[XML-SIG] XBEL / Call for extension

[Christian Junk]

Am Freitag, 28. Januar 2005 15:50 schrieb Uche Ogbuji:
> On Fri, 2005-01-28 at 11:09 +0100, David Faure wrote:
> > Yes, but at the same time it should be useable as an interoperable
> > format. So icons need to be in there - in a simple way, yes.
> > Icons are really an integral part of bookmarks nowadays.
>
> OK.  I think we have to agree to disagree on this one, but I'm just one
> person.  What do others think?  If everyone does seem to feel there
> should be one standard way to have icons, I'll defer on that point.

I think icons should be included in a new XBEL version, although it is one of 
the not so necessary features - it's not more than an eyecandy.

Including the icons as base64-encoded data is one of the easiest way, but if 
your bookmark collection is large enough, then the icon data will blow up the 
XML file.

> Meanwhile, am I correct that XBEL 1.1 is a dead letter?  It's bundled up
> into PyXML, but it's not linked from the Web site, or anything.

I think 1.1 is a dead letter, too. Perhaps this is the only official link:

http://pyxml.sourceforge.net/topics/dtds/xbel-1.1.dtd

> It's 
> also not really ready for prime time (the icons/NMTOKEN head-scratcher,
> for example).  Last time XBEL evolution came up, Martin mentioned 1.2 as
> the natural next version.

> Should we just say on the XBEL page that there was an experimental 1.1,
> but that it is withdrawn (sorta like XSLT 1.1) and that we're working on
> 1.2?
>
> Then we could start hashing out issues such as icons for 1.2, and get it
> out, already.

One of the greatest lacks is the definition of the METADATA section. So far, 
there is no solution for this problem. Please correct me if I'm not right!

Perhaps we should think of using XML Scheme instead of DTD?

Regards,
Christian

-- 
Christian Junk <junkc at fh-trier.de>
FH Trier, University of Applied Sciences
Faculty of Design and Applied Computer Science

http://christianjunk.webinternals.de
http://xbel.webinternals.de