444571 - Mozilla should distribute a "Free core fonts" set as part of the Firefox distribution

Status: RESOLVED WONTFIX

(Firefox :: General, enhancement)

Description

[Neil Harris]

What happens now:
* If a web page specifies the use of a high-quality Free font such as DejaVu, SIL Gentium or Computer Modern within a web page, very few browsers will be able to display it as intended.

What should happen:
* All Firefox users (and eventually all or almost all users of other browsers, see below) should be able to see the text displayed in the correct font

I suggest that Mozilla should make a set of these already-existing high-quality available as part of the Firefox distribution, either by packaging them with the browser, distributing them via software update or install-on-demand download, or by making them available via extensions.

Advantages:

This would greatly increase the number of high-quality fonts available for use on the Web, for Firefox users on all operating systems. Since Firefox's market share is now greater than 20% worldwide and rapidly growing, this should be sufficient to create a network effect in favour of the use of these fonts (which will be completely backwards compatible: see below). Their presence on large numbers of web pages will encourage other browser and operating system vendors to package these fonts as part of their distributions.

In particular, this will be useful for users of languages written in non-Roman scripts which are not widely supported by native fonts in most computer operating systems. Examples of non-Roman fonts include Georgian, Amharic, Burmese, Tibetan, Malayalam, Sinhalese, Cree, Inuktitut and Cherokee. This will also be useful for users of languages written in Roman scripts, such as Vietnamese, which use unusual characters or diacritic combinations which are not widely supported in mainstream fonts.

Finally, the availability of Free mathematics fonts for Firefox will greatly help the adoption of MathML.

Disadvantages:

None. Because of the pre-existing provisions for fallback in HTML/CSS font specifications, and the continuing market dominance of Internet Explorer, web page designers will continue to use font fallbacks appropriately, and the use of these fonts should not disadvantage the users of any other browser or operating system.

Comment 1

[David E. Ross]

Once @font-face is implemented per bug #70132 (see also the discussions at bug #52746), this RFE would not be necessary.  Then, Web developers will have the non-exclusive choice between either creating pages that are adaptable to whatever fonts the users have or else embedding fonts that reside on the page's server.  Users will not have to download and install fonts for viewing Web pages.  

Comment 2

[Klaus Malorny]

I don't know how far the actual implementation is, but based on the age of the bug (nearly eight years), supplying additional fonts is IMHO a reasonable interim solution. 

Comment 3

[Jesse Ruderman]

The age of the bug is irrelevant.  What is relevant is that it's about to be fixed.

Comment 4

[Neil Harris]

Can you point me to any evidence of that? All I can see in either of those two bugs is discussion, much of it devoted to intellectual property issues, rather than code development, even of an experimental testing-only implementation. 

Implementing a Free Software font installation mechanism would perform an end-run around several different problems:

* implementation difficulties of implementing dynamic temporary font loading

* intellectual property issues regarding linking vs. embedding and font copyrights (since fonts would be permanently installed, and would only involve fonts which are already explicitly licensed for universal distribution) 

* security problems regarding dynamic font loading (since Mozilla-curated fonts would have a known provenance, and could be verified before installation, either using digital signatures or hashes)

Comment 5

[Jesse Ruderman]

The bug has blocking1.9.1+ and has an active owner (John Daggett).  I also know from chatting with him in person that he is trying to determine whether it's safe to throw arbitrary files at OS font APIs.

Comment 6

[Mike Connor [:mconnor]]

bundling things that aren't Firefox isn't something we're really going to do.  I like great fonts almost as much as anyone, but distribution of stuff that's somewhat-related-but-not-required, and especially installing stuff into system locations just feels a little too bundleware for my comfort level.