Release TTFs
Closed this issue · 29 comments
To onboard the family I need TTF files, but these are not in the 1.000 release or gh-pages branch
@davelab6 What are the best tools to generate ttf's?
OTF>fontforge>TTF>ttfautohint>
cc @paarandika We have to add this ttf+ webfont build thing to make_release script, And pyftsubset thig to subset if wanted.
No, making an OTF then converting to TTF bloats the filesize :(
https://github.com/mooniak/ayanna-font/tree/master/sources has many files; which are actually used to build?
Ufo's are combined and built during the process, So combined masters are not included in the repo.
Sure, which files are actually used to build?
I see in https://github.com/mooniak/ayanna-font/blob/master/scripts/builder.sh that SFDs are the 'real' sources
So, why are the intermediate UFOs checked into the repo, if they are just temporary build files? :)
yes, kept the sfd generated ufo's considering collaboration. Since Sol's files are also coming in as UFO's.
Any luck?
I'm developing a script. Most probably it will work :)
Added a crude script to build separate ttf for Sinhala and Tamil
Awesome! I'll take a look next week
We have problems with the files. @paarandika posted me the generated files.
- Some glyphs are flipped
- Overlap is not removed (fontconvert can do --remove overlap)
I am testing the files now and will post a complete review soon.
Which glyphs are flipped?
I did checkoutlines using same script as in Hindkit. It should remove overlaps..
I'll post in detail, open the ttfs and see
On Sat, Mar 5, 2016, 2:27 PM Ayantha Randika notifications@github.com
wrote:
Which glyphs are flipped?
I did checkoutlines using same script as in Hindkit. It should remove
overlaps..—
Reply to this email directly or view it on GitHub
#52 (comment).
෴ ජය! ෴
පැතුම් එගොඩවත්ත
Pathum Egodawatta
pathum@mooniak.com
Could not find any issues
Yep, it was my mistake. It works,but there is a problem with /space. It is not interpolating it seems. I have to check again with against otf's. Lets keep this open till @davelab6 take a look at this. And @paarandika we have few things to clean up
- need separate family names for ttf's Sin+ Latin and Tamil+ Latin (#47)
- Move
ttf-buildinto./build/ttf - Move
gf-buildscripts into./scriptsfolder. So its clean. - Update readme with info on build process and how its done. A simple description is fine.
One and four I will do and I will check "space" as well.
For three and four Nope
Ooops!
One and four I will do and I will check "space" as well.
For two and three Nope
Cool. BTW https://github.com/rosettatype/yrsa-rasa may be helpful
@davelab6 The ttf's are like this, DO you think there is a better way to do this?
ttf on the left and otfs on the right
@paarandika among other things, I think its better if we can just build the combined ttf's and then split with pyftsubset. Instead of maintaining multiple feature files
Yeah, So excess points are added during conversion to TT outlines? @paarandika So we only have to figure out what's happening with glyphs with changed directions.
These glyphs have problems in curves, I think they are all using componants and FF complier mess up that.
tml_Debit, tml_Ee, tml_Four, tml_KSsU, tml_KSsUu, tml_KSsa, tml_Month, tml_NUu, tml_Nine, tml_NnUu, tml_NnnUu, tml_NyU, tml_NyUu, tml_TUu,
si_DdR.halant, si_DdRI, si_DdRIi, si_FR.halant, si_FRI, si_FRIi, si_GR.halant, si_GRI, si_GRIi, si_GhR.halant, si_GhRI, si_GhRIi, si_HR.halant, si_HRI, si_HRIi, si_JhR.halant, si_JhRI, si_JhRIi, si_KhR.halant, si_KhRI, si_KhRIi, si_MR.halant, si_MRI, si_MRIi,
Otf's are perfectly fine. I thik we could use Unlink Reference in FF python to fix this on export.
Yeah, waiting for Randika to look at the files.