RE: The pointless thread continues

From: John Hudson (
Date: Fri Jul 05 2002 - 17:59:55 EDT

At 14:19 7/5/2002, Michael Everson wrote:

>The only solution I see is that whatever data exists in ASCII hacks or
>other web font solutions has to one day be mapped and converted to proper
>Unicode sequences, if the data is to be saved. How much discussion does
>this need?

I've obviously missed some of the earlier parts of this discussion, but
when was it established that web font solutions implied ASCII hacks or
other non-Unicode approaches. When I saw the term web font being used, I
presumed it meant something like Embedded OpenType or similar server-side
resource, and I'm not aware of any reason why this should not be a clean
Unicode implementation. My understanding of Michael Jansson's approach is
that he tries to provide readable text to any browser *in a manner best
suited to that browser's capabilities*, which in effect does mean that some
browsers end up with a hacked font, but this is not intrisic to his
approach, only to the limitations of those particular browsers. At least in
early beta versions of Fairy, there was an even worse fallback position,
which rendered text in some browsers as a graphic. My understanding is also
that, in all such cases, the backing string of the original website
encoding is preserved and can be accessed by turning off the webfont display.

I can think of all sorts of instances on today's Internet in which 'install
a font' will not be an option. I have clients who want to present text on
their websites in corporate typefaces that they have paid large sums of
money to have designed. These typefaces are integral parts of the visual
identity of these companies, and having their websites viewed in Times New
Roman is not desirable. On the other hand, they do not want to distribute
the fonts in which they have invested so much money and whose exclusivity
is a key aspect of their value. Such companies require a server-sider
webfont solution. Personally, I would be happiest to see all browsers
supporting the Embedded OpenType format, because I think this provides the
best results. As long as different browsers implement competing formats, I
think there is a place for Michael Jansson's approach, which tries to
provide the best results possible in a myriad of circumstances.

John Hudson

Tiro Typeworks
Vancouver, BC

Language must belong to the Other -- to my linguistic community
as a whole -- before it can belong to me, so that the self comes to its
unique articulation in a medium which is always at some level
indifferent to it. - Terry Eagleton

This archive was generated by hypermail 2.1.2 : Fri Jul 05 2002 - 16:03:54 EDT