OASIS Mailing List ArchivesView the OASIS mailing list archive below
or browse/search using MarkMail.

 


Help: OASIS Mailing Lists Help | MarkMail Help

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

RE: Images embedded in XML



Right.  I know it can be done and have 
seen it done in earlier systems such as 
Interleaf.   I'm not sure given the existing 
capabilities it is usually a good idea 
given reuse and reliability.  Can one do it; 
sure.  Should one do it?  Only warily.  A 
spec for it isn't the differentiator; gravy maybe.

Today one usually wants to inquire if one 
wants to:

1.  Wait for a spec to emerge.
2.  Use an existing engine.
3.  Write a new engine because an 
    emerging spec is incomplete, too simple, 
    or has spawned half a dozen other specs 
    to appear to be simple which ends up
    meaning "incomplete" and "waiting".

In the *real* world we hear so much about, 
implementors are usually digging around in a 
box looking for a working widget to use instead 
of waiting to build a new widget as soon as they 
are sure the widget will be standard.  Moving 
the competition from companies to committees has 
that side effect of late bloomers.

If a document on the web were a self-contained system,  
URIs would be outlawed.  :-)

Len 
http://www.mp3.com/LenBullard

Ekam sat.h, Vipraah bahudhaa vadanti.
Daamyata. Datta. Dayadhvam.h


-----Original Message-----
From: Benjamin Franz [mailto:snowhare@nihongo.org]

On Thu, 5 Apr 2001, Bullard, Claude L (Len) wrote:

> If this works this well, why does HTML not 
> embed images in the HTML page?

You can - it isn't a limitation of HTML. URI type 'data:' from RFC2397
<URL:http://www.landfield.com/rfcs/rfc2397.html>. Netscape has supported
it for a long time - but MSIE didn't last time I checked (a few years
ago).

<IMG
    SRC="data:image/gif;base64,R0lGODdhMAAwAPAAAAAAAP///ywAAAAAMAAw
AAAC8IyPqcvt3wCcDkiLc7C0qwyGHhSWpjQu5yqmCYsapyuvUUlvONmOZtfzgFz
ByTB10QgxOR0TqBQejhRNzOfkVJ+5YiUqrXF5Y5lKh/DeuNcP5yLWGsEbtLiOSp
a/TPg7JpJHxyendzWTBfX0cxOnKPjgBzi4diinWGdkF8kjdfnycQZXZeYGejmJl
ZeGl9i2icVqaNVailT6F5iJ90m6mvuTS4OK05M0vDk0Q4XUtwvKOzrcd3iq9uis
F81M1OIcR7lEewwcLp7tuNNkM3uNna3F2JQFo97Vriy/Xl4/f1cf5VWzXyym7PH
hhx4dbgYKAAA7"
ALT="Larry">

Try it in Netscape and it will work. Try it in MSIE, and it probably will
not. Since MSIE has 86%+ of the market, designers won't use it at all.
That is the same reason designers don't use the 'OBJECT' tag, either (now
*that* was a real crime - MSIE almost certainly had to have
*intentionally* broken it. The tag was designed for transparent backward
compatibility, but MSIE throws major hissy fits over it and made it
impossible to deploy for *anything* except ActiveX.) The real question
becomes "Why doesn't MS want to support RFC2397?".

> Would one want to lose the XML payload if the image does 
> not make it, the connection drops, etc.?
> As long as I can remember, embedding binary 
> in markup has been discouraged.  
>
> When would putting the binary in the XML 
> be a good idea?

When you need to distribute a self-contained system.