Bug 68179 – Remove SVG_FONTS define as it's enabled on all major ports

• Component: New Bugs
• Status: RESOLVED WONTFIX

[Eric Seidel (no email)]

Remove SVG_FONTS define as it's enabled on all major ports

[Eric Seidel (no email)]

Created attachment 107525 [details]
Patch

[Eric Seidel (no email)]

Unfortunately I don't think we can land this until Qt ships 4.8.  Their 4.7 version disables SVG_FONTS.

I'm told 4.8 was supposed to ship 2 weeks ago, but may be another 2 weeks yet.

[Eric Seidel (no email)]

It's likely this patch will break in between now and when we can execute on this, but at least it's a starting-point/reference.

[Alexis Menard (darktears)]

(In reply to comment #2)
> Unfortunately I don't think we can land this until Qt ships 4.8.  Their 4.7 version disables SVG_FONTS.
> 
> I'm told 4.8 was supposed to ship 2 weeks ago, but may be another 2 weeks yet.

The RC is soon to be released. I bring the discussion on how it is acceptable to remove the entire SVG module when building with Qt 4.7. Stay tuned!

[Darin Adler]

Comment on attachment 107525 [details]
Patch

View in context: https://bugs.webkit.org/attachment.cgi?id=107525&action=review

> Source/WebCore/dom/DOMImplementation.cpp:75
> -#if ENABLE(FILTERS) && ENABLE(SVG_FONTS)
> +#if ENABLE(FILTERS)

Is FILTERS an SVG-only feature? If not, should this be ENABLE(FILTERS) && ENABLE(SVG)?

[Eric Seidel (no email)]

Comment on attachment 107525 [details]
Patch

It's unclear if filters is SVG only or not.  If it's SVG-only, I'll remove the FIlTERS define next.

[Eric Seidel (no email)]

I'll wait to land this until I hear back from the Qt guys.

[Adam Barth]

Is it time to land this patch now?

[Eric Seidel (no email)]

If Qt has transitioned to their next version, sure.  I don't know what their status is.

[Alexis Menard (darktears)]

Go ahead Eric as per :

http://trac.webkit.org/browser/trunk/Tools/qmake/mkspecs/features/default_pre.prf#L73

and http://trac.webkit.org/changeset/102766

we enforce Qt 4.8 in WebKit trunk.

Ping me on IRC (darktears) if any problem.

[Ryosuke Niwa]

Wait, don't we want to get rid of SVG fonts? I thought we are trying to push back on SVG fonts. e.g. acid3 no longer tests SVG fonts. Do we really want to enable this on all ports?

[Eric Seidel (no email)]

(In reply to comment #11)
> Wait, don't we want to get rid of SVG fonts? I thought we are trying to push back on SVG fonts. e.g. acid3 no longer tests SVG fonts. Do we really want to enable this on all ports?

I'm not familiar with the context of that discussion.  Do you have a mailing list thread link? (from webkit-dev, mozilla, standards, or whatever?)

[Anne van Kesteren]

See e.g. http://lists.w3.org/Archives/Public/www-svg/2011Nov/0036.html

[Ryosuke Niwa]

Acid3 has dropped tests for SVG fonts: https://plus.sandbox.google.com/107429617152575897589/posts/JdHnqpuUER4

I've asked around on #whatwg, and the general consensus appears to be that we need a better spec. And I believe some vendors such as Mozilla doesn't want to implement SVG fonts at all:
http://robert.ocallahan.org/2010/06/not-implementing-features-is-hard_03.html

(with sicking's help)

[Ryosuke Niwa]

Also see the thread on SVG WG:
http://lists.w3.org/Archives/Public/www-svg/2011Oct/0124.html
http://lists.w3.org/Archives/Public/www-svg/2011Nov/0000.html

I'm very skeptical that enabling SVG fonts on all ports is the right choice for the project given there have been significant push backs from Mozilla and Microsoft and the relevant WG is indecisive about whether they should include it in the future spec or not.