it works:
Index: kbd-chooser.c
===
--- kbd-chooser.c (revision 15477)
+++ kbd-chooser.c (working copy)
@@ -687,7 +687,7 @@
// and shouldn't have to ask the question.
if (kb->deflt) {
def = keymap_get (maplist_get (arch), kb->deflt);
- mydebconf_default_set (template, def->description);
+ mydebconf_default_set (template, kb->deflt);
}
res = mydebconf_ask ( kb->deflt ? "low" : "high", template,
&ptr);
if (res != CMD_SUCCESS)
As to being offered DEC, USB and AT keyboards: are USB and AT possible
on decstations ? (eg can you get PCI cards for decstations?) if not, can
I identify decstations by subarch, and eliminate those options?
It appears from the latest 2.6.6 kernel sources that you can identify
lkkbd modules in 2.6 kernels. What is the current status of 2.6 on
decstations and MIPS in general?
If it works, then we can autodetect DEC, AT, USB keyboards on 2.6
systema at least; I can add code to dec-kbd.c to handle this.
Regards
Alastair
Bug#247513: Debian Installer bug #247513: Spanish keymap on powerpc
tags 247513 moreinfo thanks Hi, I'm looking at fixing this bug 'Spanish keymap not working on debian installer, powerpc'. What actually happens when you try to choose 'Spanish' on the keyboard chooser on powerpc? (I do not have a powerpc machine) Thanks Alastair -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Bug#243623: #243623: Tottasten is not a german word
Hi Dennis, Can you respond to this bug report about the German translation of console-data? Thanks, Alastair -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Bug#244886: kbd-chooser: can't select keymap on sparc64 netboot 20040411 image
>Package: kbd-chooser >Severity: normal >Tags: d-i > >Hi, > >I tried 20040411 netboot sparc64 image today, and I used "expert >DEBCONF_PRIORITY=low". > >I selected fr_FR, then France. In kbd-chooser, I had this choice : > >Sun keyboard >Usb keyboard > >I selected "Sun keyboard" but I didn't have the choice to select US >Type5 keymap. So I have 2 choices : english install with right kbd or >french install with wrong kbd. > >Thanks. Thanks for your bug report. Within kbd-chooser, it appears the fr.po entry in kbd-chooser has strange characters for "sunkeymap". Could you please edit this, and try the kbd-chooser again: I am on holiday from tonight and will not have a chance to test this and fix it until Sunday, Regards, Alastair -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Full Debian install impressions and facts
>On 18.IV.2004 at 13:27 (-0500) Branden Robinson wrote: >> >> The main challenge here, as I understand it, is that d-i basically uses >> console-data's paradigm for keyboard description. So what we need is a >> gigantic mapping table in xserver-xfree86.config to translate >> console-data keyboard descriptions into XKB >> Rules/Model/Layout(/Variant?) tuples. > >The package console-cyrillic does the oposite for the Cyrillic >languages. This package is simply an inteligent wrapper arount >`consolechars' and `loadkeys'. It provides the Linux console with >flexible keyboard configuration similar to what xkb provides to X >Window. > >It is possible to extend the functionality of console-cyrillic to >cover non-Cyrillic languages, but this would require manual conversion >of the files in /etc/X11/xkb/symbols to another format. > >Anton Zinoviev One of my post-sarge TODO items for console-tools is to implement xkb keyboard parsing in loadkeys (using X11 libraries). Then merging the two keymap sets, with the X ones being definitive (ie removing separate keymaps for the console). So I would prefer to have the table in console-data and have kbd-chooser / prebaseconfig seed the debconf database for X11. Any comments? What debconf entries would need to be set? - Alastair -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Bug#243059: #243059: back button behavior in kbd-chooser
Hi Joey, Could you please verify this problem still occurs? Testing on the latest business-card image, I can't reproduce. The behavior I see is as follows: Starting with priority=HIGH, main-menu leads you directly to "choose keymap". Selecting Back here brings you to main-menu, with priority=medium. Re-entering kbd-chooser here, we see "Choose architecture". (This is as designed, but it might be nice for the user to have a visual cue explaining that priority has changed, and this is why they are seeing different questions). Here, BACK goes to main-menu, forward to "choose-keymap"; BACK from choose-keymap goes to choose-arch, as expected, and back again goes to main-menu. This is all as expected, and this hasn't changed in months; maybe a bug in main-menu that was fixed? - Alastair -- Alastair McKinstry <[EMAIL PROTECTED]> GPG Key fingerprint = CD45 260A 4546 C3C0 F595 F0F6 4132 BF90 2A38 5C57 He that would make his own liberty secure must guard even his enemy from oppression; for if he violates this duty he establishes a precedent that will reach to himself. - --Thomas Paine -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Bug#243846: (no subject)
>So the stopping and starting of PCMCIA isn't always safe either, after >all. I still wonder what's so unsafe about not stopping PCMCIA during >upgrades. Important devices would disappear during the upgrade: eg. networking, or even worse, storage: e.g I have a Compact Flash PCMCIA holder: my CF cards show up as /dev/hdd... > >-- >Pelle > > Regards, Alastair -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Automated xkb/kde/fonts configuration after install
>On Thursday 15 April 2004 02:54, Denis Barbier wrote: >> As you chose el_GR.UTF-8 as the default locale, you have to find >> a way to provide a working UTF-8 console. If jfbterm works for 2nd >> stage, it could also be used when installation is over, couldn't >> it? > >I thought about it, but IIRC I read in this list that we might change >from console-* packages to a kbd-based solution, which is supposedly >more UTF-8 compliant. All UTF-8 patches to kbd since been applied to console-* ; remaining UTF-8 issues are due to the kernel. (Which I intend to fix if I get the time; other volunteers gratefully accepted) There are issues with newt (used in d-i), for which I have patches but doesn't work within d-i (segfaulting); I'm debugging this as soon as I can. Regards, Alastair -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
[l10n]; change to console-data po files
Hi, I have made one addition to the console-data keymaps, which results in a string addition: msgid "trfu" This is 'translated' in English as msgstr "Turkish (F layout)". Can translators please check out their translations and unfuzzy the translation, or if they wish, just send me the translation to commit? Regards, Alastair -- Alastair McKinstry <[EMAIL PROTECTED]> GPG Key fingerprint = CD45 260A 4546 C3C0 F595 F0F6 4132 BF90 2A38 5C57 He that would make his own liberty secure must guard even his enemy from oppression; for if he violates this duty he establishes a precedent that will reach to himself. - --Thomas Paine -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Bug#242007: countrychooser: Translate the list of country names
Package: countrychooser Version: 0.013 Severity: normal Please translate the short list of country names, as well as the full country names list. -- System Information: Debian Release: testing/unstable APT prefers experimental APT policy: (990, 'experimental'), (500, 'unstable') Architecture: i386 (i686) Kernel: Linux 2.6.4-1-686 Locale: [EMAIL PROTECTED], [EMAIL PROTECTED] (ignored: LC_ALL set to ga_IE.UTF-8) -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Must debian-installer use iso-codes package when choosing country?
DÃ Domh, 2004-04-04 ag 02:53 +0200, scrÃobh Frans Pop: > I really think that using the 'really' short names (that is just plain Taiwan > instead of Taiwan, P.. of C..) would not be a bad compromise despite what the > official so called short UN names say. Agreed the "official short names" are ugly ; (http://www.iso.org/iso/en/prods-services/iso3166ma/02iso-3166-code-lists/list-en1.html) eg. "LAO PEOPLE'S DEMOCRATIC REPUBLIC", "LIBYAN ARAB JAMAHIRIYA" for those countries normally called "Laos" and "Libya". However, sometimes the short version is the problem too: eg. do we allow "MACEDONIA, THE FORMER YUGOSLAV REPUBLIC OF", to be shortened to "Macedonia"? I'm sure the Greeks won't agree to that ... > After all, isn't Linux for a large part about being "free" as in "able to > choose for yourselves" which is what Taiwan has been trying to to for the > past decades. Ironically, one of the main reasons I created the iso-codes package is to allow this; if someone wanted to create "Kurdish Linux" and add Kurdistan as a territory, then they would only have 1 list to override or correct on Linux, rather than n separate lists of countries and translations ... (But I don't want to fork Debian over this issue :-( ) > I think the really short names are often a lot more politically neutral than > the semi-official names in iso-3166. The countries in question have used the list to make their own political points; but the problem is that Debian deciding to change some names and not others is no longer being neutral. > FJP Regards, Alastair -- Alastair McKinstry <[EMAIL PROTECTED]> GPG Key fingerprint = CD45 260A 4546 C3C0 F595 F0F6 4132 BF90 2A38 5C57 He that would make his own liberty secure must guard even his enemy from oppression; for if he violates this duty he establishes a precedent that will reach to himself. - --Thomas Paine
Re: Re: Must debian-installer use iso-codes package when choosing country?
DÃ Domh, 2004-04-04 ag 04:17 +0800, scrÃobh Tetralet:
> My original question is:
>
> "Must debian-installer use "iso-codes" package when choosing country,
> Even though it will hurt some people, like people who living in Taiwan?"
>
> And I think your answer is "YES".
> People who install Debian via debian-installer will find that "Taiwan"
> was replaced with "Taiwan, Province of China".
> They may not detect immediately that this country code is based on ISO-3166,
> But they will think that Debian is discriminating against Taiwan people.
> It will not be a very good first impression to Debian users.
As has been noted in previous discussions on this subject, it is
basically impossible to draw up a list of countries and their names that
does not insult somebody. For Debian, picking an official list, (and
ISO-3166 is _the_ offfical list) is pretty much the only solution.
Please update the zh_TW entries, as discussed before. D-I is designed
that the first question people see is the Language; if they pick zh_CN
they should _not_see Province_of_China.
Regards,
Alastair
>
> Tetralet
>
>
> ==
> ~2UUA{150U
> http://edm-prg.epaper.com.tw/click.php?ad_code=87098
> ==
> PChome~~ \*^o^*//
> http://love.pchome.com.tw/
> ==
>
>
> --
> To UNSUBSCRIBE, email to [EMAIL PROTECTED]
> with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
--
Bug#241838: d-i Turkish keyboard issues
DÃ DÃar, 2004-04-01 ag 01:15 -0700, scrÃobh [EMAIL PROTECTED]: > [My previous postings were probably blocked due to your spam > configuration, so I am reposting from a different domain] > > Hi, > Sorry, it was my fault; I've been setting up spam filters and evolution-1.5 from experimental; some mail got lost. > Default Turkish keymap 'trq.kmap' doesn't fully work in bterm. I > couldn't type Turkish spesific characters (all other non-turkish > spesific ascii chars are okay). Though this is not a critical problem, > I want to propose a solution. The problem stems from the fact that > trq.kmap was not designed for UTF-8. There is another Turkish Q kmap > --trqu.kmap for this purpose. > > I create a simple netboot iso with a modified console-data udeb which > uses the trqu.kmap as default and it works at least in bterm (under > C.UTF-8 locale). There is also a minor problem for this case. > In the second console (Alt-F2) Turkish characters can be typed, but can > not be properly displayed, since we don't have unifont where it exists > in bterm frame buffer. But I could live with this problem. > > The real problem with this solution is that, I don't want to use UTF-8 > mode in virtual consoles after the installation is completed. There is > a lot of problem in tr_TR.UTF-8 locale. Man files can't be properly > displayed in UTF-8, readline has problems and the most important one, > there are a few serious bugs in glibc. > > I couldn't perform a full d-i installation for some other d-i bugs, so > couldn't test the whole case. Before submitting a patch, I want to > know your suggestions. Is there a mechanism just to use Unicode keymap > (trqu.kmap) for the early stages of d-i, but install the ISO keymap > (trq.kmap) in the later stage. As you can see in the language-chooser > module, we've already set the default locale as tr_TR which corresponds > to tr_TR.ISO-8859-9 locale. If we install unicode trqu.kmap by default, > it wouldn't work under this locale. > > As you may notice the keymap/fgGIod directory, there is another keyboard > layout in Turkish, known as Turkish F. I would want to add it to > kbd-chooser templates, but the same UTF-8 problem also exists here. So > before creating a patch, we should solve the above problems. > > Hope I described the whole picture, let me know if you need additional > info. Thanks for your help. > > Cheers, > > -- > roktas Ok, the best solution that appears to me is (1) Change the default keyboard in kbd-chooser for Turkish to trqu.kmap. (2) In kbd-chooser prebaseconfig, we can change the settings passed to the second stage; I can add code there to change trqu -> trq for the second stage. I will need to test this, though. Regards, Alastair McKinstry --
Bug#241838: kbd-chooser: Need non-UTF8 keymap for Turkish
Package: kbd-chooser Version: 0.47 Severity: normal For Turkish, we want the trqu UTF8 keymap for the first stage of the install, with a non-UTF-8 keymap trq on the second stage. -- System Information: Debian Release: testing/unstable APT prefers experimental APT policy: (990, 'experimental'), (500, 'unstable') Architecture: i386 (i686) Kernel: Linux 2.6.4-1-686 Locale: [EMAIL PROTECTED], [EMAIL PROTECTED] (ignored: LC_ALL set to ga_IE.UTF-8) -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
speakup and debian-installer
Hi, I'm the kbd-chooser maintainer for the new Debian Installer. I'm working on bug #239385, which appears to have problems with kbd-chooser and console-tools due to speakup not being installed; from looking at the code it appears that for the speakup keymaps to work, you need a speakup-patched kernel; is this the case? In which case, why are the speakup patches not in the mainline Debian kernels (just a question; don't want to provoke a flamewar). d-i uses a default kernel for each architecture; it may then install onto the system an optimised kernel for that particular machine. So, it appears: (1) For speakup to work in the installer, we need to use a kerne-image-*-speakup image. (Other than the speakup patches, how does this differ from the normal i386 images?) (2) If the kernel is not speakup-patched, how do we detect this in kbd-chooser (the system map will not be shipped in d-i)? Does it make more sense just to not offer speakup in kbd-chooser, or ship a speakup-variant of d-i with speakup kernel and keymap? Regards, Alastair McKinstry -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
New alioth repository for console-data, console-common
Hi, I have created new Alioth SVN repositories for the console-data and console-common packages. These contain debconf templates that are used as part of the 2nd stage installation. Translators are welcome to commit additions and changes to the .po files included . Anyone wanting access, please email me your alioth username; Debian developers should already have access, I believe. They are available from svn.debian.org via svn , using svn import svn+ssh://[EMAIL PROTECTED]/svn/console-data ditto for console-common, etc Translations currently in BTS have already been added. -- SlÃn agus Beannacht, Alastair Alastair McKinstry <[EMAIL PROTECTED]> GPG Key fingerprint (New) = CD45 260A 4546 C3C0 F595 F0F6 4132 BF90 2A38 5C57 He that would make his own liberty secure must guard even his enemy from oppression; for if he violates this duty he establishes a precedent that will reach to himself. - --Thomas Paine
Bug#241231: DECstation 5000/133 (mipsel) install with some problems
>On Wed, Mar 31, 2004 at 02:21:49PM +0100, Martin Michlmayr wrote: > >> Package: installation-reports > >> Debian-installer-version: daily r3k-kn02/boot.img snapshot, 200 >> "Select a keyboard layout": doesn't do anything (I didn't have a >> keyboard conncted, but anyway). > >This is the major problem we have on mipsel currently. The LK201 >keymaps have been splitted out into a seperate udeb and kbd-chooser >needs to handle this correctly. I had preliminary code for this but >due to several issues it does not work properly. Another thing in >kbd-chooser is that for some reason the keyboard-handling routines >seem to be passed a subarch name (as in -r4k-kn04) instead of the >keymap name on mipsel. This needs further inspection. > Actually this is right in this case, Karsten. The new serial-detection code is detecting a serial console install, and since DEBCONF_PRIORITY is higher than "low", kbd-chooser is setting serial-console=true and quitting without setting the keyboard. Whats confusing is that this option is appearing on main-menu, but apparently 'does nothing'. Martin is only seeing this because country-chooser has for some reason failed, and he sees main-menu. In the normal DEBCONF_PRIORITY=high flow, you would see the language-chooser question, country-chooser and then disk-partitioning; the kbd-chooser question would not appear. Can I be of any assistance in the kbd-chooser problem? Is the code available somewhere? I had some console-keymaps-dec code but have not commited, as you said you had some, and I don't have access to a mips/el box currently, so yours should be better :-) Regards Alastair -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Bug#241231: DECstation 5000/133 (mipsel) install with some problems
Hi Martin, >Package: installation-reports > >I first chose the language, but it didn't ask me for the country but >instead dropped me into main menu on "Select a keyboard layout". Maybe >this was because of 32 MB memory only, but I didn't get a message about >low-mem installs. > >"Select a keyboard layout": doesn't do anything (I didn't have a >keyboard conncted, but anyway). > Can you be more specific ? Do you mean you were given this option at the main-menu, which then did nothing? or you were given a kbd-chooser screen, and if so what did it show, and what happened? What is supposed to happen in this instance is that kbd-chooser, when run on a serial-console system, will autodetect the serial console and quit, unless DEBCONF_PRIORITY=low, in which case you get to configure a keyboard if it is present. This should mean you don't actually get to see a kbd-chooser screen on a serial-console machine in the "normal flow", when DEBCONF_PRIORITY=high (Its now looking for "console=ttyS*" on /proc/cmdline to see if a serial console is used). - Alastair -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Bug#238364: bug #238264: arcboot installation on mips
Thank you for your Debian Installer installation report. I am attempting to debug the arcboot installation problem. Could you please explain further what happened? Do you have the log output from the installation attempt? If possible, could you try again, and do # sh -x /var/lib/dpkg/info/arcboot.postinst to see what fails. -- SlÃn agus Beannacht, Alastair Alastair McKinstry <[EMAIL PROTECTED]> GPG Key fingerprint (New) = CD45 260A 4546 C3C0 F595 F0F6 4132 BF90 2A38 5C57 He that would make his own liberty secure must guard even his enemy from oppression; for if he violates this duty he establishes a precedent that will reach to himself. - --Thomas Paine
[l10n] Updated iso-codes 3166 and 639 lists
Hi, I have updated the iso-codes 639 language lists to include all new entries added in the last year; and the ISO-3166 country list to use the official ISO-3166 short names. This means there is some changes that need translating; could translators please look at the alioth CVS repository: https://alioth.debian.org/scm/controller.php?group_id=1438 Any translator who wants CVS access please contact me. Regards, Alastair McKinstry -- SlÃn agus Beannacht, Alastair Alastair McKinstry <[EMAIL PROTECTED]> GPG Key fingerprint (New) = CD45 260A 4546 C3C0 F595 F0F6 4132 BF90 2A38 5C57 He that would make his own liberty secure must guard even his enemy from oppression; for if he violates this duty he establishes a precedent that will reach to himself. - --Thomas Paine
Re: BIDI support in Debian Installer : help needed
>Shlomi Loubaton did some work on BIDI (BI-DIrectional langauges >support) for Debian Installer. > >He hacked the slang libraries so that some handling of BIDI is added >to it. > >This will probably need adding the libfribidi0 library to Debian >Installer. Thus a libfribidi0-udeb package will probably need to be >generated by the fribidi source package to be included in some Debian >Installer initrds > >I tried to modifiy the fribidi source package to have it generated >this udeb, but failed to do so, and I have to give this up for at >least two days. > >Can someone with good knowledge of packaging have a look into this and >try to produce a patch for the fribidi package for having it generate >a libfribidi0-udeb package. This udeb should have just ONE file�: >/usr/lib/libfribidi0.so.0.0.0 > >Copy to Baruch Even, the fribidi package maintainer > I have done so ; its present in http://people.debian.org/~mckinstry/fribidi for testing - Alastair Will diff out the patch and forward to Baruch Evan
Bug#232085: Please retest
Hi, I can't reproduce this (don't have the hardware) and can't see why this might happen. Can you please retest? There has been some changes to kbd-chooser recently; could you please test version 0.46; if the problem still occurs I would appreciate it if you built and tested a debug version (more comments printed in the logs); using export DEB_BUILD_OPTIONS=debug debuild -uc -us in kbd-chooser. -- SlÃn agus Beannacht, Alastair Alastair McKinstry <[EMAIL PROTECTED]> GPG Key fingerprint (New) = CD45 260A 4546 C3C0 F595 F0F6 4132 BF90 2A38 5C57 He that would make his own liberty secure must guard even his enemy from oppression; for if he violates this duty he establishes a precedent that will reach to himself. - --Thomas Paine
Bug#239035: mac/ADB keymaps inappropriate for powerpc
Hi Colin,
Thanks for your report.
Looking at your patch, you removed mac_kbd.c from powerpc.
I'd prefer to do the following: check in mac-kbd.c:
fd = open("/proc/sys/dev/mac_hid/keyboard_sends_linux_keycodes", O_RDONLY);
if ((fd < 0) || (read(fd,&c, 1) != 1)) {
di_info ("Failed to open
/proc/sys/dev/mac_hid/keyboard_sends_linux_keycodes;
Assuming not adb");
k->present = FALSE;
return keyboards;
}
close(fd);
if (c != '0') {
di_info ("Sending linux keycodes ; not using ADB keymaps\n");
k->present = FALSE;
return keyboards;
}
This allows kbd-chooser to be used on a running system
as well as in d-i. Can you please verify this would work
as planned on powerpc?
On review, it was a mistake to couple "architecture"
(AT / USB / etc.) to keymap choice in d-i. All the
keymaps in console-keymaps-usb are for the mac keyboards;
"PC" USB keyboards use AT keymaps; etc.
However fixing that now in sarge would be too much breakage, so I am including
your patch; and doing the following:
- Read the USB vendor code from /proc/bus/usb/devices ;
if the keyboard vendor != Apple, don't use the "USB" (Mac USB)
keymaps.
- If in kbd-chooser.c we find no keyboards, and we're not using a serial console,
use AT keymaps.
Any comments ? Do you know of a way to detect ADB keyboards
on 2.4 / 2.6 kernels?
- Alastair
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
[i18n] String Changes in iso-codes
Hi, I have just uploaded a new release of iso-codes, 0.28. This includes the iso-3166-udeb used by d-i, and this has 2 changes that need translating for d-i. They are: "Taiwan, Province of China" for TW and "United Kingdom" for GB. Can translators please translate these files? iso-codes is now present on alioth, and for any translator who wants commit rights, please email me with your alioth (debian) username. -- SlÃn agus Beannacht, Alastair Alastair McKinstry <[EMAIL PROTECTED]> GPG Key fingerprint (New) = CD45 260A 4546 C3C0 F595 F0F6 4132 BF90 2A38 5C57 He that would make his own liberty secure must guard even his enemy from oppression; for if he violates this duty he establishes a precedent that will reach to himself. - --Thomas Paine
Re: preliminary test of d-i on the mips board SWARM from Broadcom
Am Sath, 2004-03-20 ag 12:36, scrÃobh Martin Michlmayr: > * Alastair McKinstry <[EMAIL PROTECTED]> [2004-03-20 07:27]: > > kbd-chooser in d-i _should_ autodetect serial consoles and put this > > option on the menu. Could you please send me a copy of the install log; > > it should list whether serial console was detected or not. > > /dev/console? /dev/ttyS1? > > Jan 10 19:58:10 (none) syslog.info klogd: Mounted devfs on /dev > Jan 10 19:58:10 (none) syslog.info klogd: Freeing unused kernel memory: 72k freed > Jan 10 19:58:10 (none) syslog.warn klogd: Ignoring TIOCGSERIAL > Jan 10 19:58:10 (none) syslog.warn klogd: Ignoring TIOCGSERIAL > Jan 10 19:58:10 (none) daemon.info init: Starting pid 52, console /dev/console: > '/bin/sh' > Jan 10 19:58:10 (none) daemon.info init: Starting pid 54, console /dev/console: > '/sbin/debian-installer' > > Oh, I see what's going on. The UART driver for this board doesn't implement > TIOCGSERIAL. Ok, Checking through the kernel sources, the EINVAL will be returned from the ioctl(), making kbd-chooser (using the same logic as woody, BTW) decide that there is no serial console present. Is a serial console, etc mentioned on the kernel commandline (/proc/cmdline)? Can we improve the autodetection code that way? > -- > Martin Michlmayr [EMAIL PROTECTED] -- SlÃn agus Beannacht, Alastair Alastair McKinstry <[EMAIL PROTECTED]> GPG Key fingerprint (New) = CD45 260A 4546 C3C0 F595 F0F6 4132 BF90 2A38 5C57 He that would make his own liberty secure must guard even his enemy from oppression; for if he violates this duty he establishes a precedent that will reach to himself. - --Thomas Paine
Bug#233856: should be a way to disable usb-discover at boot prompt, for bad hardware that hangs kernel
Ok, I will add a boot-time option to disable usb-discover. However, I would also like to improve usb-discover's handling of this situation, eg stop it hanging on your hardware. Can you tell me: (1) the output of lspci and lspci -n (2) lsmod the list of usb modules when the system is working; ie what USB modules do you need? -- SlÃn agus Beannacht, Alastair Alastair McKinstry <[EMAIL PROTECTED]> GPG Key fingerprint (New) = CD45 260A 4546 C3C0 F595 F0F6 4132 BF90 2A38 5C57 He that would make his own liberty secure must guard even his enemy from oppression; for if he violates this duty he establishes a precedent that will reach to himself. - --Thomas Paine
Re: Bug#238973: iso-codes: Official name for TW is "Taiwan, Province of China"
Am Aoine, 2004-03-19 ag 23:01, scrÃobh Christian Perrier: > Package: iso-codes > Version: 0.27-1 > Severity: normal > > According to > http://www.iso.ch/iso/en/prods-services/iso3166ma/02iso-3166-code-lists/list-en1.html#sz > and the ISO-3166 FAQ, the official name for the TW entry is "Taiwan, > Province of China". I will change this if necessary, but would prefer not to. I think a better solution is to change the name of the list from "Country" to "Region", eg. "Select a region". The current names used in /usr/share/iso-codes/iso_3166.tab are chosen to be a drop-in replacement for the glibc file /usr/share/zoneinfo/iso3166.tab , which uses "Taiwan". I did not want to make changes in case something broke, but that does not seem to be an issue, so I'm open to changing. However, that is why there is a separate file iso-3166-1.tab with the Offical, rather than popular, names of regions. A better solution that I am working on is an XML form of the list, giving eg. ISO-3166 lists both short and official forms of country names, but some "short" forms are still verbose for a drop-down list, eg. "LAO PEOPLE'S DEMOCRATIC REPUBLIC". > Due to the importance, this subtlety has for Chinese people, both in PRC and > ROC, Agreed, Another sensitive name would be "Macedonia". Less controversially, I am changing "GB Britain (UK)" to "GB United Kingdom" in the next release, as per usability report. If you still want Taiwan to be changed, say so, and I will release with that change. -- SlÃn agus Beannacht, Alastair Alastair McKinstry <[EMAIL PROTECTED]> GPG Key fingerprint (New) = CD45 260A 4546 C3C0 F595 F0F6 4132 BF90 2A38 5C57 He that would make his own liberty secure must guard even his enemy from oppression; for if he violates this duty he establishes a precedent that will reach to himself. - --Thomas Paine
Re: preliminary test of d-i on the mips board SWARM from Broadcom
Am CÃad, 2004-03-17 ag 23:22, scrÃobh Thiemo Seufer:
> [snip]
> > In the main menu, I selected "Select a keyboard layout", and chose
> > PS2. But then I was still on this point in the main menu, rather than
> > on the next one ("Detect network hardware").
>
> Which means the PS/2 config failed, no surprise since the machine
> has no such keyboard. Selecting "No keyboard configuration" might
> be more appropriate for serial console. Wasn't this displayed on
> the SWARM?
Hi,
(Sorry about the delay in writing; phone lines down.)
kbd-chooser in d-i _should_ autodetect serial consoles and put this
option on the menu. Could you please send me a copy of the install log;
it should list whether serial console was detected or not.
On the SWARM, what device is used as console (during install)?
/dev/console? /dev/ttyS1?
>
> Thiemo
--
SlÃn agus Beannacht,
Alastair
Alastair McKinstry <[EMAIL PROTECTED]>
GPG Key fingerprint (New) = CD45 260A 4546 C3C0 F595 F0F6 4132 BF90 2A38 5C57
He that would make his own liberty secure must guard even his enemy from
oppression; for if he violates this duty he establishes a precedent that
will reach to himself.
- --Thomas Paine
Re: preliminary test of d-i on the mips board SWARM from Broadcom
Am CÃad, 2004-03-17 ag 23:22, scrÃobh Thiemo Seufer:
> [snip]
> > In the main menu, I selected "Select a keyboard layout", and chose
> > PS2. But then I was still on this point in the main menu, rather than
> > on the next one ("Detect network hardware").
>
> Which means the PS/2 config failed, no surprise since the machine
> has no such keyboard. Selecting "No keyboard configuration" might
> be more appropriate for serial console. Wasn't this displayed on
> the SWARM?
Hi,
(Sorry about the delay in writing; phone lines down.)
kbd-chooser in d-i _should_ autodetect serial consoles and put this
option on the menu. Could you please send me a copy of the install log;
it should list whether serial console was detected or not.
On the SWARM, what device is used as console (during install)?
/dev/console? /dev/ttyS1?
>
> Thiemo
--
SlÃn agus Beannacht,
Alastair
Alastair McKinstry <[EMAIL PROTECTED]>
GPG Key fingerprint (New) = CD45 260A 4546 C3C0 F595 F0F6 4132 BF90 2A38 5C57
He that would make his own liberty secure must guard even his enemy from
oppression; for if he violates this duty he establishes a precedent that
will reach to himself.
- --Thomas Paine
Re: Bug#234513: debian-installer: segfault when starting kbd-chooser
Am Aoine, 2004-03-12 ag 23:46, scrÃobh Colin Watson: > On Tue, Feb 24, 2004 at 11:48:16AM +0100, Thomas Arendsen Hein wrote: > > Package: debian-installer > > Version: N/A; reported 2004-02-24 > > Severity: important > > > > Installation using sarge-powerpc-netinst.iso downloaded from > > http://gluck.debian.org/cdimage/testing/sid_d-i/powerpc/20040223/ > > can't start kbd-chooser. > > > > The segfault is here, on line 627 of kbd-chooser.c in version 0.42: > > return (preferred->present == TRUE) ? "low" : "medium"; > In addition to this, is it a bad sign that no preferred keyboard seems > to be found? Thanks for this. I've been having a hard time getting a powerpc setup to test this on. Mac keyboards were not being included in powerpc machines! Ok, now can we get this CVS version tested quickly for beta3? - Alastair > -- > Colin Watson [EMAIL PROTECTED] -- SlÃn agus Beannacht, Alastair Alastair McKinstry <[EMAIL PROTECTED]> GPG Key fingerprint (New) = CD45 260A 4546 C3C0 F595 F0F6 4132 BF90 2A38 5C57 He that would make his own liberty secure must guard even his enemy from oppression; for if he violates this duty he establishes a precedent that will reach to himself. - --Thomas Paine -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: IA64 install experience (HP zx6000)
Am Aoine, 2004-03-12 ag 14:39, scrÃobh Jeff Bailey: > On Fri, 2004-03-12 at 00:59, Roland Dreier wrote: > > > Finally, at least with some of the earlier images, my USB keyboard did > > not work once the installer started up. Unfortunately the zx6000 does > > not have a PS/2 keyboard connector so I was only able to do anything > > through the serial console. I'm not sure whether this was still the > > case with the March 10 image, since I just stuck to serial for that > > image. > > rgh or dannf will have to comment on the scsi controller. I'm working > on solving the usb keyboard thing (if one of them hasn't already). The > machine that generates the nightly snapshots needs to be reinstalled, > and I'm getting hit on this bug myself. =) > > Tks, > Jeff Bailey On i386, this is the purpose of the usb-discover udeb: to load modules necessary for USB keyboards, in particular, to work. This is why it must be run before keyboard input is required. What USB modules are required under woody for the keyboard to operate on the zx6000? -- SlÃn agus Beannacht, Alastair Alastair McKinstry <[EMAIL PROTECTED]> GPG Key fingerprint (New) = CD45 260A 4546 C3C0 F595 F0F6 4132 BF90 2A38 5C57 He that would make his own liberty secure must guard even his enemy from oppression; for if he violates this duty he establishes a precedent that will reach to himself. - --Thomas Paine -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: base-config, kbd and console-tools
Am Aoine, 2004-02-27 ag 06:52, scrÃobh Denis Barbier: > Hi, > > is there any reason why base-config depends on console-tools instead > of "console-tools | console-utilities"? Should work, but I'm not sure of the advantage: base-config almost always just gets run at installation, so console-tools will be chosen. > It is not clear for me whether console-tools or kbd should be used; > the former seems to be unmaintained upstream but have an active > Debian maintainer, whereas this seems to be the opposite for kbd. > In a recent (01/2004) thread on the linux-utf8 ML (currently offline, so > no pointer, but Google should find its cached version) about problems > with UTF-8 and console, kbd upstream advised to throw console-tools away > and install the newest kbd. > Does someone have some experience on this topic? > Ok, A summary of status and plans: The reasons console-(tools|data) still exist are: (1) A number of developers have threatened/promised to restart work "post sarge". (*) (2) Until recently at least, kbd was unmaintained within debian, and had less integration: console-tools, etc using debconf. Recently, Wartan has packaged more recent versions o kbd; I'd need to investigate its debianisation to check if (3) is still true. I have been reluctant to get involved in the kbd package as I feel a degree of redundancy on such an essential package is useful. Andries Brouwer, the kbd developer, is cordial to re-merging the packages; so I've been working with the plan of making the diffs between console-(tools|data) easily mergable back to kbd. console-(tools|data) now use dbs, so patches are maintained individually, to make this easy. I intend to do another "major release" of console-(tools|data) soom which just merges the common patches (common to both console* and kbd) back to the "mainline", keeping just debian-specific stuff as seperate patches, feeding patches back to Andries. I raid each kbd release for new changes and merge them into console-* ; I have yet to do so for kbd-1.12, which has some 2.6 kernel fixes, I believe. * Planned (post-sarge) upcoming changes to console-tools include using X keyboard maps, so we have a more consistent, comprehensive keyboard map set. We should be able to use the libxbfile library to do this. ( Especially with a change like this, I feel it important to keep kbd seperate as a backup while such a merge takes place, so if we f*ck things up, people have something else to use) > Denis -- SlÃn agus Beannacht, Alastair Alastair McKinstry <[EMAIL PROTECTED]> GPG Key fingerprint (New) = CD45 260A 4546 C3C0 F595 F0F6 4132 BF90 2A38 5C57 He that would make his own liberty secure must guard even his enemy from oppression; for if he violates this duty he establishes a precedent that will reach to himself. - --Thomas Paine -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
[l10n] Translation needed for console-data
Hi, In the next upload of "console-data" I am including a French Atari keymap. This means one more entry that requires translation. Can translators please provide a translation for "French (Atari)" for console-data/debian-installer? Thanks, Alastair McKinstry -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
[i18n]: For translators: iso-3166-udeb has changed
Hi, This is an advanced warning that "iso-3166-udeb", used in Debian Installer, (and the iso-codes source package), need updating. I have created an Alioth project for it: http://www.pkg-isocodes.alioth.debian.org Debian developers should have direct CVS commit access (module iso-codes); for other translators, what has changed: iso_3166.pot has a new entry msgid "Ãland Island" msgstr "" that needs translating. You could just send this on to me to commit if you wish. There are also updates and spelling corrections to the iso_3166_2.pot files for the more adventurous. -- SlÃn agus Beannacht, Alastair Alastair McKinstry <[EMAIL PROTECTED]> GPG Key fingerprint = CD45 260A 4546 C3C0 F595 F0F6 4132 BF90 2A38 5C57 He that would make his own liberty secure must guard even his enemy from oppression; for if he violates this duty he establishes a precedent that will reach to himself. - --Thomas Paine -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Bug#232085: Bug #232085: please retest
Hi, I'm afraid I cannot reproduce this bug. With kbd-chooser, you should be able to choose the keymap manually with DEBCONF_PRIORITY <= medium. I am able to do so on i386, at least. Can you check if this is a mac-specific bug? Have a look in the logs (console #4, or if possible copy /var/log/messages and send it to me), and see if it is checking subarches correctly. I don't have a mac to check, and the only reason offhand I can see this breaking is if the subarch detection code is not working. Regards, Alastair -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Up/Down/PgDn/PgUp not working anymore in cdebconf?
>Quoting Alastair McKinstry ([EMAIL PROTECTED]): > >> (I'm working on broken i18n in newt/whiptail at the moment). > >Support for BIDI languages�? > >I've got no news from BIDI languages people, either Arabic or >Hebrew I'm starting to worry whether we will succeed in getting a >fully functional d-i for these languages... > Thats the plan. There are a couple of fixes for UTF-8 that also need to go in, then release. I'd like to release an experimental release this weekend, but have learned the hard way not to make promises given my current work. I'm recovering from a Death march project at work; last weekend was the first full weekend off for our group in months. The plan is to fix kbd-chooser issues tonight/tomorrow, work on an experimental newt package by Sunday, and to expect interruptions and demands by my family :-). Regards, Alastair >-- >To UNSUBSCRIBE, email to [EMAIL PROTECTED] >with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED] > > -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Up/Down/PgDn/PgUp not working anymore in cdebconf?
>On Mon, Feb 16, 2004 at 07:18:07AM +0100, Christian Perrier wrote: >> With a 20040212 netinst ISO image, the direction keys do not work >> anymore for moving around d-i menus. >> >I saw this one too. I also got left with a half finished configure once >I'd "finished" following the first reboot. It took three attempts >before I managed to get it to work. I'll be trying again tonight :) > >Andy A regression in newt broke this. Sorry. I did an upload yesterday to fix it: todays builds should work. (I'm working on broken i18n in newt/whiptail at the moment). Regards, Alastair > >-- >To UNSUBSCRIBE, email to [EMAIL PROTECTED] >with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED] > > -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Bug#232397: use menutest in bterm-unifont instead.
The postinst in bterm-unifont never worked, by misunderstanding of design. I intend to implement the menutest hack instead; this will restart bterm-unifont (my bterm design, it cannot simply be HUP'd , it needs to be restarted). The hard part of this is testing, which is why it was left so long. -- SlÃn agus Beannacht, Alastair Alastair McKinstry <[EMAIL PROTECTED]> GPG Key fingerprint = 9E64 E714 8E08 81F9 F3DC 1020 FA8E 3790 9051 38F4 He that would make his own liberty secure must guard even his enemy from oppression; for if he violates this duty he establishes a precedent that will reach to himself. - --Thomas Paine -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: D-I and UTF-8
Am Luan, 2004-02-02 ag 17:39, scrÃobh Nikolai Prokoschenko: > Hello! > > I must have missed the discussion, as this question should have been > raised several times: If the first stage of the installer uses UTF-8 > anyway, wouldn't it be wiser to configure the second step also into > UTF-8 and thus make UTF-8 the default encoding? > > BTW, I know the console is currently almost uncapable of UTF-8, but > anyway -- it won't ever be, as I see it... Redhat has done the step, > what should be done, so that Debian can also? > The problem is that it in some circumstances _reduces_ functionality at the moment. With a UTF-8 console you lose the ability to do compose characters (eg. accents). ie a de_DE.ISO_8859-15 locale allows you to enter accents, but a de_DE.UTF-8 one doesn't: the problem is that the kernel ABI (even in 2.6) defines (char base, char diactritic) -> char result_char (See struct kbdiacr in /usr/include/linux/kd.h) This works for simple 8-bit locales (eg ISO-8859, ru_KOI8), etc but breaks UTF-8, which requires strings. Using UTF-8 makes X11 easier but breaks console functionality. Lets go to UTF-8 by default when either (1) the kernel is fixed (2) We jump straight to X. - Alastair McKinstry > -- > Nikolai Prokoschenko > [EMAIL PROTECTED] / Jabber: [EMAIL PROTECTED] -- SlÃn agus Beannacht, Alastair Alastair McKinstry <[EMAIL PROTECTED]> GPG Key fingerprint = 9E64 E714 8E08 81F9 F3DC 1020 FA8E 3790 9051 38F4 He that would make his own liberty secure must guard even his enemy from oppression; for if he violates this duty he establishes a precedent that will reach to himself. - --Thomas Paine -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Bug#229761: #229761: Bad keymaps in installer
Hi, Thanks for testing the keyboards in debian-installer. In what way do you believe these keymaps are bad? Lithuanian Latvian Japanese (106 key) Macedonian Dutch Is this in the install stage, or later? If so, which kernel, etc. did you use? Thanks, Alastair McKinstry -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Countrychooser commited in CVS
Am Sath, 2004-01-17 ag 08:03, scrÃobh Christian Perrier:
> (Alastair, this is a discussion in -boot about countrychooser, a new
> d-i package, which makes use of your iso-codes package. I need to talk
> with you about some issues)
>
> Quoting Joey Hess ([EMAIL PROTECTED]):
>
> > That's interesting. This is a place where cdebconf deviates from the
> > debconf spec, apparently; ", " is the official list delimiter.
>
> I added the separating space, but I confirm that all works fine
> without it with cdebconf
>
> > Well I was wondering about geographical drill-down, or just geographical
> > groupings/headings in the list:
> >
> > Europe
> > Germany
> > France
> > Spain
> > ...
> > America
> > US
> > Canada
> > Mexico
> > ...
> >
> > But like I said, I'll reserve comment until I've seen it.
>
> Well, the problem here is that we do not have a valuable source for
> doing this. ISO code lists do no give information about such zones.
Yes. I want to avoid expanding the role of iso-codes at the moment, as
it is designed to be a small component for many packages: its hard to
convince a developer to "bloat" their package by making it depend on
yours if you keep expanding your package to include new things.
So I think the information about zones should stay in countrychooser for
the moment.
> I think users in general are quite well aware of such long lists :
> these country lists are what you find on each and every form which
> asks you for your address (those which continue to list French Guyane
> near France).
>
> As long as the form is sorted, any user may easily find his country by
> typing the first letter of its official name which I hopà they're
> aware of.
>
> The *current* list is a bt broken, according to Alastair (who
> maintains the iso-codes packages it's taken from). For instance,
> "Great-Britain" is listed there as "Britain" which is obviously
> bugguy.
Yes. It should match the "Great Britain" returned from the locale, eg.
that returned by nl_langinfo(), so you could do
localised_ctryname = dgettext("iso_3166",
nl_langinfo(_NL_IDENTIFICATION_TERRITORY));
I would also like iso_3166.tab to match /usr/share/zoneinfo/iso3166.tab
(and so replace it later), but at the moment iso3166.tab uses "Britain
(UK)", so at the moment these two goals conflict. Working on it.
> I'm waiting for Alastair to release a new version of iso-codes with
> updated codes and translations.
>
> I will also ask him for a iso-3166-udeb package as countrychooser
> needs the iso-3166.tab file as reference.
Ok, that I can do. Should it just contain the .tab file, or the .po
files too?
- Alastair
--
SlÃn agus Beannacht,
Alastair
Alastair McKinstry <[EMAIL PROTECTED]>
GPG Key fingerprint = 9E64 E714 8E08 81F9 F3DC 1020 FA8E 3790 9051 38F4
He that would make his own liberty secure must guard even his enemy from
oppression; for if he violates this duty he establishes a precedent that
will reach to himself.
- --Thomas Paine
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Bug#225897: kbd-chooser: German translation uses wrong encoding
Hi, The current version of kbd-chooser appears to be correct Unicode. Could you please check and confirm that this bug can be closed? SlÃn agus Beannacht, Alastair -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Bug#219290: [d-i] No USB keyboard after reboot
On Sun, 2003-12-28 at 11:48, Martin Michlmayr wrote: > * David Nusinow <[EMAIL PROTECTED]> [2003-12-26 18:21]: > > > Apart from this solution, which is certainly a good first step, I > > > wonder which discover doesn't handle this situation properly. It > > > seems to me that most of usb-discover should be merged into discover. > > > > How much of usb-discover are you thinking? I'll have a look at it and > > see if I can't produce a preliminary package. > > I realized in the meantime that there's probably a good reason for > usb-discover to exist. I guess this package exists so USB Keyboards > can be identified early on when discover is not available yet. As the original author, I looked over the package, and yes ,it doesn't make it clear why it exists. It is indeed there to be run as soon as possible so that USB keyboards can be used; it must be as small as possible to fit in initrds (for floppies and the 2MB limit on isofs) , hence discover is not an option at this point. However, it would be better if the list of USB modules was taken from the maintained discover[2] package automatically rather than built by hand. Regards, Alastair McKinstry > -- > Martin Michlmayr > [EMAIL PROTECTED] > -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Should we really use countries in languagechooser
Happy Holidays! On Thu, 2003-12-25 at 09:09, Christian Perrier wrote: > Quoting Anmar Oueja ([EMAIL PROTECTED]): > > > Please change it as you see fit. As a matter of fact. in the > > languagechoose.l10n I put in "choose Arabic to begin installer (Arab > > World" in arabic of course to direct it to all Arabic speakers > > regardless of their country of origin > > I'll try to make some general proposals about this. My whole feeling > is that countries are unneeded in d-i. > > But I need to speak of this with Alastair.and list. The country information is used in two places: (1) In selecting dialect; eg. US American vs British; Swiss German vs German (in principle; not done yet, I believe) (2) To select _default_ values for other things: - keyboard - ftp mirror - timezone others? Perhaps one possibility is to ask the country question at a lower priority _after_ the language question, if necessary; e.g. we might not need to ask the 'which country' bit for many languages, and it would simplify the language question. This approach was taken in woody, I believe. The default value of the country question would be 'XX' for undefined, unless the question is answered by the user (or a GPS device, etc). SlÃn agus Beannacht, - Alastair > > > -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: BIDI Status in Debina Installer
On Sun, 2003-12-21 at 20:38, Petter Reinholdtsen wrote: > [Anmar Oueja] > > Hello All: > > > > I would like to know the status of adding BIDI support to the Debian > > Text-based interface installer. > > I suspect the important bug is http://bugs.debian.org/212962 >, > and no one have added any info to it since I reported it. I suggest > starting with adding the language choice and the cdebconf > translations, and then get them to display correctly. > > > This is important since I am working on translating the PO files for > > the installer to Arabic. > > This sounds very good. I look forward to your patches. :) > I apologise for not responding to this earlier; I have been working on patches for BIDI for d-i (a fribidi udeb, patches for cdebconf and newt), but all plans have been thrown into disarray by illness; my family (me included !) we struck by flu and I'm struggling to get better by Christmas. I hope to post the patches ASAP. Regards, Alastair -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: cdebconf menu location
>Quoting Joey Hess ([EMAIL PROTECTED]): >> Having this ugly debconf priority question right after choosing the >> language is annoying. Do we really need it there? > >I haven't tested a full d-i for a while now, but I think I can get the >point. The point of putting it there was that many users weren't aware they _could_ change the priority, and ask more questions (eg the kbd-chooser default issue). It pretty much needs to be asked to be noticed; people just fail to read the boot splash screen. I agree that its ugly, though, and probably deserves to be re-written. I'd leave that until after the string freeze, though, and for people with better English than me (eg Christian Perrier :-)) Slán agus Beannacht, - Alastair McKinstry -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Fixed in NMU of lvmcfg 0.10
tag 219471 + fixed quit This message was generated automatically in response to a non-maintainer upload. The .changes file follows. -BEGIN PGP SIGNED MESSAGE- Hash: SHA1 Format: 1.7 Date: Wed, 10 Dec 2003 21:59:11 + Source: lvmcfg Binary: lvmcfg Architecture: source all Version: 0.10 Distribution: unstable Urgency: low Maintainer: Debian Install System Team <[EMAIL PROTECTED]> Changed-By: Alastair McKinstry <[EMAIL PROTECTED]> Description: lvmcfg - Configure the Logical Volume Manager (udeb) Closes: 219471 Changes: lvmcfg (0.10) unstable; urgency=low . * Verok Istvan - Initial Hungarian translation * Template polishing from Christian Perrier plus fixes by Chris Tillman. Closes: #219471 Files: 91ec25edcd2670856bf70abd13c2f4a9 587 debian-installer optional lvmcfg_0.10.dsc 37e0b0693ba1574ba8b48d4096c22b40 38943 debian-installer optional lvmcfg_0.10.tar.gz e51222ed0682eed53c03872af79fe5bd 5626 debian-installer optional lvmcfg_0.10_all.udeb -BEGIN PGP SIGNATURE- Version: GnuPG v1.2.3 (GNU/Linux) iD8DBQE/2hQZ2tp5zXiKP0wRAnEWAJwMfgiGYnE/TTg9mSG1KOumQCOoSACeKKGG wGEkHEccWKFmMh1twqhcDbU= =kqHo -END PGP SIGNATURE- -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Language choser is also keyboard choser ??
>In the last release of debian installer the language choser is also the >keyborad layout choser. I don't know how the association language - >keybord is done but i don't like this way to configure things. > >I'm italian and is very common that i use english keyborads but i like >italian language interface, in north italy is common also to use deutch >keybords with italian languge or the opposite. > >Why not present two menu, one for language and one for keyboard ? > >I dont' like also the way to select language, is not clear. I like more >a solution like: > > Please chose your languge: > > DE - (I don't know german) > EN - Select that entry to procede in English > ES - (I don't know espagnol also) > IT - Seleziona questa voce del menu per procedere in italiano > >Whe can order coutry codes alfabetically so, i think, found the right >one is more easy. > >Then we can present the chose of the keyboard layout with the default on >the default association. There is a keyboard chooser question; it is not shown by default in the beta-1 release, it is now shown by default in the daily (and future) releases. To see the keyboard question, start with DEBCONF_PRIORITY=medium on the initial prompt . The default choice of keyboard is taken from the languagechooser. Regards, Alastair McKinstry > >If you agree with that proposal i can try to submit a patch. > >Ciao >-- >Daniele > > >-- >To UNSUBSCRIBE, email to [EMAIL PROTECTED] >with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED] > > -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: updated lithuanian .po files
On Thu, 2003-12-11 at 21:32, KÄstutis BiliÅnas wrote: > Hi, > > In attachement there are some updated lithuanian (lt_LT) > .po files for d-i. > > Thanks, Committed. Thank you for your contribution to Debian. SlÃn agus Beannacht - Alastair McKinstry -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: why is dvorak the default keyboard for English locales?
On Thu, 2003-12-11 at 20:06, Joey Hess wrote: > I see that even with the latest kbd-chooser, if I pick American English > in languagechooser, it defaults to a dvorak keyboard. At least it asks > me about it now at high priority, but surely the American English > keyboard would be a much better default? This is a bug in console-data, that I have fixed. I cannot upload the fixed version ATM because there is a corrupted version sitting in /pub/UploadQueue on ftp-master.debian.org, and I can't login to clear it. I'm waiting for it to time out. Sorry about that. - Alastair -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Hungarian l10n of d-i
On Wed, 2003-12-10 at 17:58, VEROK Istvan wrote: > Hi! > > Will someone please commit this to CVS and set me straight as to where > to obtain those additional PO files external to d-i. I have no > familiarity with d-i code at all, and am no DD, so verbose replies are > welcome. Committed to CVS. Thank you for your contribution to Debian. - Alastair McKinstry -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: choose-mirror country codes
On Thu, 2003-12-04 at 20:03, Denis Barbier wrote: > On Wed, Dec 03, 2003 at 10:39:40PM +0000, Alastair McKinstry wrote: > > On Wed, 2003-12-03 at 21:52, Joey Hess wrote: > [...] > > > All well and good, but when I run choose-mirror in en_US, I see "US". > > > What needs to be done for me to see any of the above? > > > > The language should be set to en_US rather than C in the absense of > > languagechooser. Then English translations will be picked. > > > > I'll investigate as soon as I can in detail; I think it just means > > changing the C.UTF-8 to en.UTF-8 > > AFAICT there is no debian/po/en.po file containing the country > code <-> English name mapping. It should be generated by > makeheader.pl. > > Denis Doh. I can't verify but it appears that I haven't checked in my local copy of debian/en.po. Needless to say it works fine locally, and this is why I didn't understand the problem Joey was talking about ... I'll commit as soon as CVS is back. - Alastair -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: choose-mirror country codes
On Wed, 2003-12-03 at 21:52, Joey Hess wrote: > Alastair McKinstry wrote: > > The countrycodes are used to provide a more "neutral" and > > defined identity of country than English name. > > > > It is easy to confuse "English name" for country with > > "translated name" leading to bugs you won't see those bugs if you are using > > English. > > > > Also, The "English name" for countries is not unique. > > You need to avoid one package using "USA", another using > > "America", another using "United States of America". > > All well and good, but when I run choose-mirror in en_US, I see "US". > What needs to be done for me to see any of the above? The language should be set to en_US rather than C in the absense of languagechooser. Then English translations will be picked. I'll investigate as soon as I can in detail; I think it just means changing the C.UTF-8 to en.UTF-8 Regards, Alastair -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: base-config 2.0
> > >I hope to finish this up so it uses the cdebconf db for the rest of the >stuff in installer-integration, and take it further so it uss any >additonal d-i stuff it can to avoid more questions. I still need to work >out how to get the equivlant of dbootstrap_setting's KEYBD and >SERIALCONSOLE from d-i db. > Whats the issue? Have you seen kbd-chooser/debian/prebaseconfig? You can retrieve the keyboard setting from cdebconf; I'm not sure what SERIALCONSOLE is in dbootstrap; debian-installer/serial-console is a boolean set by kbd-chooser saying whether or not a serial console is in use. - Alastair -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Default bootloader
>[Geert Stappers] >> Add missing serial line (detection) support to the list. > >Are you talking about the missing serial line support in >grub-installer, or some feature of the grub boot loader? The first >should be easily fixable. At least we use grub with serial console >here at the university, and it works just fine. I am not aware that >it is possible to get grub (nor lilo) to autodetect serial lines. It >this possible with lilo? In terms of autodetecting serial lines, are you aware that kbd-chooser sets debian-installer/serial-console to yes/no to indicate whether a serial console is in use? Is this sufficient ? - Alastair -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Localizing New Debian Installer to Arabic
On Thu, 2003-11-20 at 23:06, Anmar Oueja wrote: > Great.. I will get some strings done for ya tonight :) > > > BTW, there are lots of po files for a whole bunch of archtectures. Do > you know which ones I should start with especially for you to test with ? > > How do you want me to submit the localised files ? The most useful ones to start with are main-menu, cdebconf and the languagechooser.l10n. Do you have Cvs access in Debian? if not, the easiest way to submit the files is just to send them to me. Beannacht, Alastair McKinstry > Anmar > > Alastair McKinstry wrote: > > On Thu, 2003-11-20 at 19:27, Anmar Oueja wrote: > > > >>Thanks Vincent: > >> > >>I got some answers that lead me to conclude it is impossible to do > >>Arabic on the NEW debian installer. The main problem is the darn bidi > >>(bidirectional) which isn't supported by NEWT. I tested it uing > >>NEWT-python binding and it did not work > >> > >>Are there are work arounds or something ? > >> > > > > > > I am currently adding bidi support to newt (well, the newt > > frontend for debian-installer in particular). > > > > Could you send some test translations to work on? > > > > Regards, > > Alastair McKinstry > > > > > >>Anmar > >> > >>Vincent Renardias wrote: > >> > >>>On Wed, 2003-11-19 at 21:57, Anmar Oueja wrote: > >>> > >>> > >>>>Hello: > >>>> > >>>>I would like to start localizing the Debian Installer to Arabic. Could > >>>>you please direct me to the right place to get the .po files to start > >>>>localizing ? > >>>> > >>>>How can I test my localization files ? > >>>> > >>>>Does the Debian Installer support Arabic language (right to left) ? > >>> > >>> > >>>Hello, > >>> > >>>did you get any answer? > >>>if not, tell me and I'll tell you how to proceed. > >>> > >>>Cordialement, > >>> > >> > > > -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: cvs commit to debian-installer/utils/debian by mckinstry
>Am Son, den 16.11.2003 schrieb [EMAIL PROTECTED] um 12:38: >> Update of /cvs/debian-boot/debian-installer/utils/debian >> In directory gluck:/tmp/cvs-serv16371/debian >> >> Modified Files: >> changelog control rules >> Log Message: >> Add di-utils-devicefiles package to provide the devicefiles needed >> for those archs that create them 'manually', rather than using >> devfs (or even udev) >> >How is this different from userdevfs? > >gaudenz > It is userdevfs. Its the first step to a (small) cleanup of the build Makefile; first move the code into a little udeb; once it hits the archive, the code can be removed from the Makefile and the udeb added to the pkg lists. Later we can make a build dependent on e.g. devicefiles, provided: by this, a devfs-'d kernel package or a udev'd kernel package. Regards, Alastair -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Re: Status of IPv6 in d-i?
> Actually, that doesn't have to be such a hassle. Provided a Router > Advertisement Daemon (radvd under Linux) is running for your local > network, the kernel will autoconfigure itself, without userland help; > also, globally-addressable IPv6 addresses are easy to distinguish: if > the IP address starts with 0x3ffe or 0x200, it's globally addressable > (although the first is deprecated, and may disappear soon). You're right; good news, we don't even need ping decide if ipv6 is present; just load the module. However we might still like to set up DNS (statically, or via DHCPv6, or SLP). For the limited use it will get in d-i, however, (just looking up the IP for the mirrors), we would probably be safe enough hard-coding an IP(v6) address, (ie ask the user on low priority, with a default IPv6) > It's probably better to distinguish using a negative list tha>n to > distinguish using a positive list, though. -ipv6: Isn't it better to say > "if the IP starts with 0xf, or is ::1 it's *not* globally addressable"? I think so. Beannacht, Alastair McKinstry <[EMAIL PROTECTED]> -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Localizing New Debian Installer to Arabic
Hi, Currently, support for Right-to-Left languages is unimplemented in Debian Installer (bug 212962). Nobody has been working on it as there are no Arabic or Hebrew translations. Please feel free to start, and I will work on this bug :-) You can get the debian installer .po files by checking out with CVS: see http://cvs.debian.org/ , checkout with root debian-boot. You can see the status of translations at http://people.debian.org/~tsauter/d-i/translations.html Regards, Alastair McKinstry -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Status of IPv6 in d-i?
>Hi, > >A friend of mine, who is quite the IPv6 adopter commented to me today that a >(much) earlier version of d-i that he tried out some months ago had IPv6 >support, but that it had seemed to have gone away in the newer releases. > >Are there any plans for the installer to work with IPv6 for the sarge >release? > Yes, if we can get the work done by sarge release time. At the moment, everyone looks focussed on clearing the bug backlog, but the work involved for IPv6 need not be so big: Offhand from IRC discussions I believe we need: (1) ensure wget in busybox (-cvs?) is built with IPv6 support. (2) choose-mirror works sensibly to select IPv6 mirrors (3) (bigger) netcfg tries IPv6, tries to autoconfig (ping local net?) if this works, sets a debconf flag so that choose-mirror, etc. select from IPv6 hosts. All help needed. Beannacht, Alastair McKisntry >Andrew > > >-- >To UNSUBSCRIBE, email to [EMAIL PROTECTED] >with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED] > > -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
boot-floppies ; new release for 3.0r2 ?
Hi, There is a fix in the current CVS boot-floppies, which I believe is required for the Arm RiscStation; it sets the keyboard in kbdconfig.c. Can someone confirm that this is needed for RiscStation? Is there going to be a new release of boot-floppies for the upcoming woody release? Regards, Alastair McKinstry -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Asking the priority question; was Re: Bug#221186: keymap problem fixed
On Mon, 2003-11-17 at 21:24, Denis Barbier wrote: > On Mon, Nov 17, 2003 at 10:30:16AM +0000, Alastair McKinstry wrote: > > Hi, > > > > The dvorak keymap bug is now fixed in daily builds. > > > > It is possible to work around it in two ways: > > (1) Set DEBCONF_PRIORITY=medium on the boot line, to get the > > keyboard question asked. > > > > (2) use bootkbd=us > > Not the change in syntax. > > > > I am downgrading this bug to normal, but leaving it open until > > the documentation is checked to note the change in syntax. > > Err, having to enter boot line parameters in order to have a > decent keyboard does not seem to be an option to me. (Even if > it works in this particular case since the keyboard layout is > qwerty) The bug is fixed; the workaround is for the _beta build_ only. > The keyboard question must be asked, or at least user must be > told that his locale setting assumes some default values > (which are printed at that point), and he has to lower debconf > priority in order to change it. But I really do not understand > why you insist on not displaying this question, which makes > installation much more painful, as already told by bugreports. My point is that the real solution is to ask the priority question immediately after the language question. This way users know that there are questions that they are not being asked, and can go back (via goback, at any question stage) to main-menu and change priority, and ask those questions. I would recommend asking the question, rather than just mentioning debconf priority on the opening text, because users have a habit of not noticing such things unless they are asked to make a choice. The default priority would still be "high". The alternative to asking the question is fielding bug reports every time we make a default choice: e.g. setting up a default mirror in choose-mirror, using the CDROM debs if they are present, using DHCP (when present) rather than asking for an IP, etc. as users are not aware that they have alternatives. It is a balancing act; beginners do get confused by the keyboard question, not knowing which one they should choose. People do evaluate installers based on how many questions they get asked. I recommend we insert the question now; if we get bug reports in beta2 reporting it as an issue, I will change the kbd-chooser priority. I'm sending this mail to debian-boot, and particularly to Joey Hess, as I do not want to change the menu question order without agreement. Does anyone object to such a change? > Denis > Regards, Alastair -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
[Fwd: Alpha Debian Netinst]
A report that was forwarded to me; I am updating the ports-status page with details on submitting reports the offical way. -- Alastair McKinstry <[EMAIL PROTECTED]> GPG Key fingerprint = 9E64 E714 8E08 81F9 F3DC 1020 FA8E 3790 9051 38F4 He that would make his own liberty secure must guard even his enemy from oppression; for if he violates this duty he establishes a precedent that will reach to himself. - --Thomas Paine --- Begin Message --- http://www.debian.org/devel/debian-installer/ports-status It says on that page to report any installations to you. Well, I'm trying the "sarge-alpha-netinst.iso" from testing (both daily and non-daily) on an alpha PWS 500au with 128mb ram using aboot/SRM and so far am unsuccessful. It dies while trying to mount /dev/rd/0. If I tell it to use /dev/ram0 as a root, it says it can't find init. If I pass init=/linurc to the kernel, it still says no init found. So I'm slowly hopefully getting there. I don't have a full dump of the boot process, otherwise I'd dump it to you. Any Ideas? Otherwise I guess it doesn't work... *shrug* --- End Message ---
Bug#221186: keymap problem fixed
Hi, The dvorak keymap bug is now fixed in daily builds. It is possible to work around it in two ways: (1) Set DEBCONF_PRIORITY=medium on the boot line, to get the keyboard question asked. (2) use bootkbd=us Not the change in syntax. I am downgrading this bug to normal, but leaving it open until the documentation is checked to note the change in syntax. Regards, Alastair McKinstry -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Second stage console problem
On Fri, 2003-11-14 at 06:21, Miroslav Kure wrote: > On Thu, Nov 13, 2003 at 05:05:43PM -0500, Joey Hess wrote: > > Kenshi Muto wrote: > > > How do you think about following flowchart, Joey? > > > > > > n > > > check tty -> serial. 2nd stage as current LANG (*1) > > > | y > > > v y > > > check fb0(*2) -> run bogl-term as LANG(*3) > > > | n > > > v y > > > acceptable LANG without bogl-bterm?(*4) -> 2nd stage as LANG > > > | n > > > v > > > cancel LANG (use English) > > > | > > > v > > > 2nd stage > > > > > > *1: If you want to install in serial console/non-FB machine, you can > > > use only English installer, it means LANG of 2nd stage is always > > > English (C). > > > *2: Current termwrap looks don't implement yet. > > > *3: Current termwrap looks don't implement yet. > > > *4: Current termwrap looks implement partly? I think this needs > > > various people's help. > > > > Yep, that sounds right. > > Yet there is the issue with latin-* languages, who don't need > special terminal, but only need to set correct font in the > console-tools package. > E.g. accented characters in iso-8859-2 (Czech, Slovak, maybe > others) won't display correctly until you set lat2 console font, > which makes base-config almost unreadable. > > Regards, > Miroslav Kure Ok, Two things to take note of: (1) I'm implementing a default font for linux console in console-data that will provide Latin, Hebrew, Cyrillic. I can detect and change that based on default locale if necessary - setting eg an iso02 based one. However, there is a catch: if you set a vesa screen mode, e.g. to set a font or console screen size, then setting a font will undo that. So I have to detect a vesa mode and _not_ set font in that case, but leave it to the user. I'm detecting VESA modes by looking for "vga=" on /proc/cmdline; if a framebuffer is set by default here, the font will not be set. (This need not be a problem for base-config, as it can use bogl and load a framebuffer after boot. But help will be needed by various arch people to debug this). (2) Important: There is a bug in the linux kernel API that makes combining characters (accents) at the keyboard effectively impossible in UTF-8 mode: (the API only allows a _char_ for the result of a diacritic -type key: hence compose chars don't work properly in UTF-8 modes). I don't think bogl does anything smart about this; it means that _at_the_console_ entering accents in UTF-8 locales is broken. While this does not affect d-i at the moment (no dialogs that want arbitrary text), it means most users will not want UTF-8 locales on the console. Regards, Alastair McKinstry > -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Bug#220117: Installation report
On Mon, 2003-11-10 at 23:46, Wilmer van der Gaast wrote: > Package: installation-reports > Priority: normal > > INSTALL REPORT > > Second time I tried, everything went better. I picked the Dutch language > from the list, and noticed that it also assumes a Dutch keyboard then. I > can tell you, Dutch keyboards still exist over here, but there are less > Dutch keyboards than Debian users in this country. Please change that > default, or better, let the user pick a layout. Is a lower debconf priority is chosen, a question offering a choice of keyboards will be offered (A bug is filed to make this clearer). The default keyboard is the same as it was for woody; whar is the best default keyboard for Dutch language users ? Note, a different choice may be selected for Dutch Belgian speakers. Regards, Alastair McKinstry > Wilmer van der Gaast. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Bug#220182: netcfg: Please re-order the TCP/IP configuration
Hi, If we do reorder the questions to -IP address -Netmask [guessed from ip address] -Gateway [guessed from previous answers] -DNS servers -Machine name -Domain name Then it may also be possible to retrieve Machine and domain name from DNS, if present. Regards, Alastair McKinstry -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: Bug#215469: don't include unnecessary udebs on the cd
>
>
>
>
>I attach some annotated lists:
>
> - netinst-udebs: udebs that should go on all netinst cds
> - netinst-udebs-i386: additional udebs that should go on i386 netinst cds
> - exclude-udebs: udebs that have no reason to go on any cd
> - exclude-udebs-i386: udebs that have no reason to go on any i386 cd
>
>File format uses # for comments. ${kernel:Version} should be substituted
>with the current installer kernel version, currently 2.4.22 on i386. To
>get the udebs that can go on non-netinst cds, just remove the excluded
>ones from the full list of available udebs.
>
>Someone else will have to take care of the other architectures, and
>these lists could use some review by the other d-i developers. In
>particular:
>
> - I don't understand why console-keymaps for other architectures
>are available on i386.
It is apparently possible to connect other keyboard types to a PC, typically
via serial. The only such adapter I have is
ps2-to-serial, so I can't test this very well yet.
Nevertheless, its so esoteric that anyone with such wierd
hardware can supply the udeb for themselves (it would require
recompiling kbd-chooser anyway), so drop them from the CD.
- Alastair
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Bug#219868: choose-mirror: Go Back does not work properly
On Sun, 2003-11-09 at 20:58, Denis Barbier wrote: > Package: choose-mirror > Version: CVS > Severity: normal > Tags: d-i > > Hi, > > if you select http protocol, continue until the last question, > go back to the first question and select ftp protocol, then > choose-mirror exits before reaching the last question. > > Denis > I've just fixed this bug in CVS. Could you please retest choose-mirror? Thanks, Alastair McKinstry -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: setting /etc/modules in d-i / base-config?
On Sat, 2003-11-08 at 23:59, Petter Reinholdtsen wrote: > [Alastair McKinstry] > > Do we (expect to) prime the /etc/modules on /target in d-i, in > > general, eg. for network interfaces, or should kbd-chooser add the > > USB modules if it spots a USB keyboard? > > My plan is to leave this to the HW detection system (discover), and > make sure it detect and loads the correct modules at boot time. > When it comes to USB devices, I guess hotplug is a must as well, to > make sure the required USB modules are loaded when the devices are > inserted as well. > This will not be popular with those who want to keep base small, and fast boots; but would probably be a good idea for modern desktops. It should be possible to do both: copy the current modules from d-i to /target, then the system will work; discover will then not be necessary, just a good addition. > > This could be a generic issue: eg if a PC has multiple network > > cards, and the modules are loaded at boot-time in a different order > > to that in d-i, the eth0 / eth1 aliases will be different. > > The detection order is as far as I know, the order on the PCI bus, and > this do not change without moving cards in the PC. For Disks, I believe so, but for ethernet cards, it depends on the order of modules loaded, or aliases in /etc/modules; eg alias eth0 e100 alias eth1 tulip sets things the reverse order to which they are setup in d-i. I have this in my modules setup to "autoload" e100 on reference to eth0. Regards, Alastair -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: updated lithuanian .po files
On Sun, 2003-11-09 at 00:04, KÄstutis BiliÅnas wrote: > Hi, > > In attachement there are some updated lithuanian (lt_LT) > .po files for d-i. Could someone commit them for me? > > Thanks, Committed. Thanks for your contribution to Debian. Alastair -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Bug#219765: Languagechooser 1.03 dies on mipsel
On Sat, 2003-11-08 at 20:48, Karsten Merker wrote: > Package: languagechooser > Version: 1.03 and 1.04-cvs > > On mipsel, languagechooser dies with a return-code of 1 when called > from main-menu (probably it segfaults). When called in a shell > via "udpkg --configure language-chooser" it just hangs. Further tests > have narrowed down the problem to the following code segment: > > # Only display the translated texts (ie the English "translation") > # when in UTF-8 mode. > if echo $LANG $LC_CTYPE | grep -q UTF-8 ; then > db_set debconf/language en > else > db_set debconf/language C > fi Hi, The code actually goes: echo $LANG $LC_CTYPE | grep -q UTF-8 if [ $? ] ; then ... Could you please test this on mipsel? I can't log into a mipsel box at the moment. if this fails, try the variant if `echo $LANG $LC_CTYPE | grep -q UTF-8 ` ; then ... Regards, Alastair McKinstry > The echo and grep processes are started, but the db_set is not called. > When the same sequence is run manually in a shell, it works as expected. > Exchanging the busybox-ash by a dash as /bin/sh does not change the > behaviour. > > If the sequence above is removed and replaced by a single > > db_set debconf/language C > > as a temporary workaround, language-chooser works on mipsel. > > Regards, > Karsten > -- > #include > Nach Paragraph 28 Abs. 3 Bundesdatenschutzgesetz widerspreche ich der Nutzung > oder Uebermittlung meiner Daten fuer Werbezwecke oder fuer die Markt- oder > Meinungsforschung. > -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
setting /etc/modules in d-i / base-config?
Hi, #219290 points out that USB keyboards don't work after reboot, as the necessary modules aren't loaded properly. Do we (expect to) prime the /etc/modules on /target in d-i, in general, eg. for network interfaces, or should kbd-chooser add the USB modules if it spots a USB keyboard? This could be a generic issue: eg if a PC has multiple network cards, and the modules are loaded at boot-time in a different order to that in d-i, the eth0 / eth1 aliases will be different. Regards, Alastair McKinstry -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: GOBACK in kbd-chooser
>Am Fr, den 07.11.2003 schrieb Petter Reinholdtsen um 10:52: >> > On user selecting GOBACK out of the module, exit 0, not 30. Its not >> > an error and main-menu shouldn't lower the debconf priority. >> >> How do you expect this to work? If the user chooses GOBACK, he expect >> to get to the previous question, not just continue like nothing >> happened. Returning OK (0) will make main-menu just select the next >> entry in the menu, right? > >Even worse, since the udeb will be left unconfigured it will select the >same entry again. Wit debconf priority=high, you won't have a chance to >break out, because the item will be selected automatically. > >So with the current model, the priority must be lowered to prevent such >a deadlock. > >Sebastian Ok, bad fix for a problem that was fixed properly elsewhere (in anna). Will back out tonight when I get home. - Alastair >-- >PGP-Key: http://www.mmweg.rwth-aachen.de/~sebastian.ley/public.key >Fingerprint: A46A 753F AEDC 2C01 BE6E F6DB 97E0 3309 9FD6 E3E6 > > > -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
Re: bogl-bterm: font reloading
>[Daniel Jacobowitz] >> Can you just restart bterm? Yes, I realize that's got some issues >> since you're running inside of it. But it's really not set up to >> re-parse the font. > >I would rather not restart debian-installer and main-menu, and it seem >to be a requirement to be able to restart bterm. > Why not? (I'd like to know what issues exist with restarting main-menu in particular). It appears from the bterm code that its going to be hard to get it to reparse the font without restarting the "client" program main-menu. - Alastair >-- >To UNSUBSCRIBE, email to [EMAIL PROTECTED] >with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED] > > -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
