RE: WOFF and extended metadata

On Monday, May 24, 2010 10:03 AM Jonathan Kew wrote:
> 
> On 24 May 2010, at 14:34, Levantovsky, Vladimir wrote:
> >
> > The important point here (and I just want to emphasize it once again)
> is that neither David nor I (nor anyone else at this point) is asking
> to reject the font if something is wrong with the metadata.
> >
> > The key requirement, and is the only one we are asking for, is for a
> browser to be able to parse the extended metadata *at user's request*!
> 
> I completely agree that it's desireable for browsers to be able to
> parse (and display) metadata.
> 
> However, I don't think it is appropriate for the WOFF spec to try and
> mandate this. There will be some situations (e.g., mobile phones) where
> it may not be reasonable to provide "Page Info" or similar UI options
> where displaying font metadata makes sense, but my feeling is that
> trying to define exactly which UAs are required to have such a feature
> is a mistake. 

I agree with you. This is exactly the position I took when I proposed the subject of metadata up for a discussion. [1]

> We could argue almost endlessly over how to specify this
> for today's range of products, let alone dealing with tomorrow's UAs
> that may not fit neatly into whatever categories we can currently
> imagine and define.
> 

+1

> The WOFF spec should ensure that there's a well-understood,
> interoperable format for the kinds of metadata that font vendors want
> to provide, and that users may want to see. 

Agree.

> How (and even whether) that
> data is exposed by a given UA should be determined by the UA developers
> in the light of their particular product's design and environment, user
> requirements, etc.
> 

I agree that *how* the data is exposed by a given UA should be up to that UA developers.
As to *whether* it is exposed or not - I believe this is one of those rare cases when the use of SHOULD, as defined by the RFC 2119 [2] would be a perfect fit for the situation at hand. As defined by [2], "SHOULD" means that

   "there may exist valid reasons in particular circumstances to ignore
   a particular item, but the full implications must be understood and
   carefully weighed before choosing a different course."

I've never asked for user interface displaying font metadata and its features be mandated by the WOFF spec, and I've never asked to REQUIRE that browsers provide it. What I am asking for is the statement be added to Extended Metadata section that would say something along the following lines:

"If the metadata is present, browsers SHOULD provide the means for the Extended Metadata content to be displayed at user's request."

That's all! I believe this would satisfy all the needs of both browser and font vendors. This statement would acknowledge the fact that UA developers may have valid reasons in particular circumstances to ignore this item, but it would also allow the WOFF spec to define recommended behavior we all seem to agree on. Having this statement will also emphasize the importance for metadata to be included, and would clearly state the fact that the metadata content is intended to be seen by a user, with display being the only function expected to be supported by UA.

Regards,
Vlad

[1] http://lists.w3.org/Archives/Public/public-webfonts-wg/2010May/0126.html
[2] http://tools.ietf.org/html/rfc2119

Received on Monday, 24 May 2010 15:19:11 UTC