FontLab Forum
2012-02-09, 03:16:20 *
Welcome, Guest. Please login or register.

Login with username, password and session length
News: Welcome to the FontLab forum, read how to use it! Update: Archives from old MSN forums are now available on our forum.
 
   Home   Help Search Calendar Downloads Tags Login Register  
Del.icio.us Digg FURL FaceBook Stumble Upon Reddit SlashDot

Pages: [1]
  Print  
Author Topic: user error? [posted in proper subforum this time.]  (Read 2407 times)
Interested_User
Jr. Member
**

Karma: +0/-0
United States United States

Posts: 1


« on: 2009-06-12, 00:21:52 »

Hello. Long time user, second time poster. (I'd like to say that I really like the SBL Hebrew font and think it's magnificent.)

I have some text that doesn't quite render properly, and so I'm wondering if I'm doing something incorrectly. Everything mentioned is (AFAIK) up-to-date and the most recent version.

I have the text <shin + shin dot + segol + tav>. This displays fine.
If this undergoes normalization to NFC, it becomes <shin + segol + shin dot + tav>. This too displays fine.

So far, so good. If for other reasons, however, I want to prevent the above reordering, I think I'm supposed to insert a CGJ character mark such that I have the sequence <shin + shin dot + CGJ + segol + tav>. This does not display fine. Attached are the outputs as given by both Word 2007 and xelatex.

Is my understanding of the "mark reordering" section of the SBL manual defective? Is there something defective with Uniscribe/whatever's underneath xetex? Or is this behavior caused by the font itself?

Thanks,

-Interested User.
Logged
tiro_hudson
Moderator
Hero Member
*****

Karma: +8/-0
Canada Canada

Posts: 85


WWW Email
« Reply #1 on: 2009-06-12, 17:11:14 »

The doubling of the segol and, hence, insertion of the dotted ring, is a known bug in Word 2007. This is a new bug that was not present in earlier versions. Microsoft are aware of the problem, but I don't know whether we can expect a patch update.

The xetex rendering problem is new to me. It looks like their rendering engine is not removing the CGJ from the glyph string prior to GPOS, which is necessary to allow the mark to be correctly positioned relative to the base glyph.

I can experiment with removing the CGJ using a font lookup, but I am not sure whether this will work reliably and will need to ensure that it won't break anything.
Logged
Tags:
Pages: [1]
  Print  
 
Jump to:  

Powered by MySQL Powered by PHP Powered by SMF | SMF © 2011, Simple Machines Valid XHTML 1.0! Valid CSS!