[odf-discuss] Re: Interop between multiple standards and multiple applications [new thread]

[Lars Noodén]

robert_weir at us.ibm.com wrote:

> OK.  Let me work this through in my head.  So we define a "ODF Interop 
> Profile" which is a restriction of the full ODF, essentially a subset of 
> the schema, plus perhaps some other restrictions (all fonts must be 
> embedded?  ...

OMG no!  It's a good topic to have brought up, but embedding the fonts
adds a whole slew of technical and licensing problems that are unlikely
to be solved in the foreseeable future.  A *separate* initiative to
develop and market Free fonts, however, would dovetail nicely.

An intermediate solution might be to refer to fonts the way that DTDs
are supposed to be referred to and used: have a standard URI for the
font, whether Free or proprietary.  However, the fonts problem is an old
plague for all users, esp. those moving around between different
machines with MS Office.

> So if we had this, then I think we solve most interop issues between ODF 
> editors, i.e., between OpenOffice, KOffice, etc. 

However, to get there, an essential first-step, technically and
strategically, is to have ODF-compliant applications.

In regards to legacy apps like MS Office, MS talked a lot of crap a few
years back regarding XML support.  IFF (If and Only If) MS Office can
support arbitrary, user-supplied XML schemas, then ODF support can be
achieved by loading the XML-schema.  If not, then a false advertising
lawsuit on behalf of consumers is in order.

Increasing general XML support in applications would go a long way in
furthering interoperability with or without ODF.  I've yammered on and
off since 1995 / 1996 about making web browsers into generalized
rendering engines with a small handful of pre-loaded DTDs/schemas +
stylesheets.  That would make it possible to have at least universal
viewing of ODF documents: browser could be pre-loaded with XHTML,
Docbook and ODF (along with supporting stylesheets), users could then
load in additional DTD/schema as needed.

Back 11- 12 years ago, Softquad's Panorama was an SGML-based example,
proving that it is both doable and marketable.  What's needed now would
be for the leaders (Opera, Konqueror and Mozilla) to get on board.  One
might also look for allies among whoever is currently making Dreamweaver
and XMetal.

Regards
-Lars