Posted by:
John Hudson I'm looking into the low punctum problem, but I'm 99% sure that it is due to a bug in Internet Explorer. The codepoint we are currently using for the low punctum U+0323 is not specifically a Hebrew mark, and some applications have trouble handling it in right-to-left text. Other Microsoft applications used to have this bug, including Office Xp, but it has been fixed in Office 2003.
Note, Chris, that your 'crude fix' is only removing the dotted circles (indicating invalid mark sequences): another symptom of the problem -- at least on my system -- is that the font is being switched, resulting in those ugly square dots instead of the SBL Hebrew puncta. This is what I am seeing with your HTML, with the mark incorrectly encoded before the base:
And this is what I get in Word 2003, with the mark correctly encoded after the base:
Important note: the upper puncta are actually positioned outside the vertical metrics of the font, which means they will be clipped and invisible in some applications, but they should print correctly (I made the above screenshot by printing to a PDF). Because the puncta are so seldom used, I did not want to have them dictate the default linespacing for the font. This means, apart from the clipping issue, that passages using the puncta cannot be 'set solid' but must have additional linespacing added in the application.
