April 2003 - openSUSE M17n - openSUSE Mailing Lists

update: yudit 2.7 packages available for SuSE Linux 8.1 and 8.0
by Mike FABIAN 15 Jan '04

15 Jan '04

Updated yudit packages are available for SuSE Linux 8.0: ftp://ftp.suse.com/pub/people/mfabian/8.1-i586/yudit-2.7-0.i586.rpm ftp://ftp.suse.com/pub/people/mfabian/8.1-src/yudit-2.7-0.src.rpm ftp://ftp.suse.com/pub/people/mfabian/8.0-i386/yudit-2.7-0.i386.rpm ftp://ftp.suse.com/pub/people/mfabian/8.0-src/yudit-2.7-0.src.rpm >From the NEWS file of yudit: NEWS> Changes since version 2.6.4: NEWS> --------------------------- NEWS> NEWS> o Added full Unicode compliant bidirectional text support NEWS> o Added fallback to root xinput style if no suitable style is found NEWS> o Received and added Polish gui translations, Persian.kmap, Sanskrit.kmap NEWS> Inuktitut-ICI.kmap, Inuktitut-KBD.kmap, Chinese-Pinyin.kmap, NEWS> Chinese-WB.kmap, ArmenianEastPhon.kmap NEWS> o Speed Optimizations NEWS> o OpenType GPOS support for composing characters NEWS> o Better TAB control NEWS> NEWS> Other Notes: NEWS> ----------- NEWS> Even if you can not read bidirectional scripts, it will be NEWS> useful to type this in the editor command area: NEWS> NEWS> howto bidi NEWS> NEWS> to get some ideas what bidi is, and how it works in Yudit. NEWS> Even if you can not read RL scripts, RL words will be indicated NEWS> in yellow (the default RL color). >From yudit's home page at http://www.yudit.org: > Yudit is a unicode text editor for the X Window System. She can do > True Type font rendering, printing, transliterated keyboard input and > handwriting recognition with no dependencies on external engines. Her > conversion utilities can convert text between various encodings. > Keyboard input maps can also act like text converters. There is no > need for a pre-installed multi-lingual environment. Menus are > translated into many languages. -- Mike Fabian <mfabian(a)suse.de> http://www.suse.de/~mfabian 睡眠不足はいい仕事の敵だ。

1 4

update: mined 2000.3 package available for SuSE Linux 8.0
by Mike Fabian 04 Aug '03

04 Aug '03

An updated package of mined is available for SuSE Linux 8.0: ftp://ftp.suse.com/pub/people/mfabian/8.0-i386/mined-2000.3-0.i386.rpm ftp://ftp.suse.com/pub/people/mfabian/8.0-src/mined-2000.3-0.src.rpm Mined is a text editor, small and easy to use, yet powerful. It is Unicode-capable, using UTF-8 encoding (also converts UTF-16 input). >From the CHANGES file: ============================================================================= Changes from mined 2000.2 -> mined 2000.3 ========================================= Documentation: -------------- The manual page was updated and thoroughly revised. Its primary source was transformed into HTML, roff/man format being generated. The mined web pages were updated and revised. The file "doc/compilation" (with compilation hints) was updated. Features: --------- HOP '/' enters an indented Javadoc comment frame. Interface: ---------- Options can be concatenated on the command line. (mined -uu instead of mined -u -u enables Unicode line separators.) Right-to-left script input support is now enabled by default. The option -b toggles it. Option -G disables (actually toggles) the display of certain control characters as block graphics characters (enabled by default). Option +G enforces use of block graphics for display of menu borders. May be used if the "alternative character set" capability is not configured in your system but your terminal does have the capability. Set output delay (-d0..-d9) to none (-d-) by default in all versions. The help command was unified on Unix and DOS versions. Not completely, however, as calling a sub-programm ("less") through the "system" call doesn't seem to work on DOS (it crashes or blocks terminal input afterwards, even with cygwin). See also next comment. A second help viewing mode is now avaible (HOP HELP, e.g. HOP F1). It displays help information within mined itself, restoring the previous editing state afterwards. This is the default on MSDOS for the reason mentioned before. Bug fixes: ---------- Fixed a screen-related pointer confusion after replacement with multiple lines (embedded newline) which could result in a display bug and page down could be blocked. (Very minor) Just deleting the trailing line-end of a file is also considered a modification (a file only modified this way would previously not have been saved automatically). (Minor) With UTF-8 auto-detection involved, the character count after reading the file could be wrong (ESC ? would have been correct). (Minor) The pop-up menu, when modified with HOP and thus becoming smaller, used to leave the frame of its previous width on the screen. (Embarassing) Although I was proud of "perfect responsiveness to window size changes" I had just forgotten to implement that for the case of a menu being open. On DOS, editing a file with Unix line-ends, the function "append to buffer" used to append the lines with MSDOS line-ends. -- Mike Fabian <mfabian(a)suse.de> http://www.suse.de/~mfabian 睡眠不足はいい仕事の敵だ。

3 10

Japanese menu problems under IceWM
by ピアスエリック 28 May '03

28 May '03

Howdy. I don't think this is SuSE specific, so this post may not belong here. But it is Japanese related, and I've posted this twice to IceWM's mailing lists and received 0 replies. I compiled IceWM 1.2.0 for international support (--enable-i18n). In the program menu, IceWM seems to have a problem (bug?) with any folder or program with Japanese characters for its name. Here's a screenshot which demonstrates the problem. http://epierce.freeshell.org/menu.png If you'll note, the four (automated by IceWM) Japanese entries up from the bottom have proper quick-keys assigned (i.e., the (A), (T), etc.), but the 7-8 Japanese entries that I added all have a problem (i.e., just a little dot shows up). After a quick look at 'ja/LC_MESSAGES/icewm.mo', I tried to copy what was going on there by adding something like (_M) after the Japanese in the 'menu' file in hopes of assigning my own quick-key, but that didn't work. IceWM only seems to be looking for roman characters for its automatic quick-key assignment. Anyone have any ideas here? Thanks for reading... Eric P __________________________________________________ Do You Yahoo!? Yahoo! BB is Broadband by Yahoo! http://bb.yahoo.co.jp/

3 6

[m17n]: SuSE 8.1 / kinput2 / canna
by David Nettles 14 May '03

14 May '03

I have successfully installed SuSE 8.1 on 4 different environments with Japanese support enabled. All of the installations have occured on Western hardware (in the US), but the installation was done in Japanese (changing to Japanese mode at the beginning of the installation). On one machine, though, things are not working quite right. It seems that I am missing something. As a user, I log in by selecting "IceWM" from the kdm selector. I then perform the following steps: * open a kterm * in the kterm I type the following on the command line "echo おに" * before pressing return, I press the space bar with the intent of having it do a lookup for a reasonable kanji (in this case I intend to search for "鬼") * instead of giving me a first reasonable kanji or a box with choices, it gives me a small popup windows with the message "かな漢字変換サーバと通信できません" I later re-ran kinput from the command line ("kinput2 -xim -kinput -canna") and received the same error message as above. As root I ran "/etc/init.d/canna full-status" and it reported the following: Checking for Canna Kanji Server: running Error Disconnected This machine is configured to get all of its network stuff via the default DHCP config that was configured by SuSE 8.1's installer. I have verified that files that should exist in /etc as well as /var/lib/canna match the files as found on a model working system (the one that I am typing this on now). I've searched groups.google.co.jp for other folks with config issues looking for more things to check, but in general my systems are pretty safe. No tricks installed. All users have LANG=ja_JP by default (I did trying changing it to LANG=ja_JP.EUC per another's recommendation, but it did not change the behavior). The .profiles are unchanged defaults. For the Japanese users. No changes made. This problem effects the users on only this one new installed system. 3 of the systems (including the problematic one ) are light "KDE, GNOME, windowmaker, ICEWM, mwm, qvwm, twm, failsafe" installations. All help is appreciated. -- david email: tetsuoni3000(a)yahoo.co.jp web: http://www.miteyo.org

2 2

new utf-8 support for LaTeX
by Gerhard Schuck 25 Apr '03

25 Apr '03

Has anyone noticed that there exists an alternative to ucs-style to use utf-8 encoding with latex? Have a look at: http://www.latex-project.org/cgi-bin/ltxbugs2html?pr=latex%2F3480 I tried it and it seems to work, also together with cjk-latex. But I didn't really test it. It is said to be more lightweighted and much less ressource hungry than ucs-style. However I have no idea, whether this is an advantage or not. Basic installation info: You have to create two files by copy and paste from the last part of the above homepage: utf8ienc.ins utf8ienc.dtx With "latex utf8ienc.ins" you produce some more files. Copy utf8.dev, omsenc.dfu, ot1enc.dfu, t1enc.dfu to a directory recognized by latex and execute texhash. Then try "latex utf8-test". It should produce a dvi file with Umlauts. Notice that the latex run produces one error, which is intentionally added. If you look at the utf8-test.tex file you understand. Regards Gerhard Schuck

1 0

WindowMaker update for SuSE Linux 8.2
by Mike FABIAN 24 Apr '03

24 Apr '03

Here are updated WindowMaker packages for SuSE Linux 8.2 with improved UTF-8 and Xft2 support: ftp://ftp.suse.com/pub/people/mfabian/8.2-i586/WindowMaker-0.80.2.20030423-… ftp://ftp.suse.com/pub/people/mfabian/8.2-i586/WindowMaker-devel-0.80.2.200… ftp://ftp.suse.com/pub/people/mfabian/8.2-src/WindowMaker-0.80.2.20030423-0… ftp://ftp.suse.com/pub/people/mfabian/8.2-i586/wdm-1.20-500.i586.rpm ftp://ftp.suse.com/pub/people/mfabian/8.2-src/wdm-1.20-500.src.rpm ftp://ftp.suse.com/pub/people/mfabian/8.2-i586/wmakerconf-2.9-93.i586.rpm ftp://ftp.suse.com/pub/people/mfabian/8.2-src/wmakerconf-2.9-93.src.rpm The bug reported by Togan Muftuoglu on this list about "Loosing locale settings" (see http://lists.suse.com/archive/m17n/2003-Jan/0019.html) should also be fixed in these packages. These packages are a CVS snapshot of WindowMaker where I added some patches to enable Xft support *always*, even on displays without RENDER extension, because Xft2 can be used always. You can still switch Xft support off in the config files by setting "AntialiasedText = no;", but it is now possible to use the same fonts via Xft also on displays without RENDER. This won't work on 8.1 because 8.1 didn't yet have Xft2, therefore I uploaded these updates only for 8.2. I think the anti-aliased fonts look quite nice in WindowMaker and the font setup is dramatically easier. wsetfont sans or wsetfont serif now writes font patterns like WindowTitleFont=-xftdummy-sans-bold-r-normal-*-12-*-*-*-*-*-*-* ~/GNUstep/Defaults/WindowMaker and ~/GNUstep/Defaults/WMGLOBAL which works in almost all locales without change. And it is very nice to be able to setup preferred sans or serif fonts in ~/.fonts.conf, and get the effect throughout the system for all applications using Xft2. Until now mostly Qt3/KDE3 and Gnome2 applications, but now also WindowMaker. If you use wdm and wmakerconf, you have to use the updated packages from pub/people/mfabian as well because they use libs from WindowMaker and needed to be updated because of the WindowMaker update (But wdm is pretty buggy, better don't use it). wmakerconf still has some problems when running in UTF-8 locales, I'll try to fix that later. -- Mike Fabian <mfabian(a)suse.de> http://www.suse.de/~mfabian 睡眠不足はいい仕事の敵だ。

1 0

Loosing locale settings
by Togan Muftuoglu 24 Apr '03

24 Apr '03

Hi, I do not know if this was happing before yet recently I discovered that sometimes I am loosing my locale settings when I have more then 2 Xterms opened. The weird thing is it does not happen always SuSE 8.1 Kernel k_deflt-2.4.19-74 Windowmaker (WindowMaker-0.80.2-1 from Mike Fabian's directory) I have the locale settings in $HOME/.profile as export LANG=tr_TR.UTF-8 export LC_MESSAGES=en_US export LC_TIME=POSIX export LC_COLLATE=tr_TR.UTF-8 I am using "utf8xterm" which comes with shtools-2002.01.30-117. So When I have the Xterm for the first time the locale is toganm@earth:~> locale LANG=tr_TR.UTF-8 LC_CTYPE="tr_TR.UTF-8" LC_NUMERIC="tr_TR.UTF-8" LC_TIME=POSIX LC_COLLATE=tr_TR.UTF-8 LC_MONETARY="tr_TR.UTF-8" LC_MESSAGES=en_US LC_PAPER="tr_TR.UTF-8" LC_NAME="tr_TR.UTF-8" LC_ADDRESS="tr_TR.UTF-8" LC_TELEPHONE="tr_TR.UTF-8" LC_MEASUREMENT="tr_TR.UTF-8" LC_IDENTIFICATION="tr_TR.UTF-8" LC_ALL= When I encounter the problem of loosing locales the output is like this LANG=tr_TR LC_CTYPE="tr_TR" LC_NUMERIC="tr_TR" LC_TIME=POSIX LC_COLLATE=tr_TR.UTF-8 LC_MONETARY="tr_TR" LC_MESSAGES=en_US LC_PAPER="tr_TR" LC_NAME="tr_TR" LC_ADDRESS="tr_TR" LC_TELEPHONE="tr_TR" LC_MEASUREMENT="tr_TR" LC_IDENTIFICATION="tr_TR" LC_ALL= I have no idea whatsoever how come the locales change does anybody have ? -- Togan Muftuoglu Unofficial SuSE FAQ Maintainer http://dinamizm.ath.cx

2 5

Japanese input packages for LiveEval
by Yasunari Imado 17 Apr '03

17 Apr '03

Hello. Isn't there any good idea which includes a Japanese input package(e.g. Kinput2 and Canna) in a LiveEval version? -- -O_O-O_O-O_O-O_O-O_O-O_O-O_O-O_O-O_O-O_O-O_O-O_O- Yasunari Imado sjv(a)iris.eonet.ne.jp (PGP Key ID: 2582B3D6) [SSJV] http://ssjv.hmc5.com/ [DTL] http://www.eonet.ne.jp/~vazzy/ -O_O-O_O-O_O-O_O-O_O-O_O-O_O-O_O-O_O-O_O-O_O-O_O-

2 6

Unicode, Russian and SLPro 8.2
by Anders Karlsson 17 Apr '03

17 Apr '03

Hi there, I have not yet managed to try out SLPro 8.2, and I wanted to check here on the list what changes has been made to the language environments between 8.1 and 8.2. Is there anyone that can give a quick run-down of the tweaks that has been made? Regards, -- Anders Karlsson <anders(a)trudheim.com> Trudheim Technology Limited

1 0