Ad by jey

Forum

 

15 posts

"The kerning is set in a way that doesn't work at dafont (we use the GD library of PHP)"

Jun 01, 2019 at 01:18

Hi, I've recently submitted a font called Running Start Basic, and I've gotten this message as the reason it wasn't accepted. I've never had that problem before, so I'm lost as to what to do. I'm using FontForge to design this font. I need help diagnosing the issue.

Is kerning not working at all? For example, does the sequence VA leave a giant space in the middle, or do they sit close to each other?

I've seen this problem mentioned elsewhere on the forum: is the problem the fact that I have contextual alternatives for K, R, and T, which do not have Unicode code points? Will I need to map the glyphs to a character value in Private Use Area?

Thanks for your cooperation! I hope this issue will be smoothly resolved.


Jun 01, 2019 at 05:07

Kerning individual pairs or Use a matrix of kerning classes

http://designwithfontforge.com/en-US/Spacing_Metrics_and_Kerning.html


Jun 01, 2019 at 11:49

claudeserieux said  
Kerning individual pairs or Use a matrix of kerning classes

http://designwithfontforge.com/en-US/Spacing_Metrics_and_Kerning.html

That's what I'm doing.


Jun 01, 2019 at 14:02

I received the same message from Dafont after I submitted HFF Warped Zone. The font only used class kerning and only the key characters were in the old-style kern table. The rest of the kerning were in the kern feature. What I did was expand the class kerning and created a flat old-style kern table. I retained the class kerned kern feature. I resubmitted the font and it was accepted by Dafont. No other changes was made to the font.

I do not know what GD library of PHP is or does but it appears that it does not support OT kern feature.


Jun 01, 2019 at 14:14

I'm indeed using a kerning table. How would I expand the table?


Jun 01, 2019 at 15:53


Edited 2 times. Last edit on Jun 03, 2019 at 00:46 by Lemmiwinks


Jun 01, 2019 at 15:59

I meant how would I convert my kerning matrix into that?


Jun 01, 2019 at 17:46

If you are using Fontlab, there is an option that will automatically create an old style kern table out of your class kerning when you generate a font.


Jun 03, 2019 at 09:44

I'm not using Fontlab, and I will not pay $459 to upload one free font onto one website.

I've sent the updated version and they haven't gotten back to me.

http://www.mediafire.com/file/aulbr0d97ldgakh/Running_Start_Basic_1.1.zip/file

I've never had the problem with kerning tables before, and I doubt that that's the problem here. Do get back to me as swiftly as you can.

Edited on Jun 03, 2019 at 09:46 by JapanYoshi


Jun 03, 2019 at 17:48



Jun 03, 2019 at 19:35

JapanYoshi said  
I'm not using Fontlab, and I will not pay $459 to upload one free font onto one website..

$48

https://www.fontlab.com/font-editor/typetool/


Jun 04, 2019 at 07:26


I don't know how that second image can ever occur, because R.salt only appears before an A, and never an R.


Jun 04, 2019 at 17:12

What Claude is trying to say is what's with feature RQD? That's probably the reason why he provided the link to the filespec.

As for the R.salt [R/r] pair, Cluade was trying to show you the kerning issue between the two characters. That kerning issue was not addressed by your kern feature.

Lastly, what's with those PURCHASE PREMIUM TO UNLOCK glyphs? I count over 500 of them. If you did not intentionally put them there, I suggest that you remove those glyphs as they just unnecessarily bloat the font's file size.


Jun 04, 2019 at 19:30

Au lieu RQD -> ss01

https://docs.microsoft.com/en-us/typography/opentype/spec/featurelist

Use the .notdef glyph for PURCHASE PREMIUM TO UNLOCK ( au lieu d'utiliser le Catalan )

https://docs.microsoft.com/en-us/typography/opentype/spec/recom


Jun 05, 2019 at 01:14

toto@k22 said  
What Claude is trying to say is what's with feature RQD? That's probably the reason why he provided the link to the filespec.

I don't see the problem; on all the programs I've used, including Notepad and Inkscape, the character substitution works correctly.

toto@k22 said  
As for the R.salt [R/r] pair, Cluade was trying to show you the kerning issue between the two characters. That kerning issue was not addressed by your kern feature.

That letter sequence never occurs; K.salt, R.salt, and T.salt only appear before [A/a], so the right side bearing is adjusted for that glyph. Would I have to use CALT for this instead?

toto@k22 said  
Lastly, what's with those PURCHASE PREMIUM TO UNLOCK glyphs? I count over 500 of them. If you did not intentionally put them there, I suggest that you remove those glyphs as they just unnecessarily bloat the font's file size.

That glyph replaces all the special characters supported by the paid version, Running Start Premium, as opposed to the glyphs that are NOT in the paid version, which fall back to the .notdef glyph.

Edited on Jun 05, 2019 at 01:31 by JapanYoshi



All times are CEST. The time is now 23:47


 
Privacy Policy  -  Contact