Tarobish / Jomhuria

http://tarobish.github.io/Jomhuria/
SIL Open Font License 1.1
23 stars 6 forks source link

Categorising Arabic fonts #40

Closed Tarobish closed 8 years ago

Tarobish commented 9 years ago

Hi All, Here is a problem, normally in the softwares like (illustrator, indesign or Microsoft word) Persian/Arabic fonts are divided by a tiny line to separate them from latin, Chinese, etc... but Jomhuria, after installing it goes under Latin category (Amiri also facing a same problem) Does anyone has an idea?

davelab6 commented 9 years ago

cc @khaledhosny

I guessed maybe the codepage info was not set, but it was:

screen shot 2015-09-26 at 12 20 03

khaledhosny commented 9 years ago

No idea really, anyone knows what Adobe bases its classification on?

davelab6 commented 9 years ago

@readroberts is there an afdko method for setting this?

graphicore commented 9 years ago

ping

Oldouze commented 9 years ago

Hi there, Wasn't sure where else I could comment but just wanted to let you know that I downloaded and used this font and I really like it! It's beautiful and versatile. Quite excited to have this as part of my font collection. Well done!

solargaria commented 9 years ago

I trying it in Adobe Indesign. for make books in Jawi language (Old malay with Arabic scripts), It's working good... thank you

davelab6 commented 8 years ago

It seems that adobe software uses an unpublished heuristic to determine this, and only setting the codepage support to Arabic may work.

khaledhosny commented 8 years ago

Setting it to Arabic only?

graphicore commented 8 years ago

@Tarobish is Mirza and/or Katibeh working in this context?

I could compare some of the tables. OS/2, name what could be relevant here?

Tarobish commented 8 years ago

Yes, both are working perfect. i think "Glyphs App" did something by default to categorized them

On Feb 18, 2016, at 11:22 AM, Lasse Fister notifications@github.com wrote:

@Tarobish https://github.com/Tarobish is Mirza and/or Katibeh working in this context?

I could compare some of the tables. OS/2, name what could be relevant here?

— Reply to this email directly or view it on GitHub https://github.com/Tarobish/Jomhuria/issues/40#issuecomment-185873926.

davelab6 commented 8 years ago

This was discussed on the glyphs forum. https://forum.glyphsapp.com/t/place-font-on-hebrew-font-list/3856/10

davelab6 commented 8 years ago

@graphicore any luck with this?

graphicore commented 8 years ago

To speed the needle in the haystack search up a bit. @Tarobish for Mirza and Katibeh, under which names are the fonts reported in your InDesign menu for Persian/Arabic fonts?

Tarobish commented 8 years ago

The whole Arabic, Persian fonts are in the same place included Mirza and Katibeh

KB STUDIO

On Feb 26, 2016, at 5:09 PM, Lasse Fister notifications@github.com wrote:

To speed the needle in the haystack search up a bit. @Tarobish for Mirza and Katibeh, under which names are the fonts reported in your InDesign menu for Persian/Arabic fonts?

— Reply to this email directly or view it on GitHub.

graphicore commented 8 years ago

But not in Arabic script? It says in Latin "Mirza" and "Katibeh"?

davelab6 commented 8 years ago

Screenshot + unicode text that you see would be ideal

Tarobish commented 8 years ago

No, in the Arabic script, we just have this problem with jomhuria

screen shot 2016-02-26 at 6 11 52 pm

graphicore commented 8 years ago

@Tarobish can you please download and test this file: https://github.com/Tarobish/Jomhuria/raw/issue_40/generated/Jomhuria-Regular.ttf

graphicore commented 8 years ago

That's very off-topic. Please make an issue if it is so in the versions in gh-pages as well. The font at the link I posted is to work on the category problem.

Tarobish commented 8 years ago

Sorry :) Just deleted them all No still it doesn't

graphicore commented 8 years ago

NP, just make a new issue, describe what you did and the software and post a live testing page.

Tarobish commented 8 years ago

Sure :)

graphicore commented 8 years ago

@Tarobish: please download and test again: file: https://github.com/Tarobish/Jomhuria/raw/issue_40/generated/Jomhuria-Regular.ttf (Make sure you restart InDesign after installing the font.)

Tarobish commented 8 years ago

Still it doesn't

graphicore commented 8 years ago

This is a mystery. In the issue_40 I did:

>>> font['OS/2'].ulCodePageRange1 = 64
>>> font['OS/2'].ulCodePageRange2 = 0

and

>>> font['OS/2'].ulUnicodeRange1 = 8193
>>> font['OS/2'].ulUnicodeRange2 = 0
>>> font['OS/2'].ulUnicodeRange3 = 8
>>> font['OS/2'].ulUnicodeRaneg4 = 0

So that these values match exactly the values of Katibeh, which is reported to work. I don't have an idea which other value could be responsible for this. I coud try to use the Katibeh name and OS/2 tables completely and then, and if it works, go towards the Jomhuria tables until it breaks? Any Idea anyone? Could any other table be responisble? (name looks just fine for the purpose and is just like Katibeh, it has the same fields, so I rule that out)

@Tarobish are you sure your InDesign updated the font you used? Do you use your system-wide font installation or the InDesign folder?

https://helpx.adobe.com/indesign/using/using-fonts.html#installing_fonts

You can make fonts available in InDesign by copying the font files into the Fonts folder inside the InDesign application folder on your hard drive. However, fonts in this Fonts folder are available only to InDesign.

I believe I've heard that the InDesign folder was better to test fonts and I want to rule out it being a caching issue or such.

Tarobish commented 8 years ago

Yes I did but let me try one more time I'll delete everything, restart the whole system and re-install again

KB STUDIO

On Mar 1, 2016, at 1:09 PM, Lasse Fister notifications@github.com wrote:

This is a mystery. In the issue_40 I did:

font['OS/2'].ulCodePageRange1 = 64 font['OS/2'].ulCodePageRange2 = 0 and

font['OS/2'].ulUnicodeRange1 = 8193 font['OS/2'].ulUnicodeRange2 = 0 font['OS/2'].ulUnicodeRange3 = 8 font['OS/2'].ulUnicodeRaneg4 = 0 So that these values match exactly the values of Katibeh, which is reported to work. I don't have an idea which other value could be responsible for this. I coud try to use the Katibeh name and OS/2 tables completely and then, and if it works, go towards the Jomhuria tables until it breaks? Any Idea anyone? Could any other table be responisble? (name looks just fine for the purpose and is just like Katibeh, it has the same fields, so I rule that out)

@Tarobish are you sure your InDesign updated the font you used? Do you use your system-wide font installation or the InDesign folder?

https://helpx.adobe.com/indesign/using/using-fonts.html#installing_fonts

You can make fonts available in InDesign by copying the font files into the Fonts folder inside the InDesign application folder on your hard drive. However, fonts in this Fonts folder are available only to InDesign.

I believe I've heard that the InDesign folder was better to test fonts and I want to rule out it being a caching issue or such.

— Reply to this email directly or view it on GitHub.

Tarobish commented 8 years ago

just updated the indesign and we are fine :)

graphicore commented 8 years ago

Really? Which font did work? Is there a difference between the one in the issue_40 branch and the version in gh_pages?

Tarobish commented 8 years ago

Yes :) the one in the issue #40 the second one that you sent me this one https://github.com/Tarobish/Jomhuria/raw/issue_40/generated/Jomhuria-Regular.ttf

graphicore commented 8 years ago

Yeah, give it a trial! But I think the version in between (the first I posted today) is the one that made the difference. So, please can you test this one as well: https://github.com/Tarobish/Jomhuria/raw/94ae05b2f5f987ea91c25c32bfc292dbfff58359/generated/Jomhuria-Regular.ttf (use the link here, it's from the history of issue_40)

Tarobish commented 8 years ago

Sure :)

KB STUDIO

On Mar 1, 2016, at 2:48 PM, Lasse Fister notifications@github.com wrote:

Yeah, give it a trial! But I think the version in between (the first I posted today) is the one that made the difference. So, please can you test this one as well: https://github.com/Tarobish/Jomhuria/raw/94ae05b2f5f987ea91c25c32bfc292dbfff58359/generated/Jomhuria-Regular.ttf (use the link here, it's from the history of issue_40)

— Reply to this email directly or view it on GitHub.

Tarobish commented 8 years ago

yeah, its fine :)

graphicore commented 8 years ago

yeah, its fine :)

Which one?

Tarobish commented 8 years ago

The link in you last comment

Yeah, give it a trial! But I think the version in between (the first I posted today) is the one that made the difference. So, please can you test this one as well: https://github.com/Tarobish/Jomhuria/raw/94ae05b2f5f987ea91c25c32bfc292dbfff58359/generated/Jomhuria-Regular.ttf (use the link here, it's from the history of issue_40)

graphicore commented 8 years ago

OK, so that refers to this change:

>>> font['OS/2'].ulCodePageRange1 = 64
>>> font['OS/2'].ulCodePageRange2 = 0

Which essentially says the only codepage that "is considered functional" is Arabic.

Before we had bit's set for a lot of Latin based codepages as well. Seems that Adobe's algorithm just prefers Latin. I'll update the main build script to set these Flags.

graphicore commented 8 years ago

I updated the font in gh-pages. Feel free to test it one last time.

davelab6 commented 8 years ago

Thought it would be codepages! Cool!