Re: [wmvs] Requirements for "Valid" Badges

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Janet Daly <janet@w3.org> wrote:

>Your complaints of delays are valid, and I am responsible for those
>delays in response, as the former Comm Team Lead. Please accept my
>apologies.

Gladly; and I appreciate the reply.


>Perhaps we'll have them both solved in a short time.

That would be wonderfull.


>> 4. The fonts used must be embeddable — including optimization for
>>    download time by removing unused glyphs from badge files — and not
>>    encumbered by any “Intelectual Property” restrictions (since they
>>    will be part of the Validator source code and published under an OSS
>>    License) beyond the normal Logo and Trademark guidelines.
>
>What does this mean? Do you mean that you would like us to use only
>OpenSource fonts in our communications materials? I don't know if I can
>promise that; I only want to be clear on your request.

Well, ideally “Open Source” font faces — such as “Bitstream Vera” — would be
used. This would let us include SVG format masters in the source code for the
Validator under the same license as the rest of the code.

Any font with a limitation or encumberance on its distribution would require
us to go to ridiculous lengths in order to satisfy those limitations, our
internal development requirements, and external user requirements.

For instance, any font whose license does not allow us to include the original
font files themselves in the source distribution is probably out of the
question. Any font which would not let us embed a few of the glyphs — e.g. in
an SVG master — would make it very hard to use.

In addition to the straightforward Copyright-related license issues, I'm told
some fonts have yet further restrictions in some juridistictions. e.g. while
the US does not allow restrictions on the “shape” of a font (IIRC), some other
juridistictions /do/; and thus even a free clone or derivative of such a font
would become encumbered in those juridistictions.

And finally, some fonts might be distributed under a license that prohibits
subsetting of the available glyphs or similar modifications. In a context such
as providing badges in SVG format, this could lead to the ridiculous situation
of having to either include the full font in each badge (leading to very large
file sizes) or not embedding the font at all.


In any case; our main concern with fonts useage for the badges is not whether
they're “pretty”, suitable for print or web use, legible, etc. (I expect you
have all those sorts of issues well in hand). Our requirements stem from the
fonts being, from the perspective of the Validator, essentially “source code”
and thus needs to allow modification and redistribution under the same terms
as the surrounding source code.

In particular, it's not enough to distribute binary objects (PNG, GIF, SVGz,
etc.) of the pre-rendered badges. We need to include the “source code” masters
in a way such that the recipient could conceivably regenerate — using Open
Source tools —  all the badges from the master templates.

Typically this would be to implement a new compression algorithm, converting
PNG alpha channel to simple transparency for MSIE:win (a specific use case
which has cropped up lately), adding Dublin Core or RDF metadata, adding a
custom glyph for the checkmark in the badges, or similar purely
technical/format transformations.


But do note that this is in regards fonts __as_used_in_the_badges__ and not
w3.org font useage in general. I assume validator.w3.org and related sites
would quickly adopt any w3.org general font policy for web pages, but here the
CSS font fallback mechanism would make any encumbered fonts non-problematic.

It's specifically the font used for the text in the “Valid This” badges that
concern us.



Back in 1998 or thereabouts, my then employer set about on a grand project to
replace their original web site — which was rather austere and simplistic, but
Valid HTML 3.2; guess who designed it? :-) — with one designed by a local ad
agency. When the time came for the launch, every employee computer had one of
the “Copperplate” (98BC?) fonts installed so the site would “look right”...

It was quite amusing to hear the complaints of the marketing department, upon
returning from a customer and having demonstrated our shiny new web site, that
the pages looked horribly ugly and the pretty font didn't show up. Not quite
as amusing when the proposed workaround was to turn the pages into a single
large image with an imagemap for links.


The moral of the story? Font useage needs to be _useable_ first, and æstethic
second; and in our case, we have some additional use requirements above what
is normally necessary for design for web publishing.


- -- 
My mom is a professional botanist, or, as her spousal equivalent described
it, they'll be out hiking in the woods, she'll see a plant off by the side
of the trail, run up to it, bend down, and start talking Latin at it.
                                                      -- Steve VanDevender

-----BEGIN PGP SIGNATURE-----
Version: PGP SDK 3.2.2

iQA/AwUBQoMCj6PyPrIkdfXsEQJ69ACfY6IxAUfJc6s5T1t85B74q/9Df8EAmweW
t1DNOBMjRqnE4/mP2elreEe7
=d7nW
-----END PGP SIGNATURE-----

Received on Thursday, 12 May 2005 07:15:36 UTC