Maintenance of this package and alternatives

[Goldziher]

Hi there,

This package appears to be unmaintained.

We forked it - and are publishing our own maintained fork as xberg-ttf-parser. Its a part of the https://github.com/xberg-io/xberg repository (see crates).

You can point further fixes and issues there if you rely on it.

Edit:

If you are a maintainer of ttf-parser itself and are reading this, please respond to this issue and tell us if you do intend to maintain this package going forwards.

[Goldziher]

@tobocop2 FYI

[alerque]

I think the main development effort by most interested parties has completely shifted to fontations as replacing this entirely and eventually seeing this project deprecated. In fact I think @behdad had in mind doing that sooner rather than later (c.f. harfbuzz/rustybuzz#166 (comment)).

I might be able to jump in and help drag out the maintenance a bit longer with patch releases if there is an interest and use for that. I think it would be necessary to say what the value proposition is for that? And why would migrating to fontations not be more viable than forking or contributing to the maintenance of this?

[Goldziher]

To be explicit about what the fork actually is, and to credit the people whose work it carries: it is v0.25.1 plus nine open pull requests from this repo, cherry-picked as-is. None of it is ours except #228.

Upstream PR	Author	Fix
#203	@dereified	Document that Face::style() falls back to the head mac style bits
#207	@asibahi	Face::set_variation returns None for an axis the face does not define
#216	@SamusAranX	Read the fvar HIDDEN_AXIS flag from bit 0, not reserved bit 3
#222	@scadastrangelove	Guard the CFF2 blend operator against an empty argument stack
#223	@scadastrangelove	Promote avar axis value mapping to i32 so it cannot silently wrap
#224	@scadastrangelove	Cap total component visits in glyf / gvar composite outlining
#225	@scadastrangelove	Cap total paint-graph node visits in COLRv1 painting
#226	@asibahi	Parse loca for a font carrying the maximum 65535 glyphs
#228	@tobocop2	Ignore the deprecated dotsection operator in CFF charstrings

Each row is recorded in the fork's README with the upstream PR and the cherry-pick commit:
https://github.com/xberg-io/xberg/blob/main/crates/xberg-ttf-parser/README.md#carried-upstream-prs

Nothing was rewritten or reattributed. Licensing is unchanged — MIT, retaining the upstream copyright notice.

Worth saying that several of these are correctness bugs rather than cleanups, and they are all still open here with no maintainer response: #226 drops every TrueType outline in a 65535-glyph face, and #228 makes i, j, ! and . render blank in fonts converted from Type 1.

[Goldziher]

Thanks for responding @alerque - we would of course prefer not to maintain this crate. Lets see. If you do decide to deprcate this project, we are fine with maintaining our own fork. So lemme know.

Our fork carries most of the content of the open PRs presently on this repo and some modernization.

[behdad]

Can you please send a PR to add your fork note to this package and declare it deprecated?

[Goldziher]

Can you please send a PR to add your fork note to this package and declare it deprecated?

sure

[behdad]

I mean, I'm also fine giving you access to maintain here. With a proper "maintenance mode" note.

[Goldziher]

That would be preferable. If you could give me and @tobocop2 maintainer access, we can take over. happy to do so.

@behdad

[behdad]

Done.

[Goldziher]

closing this issue first. Thanks!

[asibahi]

You should include this PR as well as it is a bugfix to match code with spec: #212

[tobocop2]

Done.

thank you!

[behdad]

If you don't mind, please add big "maintenance-only" and "fontations" replacement notice at the beginning of the readme. Thanks.

[Goldziher]

Sure.