[multikulti-tech] Bengali help pages errors
Dan McQuillan
dmcquillan at lasa.org.uk
Tue May 18 17:04:04 BST 2004
hi andrew
it's great to hear from you! the original Multikulti pilot project took some of its inspiration from the open road project and from centrelink.gov.au.
i remember exchanging emails with larry stillman in the early days, after seeing a short newspaper article about vicnet/openroad.
> -----Original Message-----
> From: andrewc at vicnet.net.au [mailto:andrewc at vicnet.net.au]
> Subject: [multikulti-tech] Bengali help pages errors
> Message = Hi I was just looking at your site, and reading the
> Bengali help pages. There are a number of errors on the page.
yes, it's taken a while to figure out what's going on with bengali :(
i'll update the help pages using your info.
> The only version of uniscribe that supports Bengali, is the
> version available with Office 2003. Office uses its own
> version of uniscribe, rather than the system version.
aha, i didn't know that.
is it possible to copy that over the system version so it's available to other programs?
>To use Bengali iin other uniscribe aware programs it is necessary to
> unpdate the system version of uniscribe.
yes, i notice that the site for Avro Keyboard offers a download of an updated usp10.dll
http://www.omicronlab.com/avrokeyboard/configuring_system.htm but i don't know how reliable/legal that is. I've seen other people talking about having a VOLT licence to get access to updated uniscribe versions...
>The first available
> uniscribe update form Microsoft will be in the Windows XP
> Service Pack 2, allowing anyone using WinXPSP2 to correctly
> render Bengali Unicode, assuming they have a Bengali OpenType font.
on the subject of fonts, have you any experience of the fonts from the Free Bangla Fonts Project
http://www.nongnu.org/freebangfont/? i'd prefer to be recommending open source fonts if possible.
Deepayan Sarkar, one of the developers from that project, has sent me some helpful emails about how the whole open type/rendering thing works on linux systems.
> Its possible to fudge and get IE and Mozilla Firefox
> correctly displaying Bengali Unicode, but meesing about with
> system DLLs is probably not advisable for the general user of
> your site.
i'd be interested to know more about that :)
it occured to me a while ago that we should provide links to localised versions of mozilla/firebird from the multikulti site. localisation seems to be one of the strengths of the open source movement.
i saw in a recent newsletter on IndLinux.org http://www.indlinux.org/nl/nl150404.html that there is a Mozilla binary built with Pango for enabling complex script rendering available at
ftp://ftp.mozilla.org/pub/mozilla.org/mozilla/releases/mozilla1.6/contrib/mozilla-i686-pc-linux-gnu-gtk2-pango.tar.gz but i haven't tried it.
> The second problem on the page is that you recommend the use
> of the font "Arial Unicode MS" for Bengali. I know of two
> versions of this font, shipped in Microsoft products at
> different times. The more recent version (v1.0) has much more
> support than the earlier version.
i seem to have version 0.84. would an XP Pro system have the more recent version?
>Neither version contains
> OpenType layout tables for Bengali and neither is suitable
> for displaying Bengali Unicode.
ah. back to my question about open source fonts then.
there seem to be several being developed for indic languages http://www.indictrans.org/src/fonts/
which is positive.
> >From my knowledge of Microsoft Typographies timeline, full
> support for Bengali can be expected in Longhorn when it is
> finally released.
one of the headaches for me has been finding unicode based software for bengali & gujarati so we can generate utf-8 content. we have been using specialist (non-unicode) software up to now.
(hence my interest in Avro Keyboard).
we've decided to switch to Word for gujarati, but we'll be using a new unicode version of the specialist software for bengali. (have you heard anything about how good Word is for gujarati?)
we had some problems when we first switched to unicode for farsi, includin the well-know farsi letter yeh glitch. i actually found the open road farsi support page helpful in solving that - thanks :)
i'm having problems right now with urdu. our translator is using Word 2003, and it mostly seems OK except with the letter HAMZA (\u 0621) followed by the letter SHEEN (\u 0634). Do you know of a way to get these characters to combine correctly?
anyway, thanks again for writing with some helpful suggestions about bengali. i'm trying to pick your brains some more because it's hard to find people knowledgeable about this stuff.
i'd also be like to know more about what you're up to with the openroad project. maybe there's more ways we can collaborate.
best
dan mcquillan
[Dr. Dan McQuillan/Project Leader/Multikulti Project/020 7247 7226/www.multikulti.org.uk]
p.s. although Multikulti is a general resource for UK community groups, it's been really noticeable that libraries are some of the biggest fans of the site. We've just tranlsated a leaflet about libraries for the education section of multikulti, in collaboration with a project called Welcome To Your Library.
More information about the multikulti-tech
mailing list