I've discovered that FontLab appears to ignore whatever code I put into my 'aalt' feature and simply generates its own code for this. There are some features which I would like to exclude from 'aalt' but which FLS seems determined to add back in. Is there some preference that I am missing in FLS, or is there simply no way to control the 'aalt' feature from within FontLab?
André
Have you verified this e.g. with TTX, OTM, or another tool?
I could be wrong, but vaguely remember that 'aalt' is exported correctly when generating OTFs/TTFs with FLS5, but FLS5 creates a new aalt when importing OTFs/TTFs. So that what you see after importing to FLS5 is not what FLS5 exported ... (Curious for results of your check.)
Thanks for your response, Karsten.
My original suspicion was due to opening the generated font files in FontLab.
I've now tried opening them in both ttx and OTM, and in neither of these can I find *any* evidence of an 'aalt' feature at all, though I admit that I'm not very comfortable with interpreting the output of either of these tools.
André
UPDATE: While I can't find my 'aalt' feature in either ttx or OTM Lite, I am getting little triangles showing up in InDesign's glyph palette, and they seem to be what I'd expect from my aalt feature rather than the one FLS5 is displaying, so it appears that your assessment was correct.
Hello André, I made a quick test too and export seems to work fine – subtables referenced by lookups referenced by 'aalt' have the expected content.
Karsten (Vielen Dank!) was able to determine that my 'aalt' feature was indeed not being generated and that the problem was with my code rather than with FontLab Studio.
Now I have a followup question:
How many users would be terribly annoyed by a font which lacked an 'aalt' feature altogether?
I ask this because the face I'm working on has a rather large glyph complement and some fairly complex code. I've spent a lot of time trying to optimize the code because earlier versions were getting coverage offset overflow errors when compiling. I'd finally managed to pare things down so that it would compile, but that was while I was under the mistaken impression that my 'aalt' feature was in fact being generated. Once I restore it, the GSUB table again becomes too large.
To me, 'aalt' seems less critical than the other features in the font, but I don't know what problems (if any) might result from omitting it.
Opinions?
André
Captured from commoncrawl.org on 18 Apr 2015. original URL