============================================================================== Bengali Khanda Ta In Bengali a dead consonant TA shows up as U+09CE BENGALI LETTER KHANDA TA in all contexts except where it is immediately followed by one of the consonants TA, THA, NA, BA, MA, YA, or RA. Khanda-Ta cannot bear a vowel matra or combine with a following consonant to form a conjunct aksara. It can form a conjunct aksara only with a preceding dead consonant RA, with the latter showing up as a REPH placed on the Khanda Ta. Previous versions of the Unicode Standard recommended that Khanda-Ta be encoded as TA + VIRAMA + ZWJ. Instead, the Khanda-Ta should be used explicitly in new text, but users are cautioned that instances of the old encoding may exist. ============================================================================== Date: 23 Apr 2004 09:01:11 -0000 From: Dr Anirban Mitra Hope you are in the process of developing the next version of Yudit. Your script processor is the best in Free software as far as Bengali (an other indic) script processing is concerned. It has helped us a lot in creating the first Bengali GNU/Linux distro called "Ankur Bangla LiveCD" . You will be happy to know that Yudit is included as third party software in the distro to help translator in translating the GUIs like GNOME & KDE in Bengali. 1. As per specification of Unicode 4.01 ra+ yaphala(ra + post-base form of ya) is to be separately coded from Ya+reph (above-base form of ra + ya) as below: ra+zwj+virama+ya > ra+yaphalaa ra+virama+ya > ya+reph The rendering has to be adjusted accordingly. 2. As per detailed discussion in unicode indic mailing list, The Khanda-Ta issue has been decided thus: Khanda-Ta will be accepted as a special type of Half form of ta that behaves like a halant form of ta. For example the cluster ta+virama+zwj+ka+matra-i will be rearranged as khanda-ta+matra-i+ka and not matra-i+khanda-ta+ka as being done now. 3. It has been a good idea to assign specific range to specific fonts so two fonts don't mix. But there is a problem with Devanagari Danda (x0964) & double danda (xo965) as well as vedic stress marks udatta ,anudatta etc. (x0951-0953). Though they are included in Devanagari block they are common punctuation mark for all indic language and they should be set to font of the character immidiately preceding the mark for aesthetic look. Now some personal request: 4. Please include MuktiNarrow.ttf available from http://savannah.nongnu.org/download/freebangfont/MuktiNarrow-0.94.tar.bz2 as the default Bengali font as it is good looking and is packaged as the default Bengali UI font for most GNU/Linux distros like Fedora, Debian, Mandrake as well as our AnkurBangla distro. ============================================================================== Date: Sun, 9 Nov 2003 22:09:17 +0100 From: eisvogel To: Gaspar Sinai Subject: Patch Yudit 2.7.6 / kmap / termination sequence & encoding Parts/Attachments: 1 Shown 32 lines Text 2 Shown 110 lines Text ---------------------------------------- Dear Gaspar, The files listed below proableby do have an encoding other than utf-8. When read those, a lot of 0xFFFD show up. mabye iso-8859-7: GreekBible.kmap GreekPolytonic.kmap maybe koi8-r: Russian-Translit-Nordic.kmap Russian-Translit-Slovene.kmap Ukrainian-Translit.kmap Eisvogel ============================================================================== Gaspar Sinai 2002-11-12 Need to enhance Lao: Requestor: Anthony Souphavanh font: http://laonux.muanglao.com/fonts/hongkadu.ttf Need to enhance Tibetan Requestor: Gregory Mokhin font: http://www.yudit.org/download/fonts/utibetan.ttf ============================================================================== Need kmap for Lao, Thai otherwise in theory they work In addition extra work may need to be done to peroperly support: o Sinhala o Khmer U+1780-U+17FF o Myanmar U+1000-U+109F o Tone letters 02E5..02EB Version 2.5 broke background printing. It shuld be restored. The problem is that all events are handled now in a job and when a job is executed the job is removed already. So event next does not call the updating job! Flickering should be eliminated by using off-screen rendering. Availability of on-screen rendering should be kept because of the low memory consumption. The following classes need cleanup: swindow/SX11* Make focus work on multiple panes, modal handling cleanup. Remove unncecessary code Handle colormap change. Add one big happy pixmap instead of creating many small ones (like in swindow/swin32/SWin32.cpp) Layered unicode support should be done in 3.0. A good reference: Metacode: http://www.cs.fit.edu/~satkin/docs/dissertation.pdf 3. Make a configuration screen A configuration screen would make Yudit more easy to use 4. Add an ability to specify more unicode ranges for fonts. (currently indic range is covered) 5. Adjusting of the line gaps as some fonts. 6. Make kmap list translatable. 7. Add more tables to: doc/otfsupport.txt ============================================================================== Gaspar Sinai gsinai@yudit.org Tokyo 2006-05-21