Re: Request: Techniques for reducing the size of XML instances

[mrossi@csc.com]

     Well we've really come full-circle haven't we? :-) The various
minimization features of SGML (SHORTTAG, DATATAG, OMITTAG) would mostly
allow for your first option. You'd need SOME tags, but how many depends on
the structure of your vocabulary. As for your second option, you're right.
Lots of people are pursuing some form of binary XML. Unfortunately, someone
else will have to provide the roundup as I don't have a good grip on all of
those at the moment.

     This topic comes up occasionally, so you might want to check back
through the archives for threads on compression and such. I recall there
were also some techniques discussed for compression based on lement
patterns or some such thing, very similar to SGML minimization. Hope that
helps.

Michael A. Rossi
Computer Sciences Corporation
mailto:mrossi@csc.com
856-983-4400 x4911


                                                                                                                   
                    "Roger L.                                                                                      
                    Costello"            To:     xml-dev@lists.xml.org                                             
                    <costello@mit        cc:                                                                       
                    re.org>              Subject:     Request: Techniques for reducing the size of XML instances   
                                                                                                                   
                    07/26/01                                                                                       
                    12:06 PM                                                                                       
                                                                                                                   
                                                                                                                   


Hi Folks,

Does anyone have a summary of techniques for reducing the size of XML
instances (as would be required in limited bandwidth applications)?

XML instance minus tags:

Are there techniques for reducing the size of XML instances by simply
stripping off the tags (thus maintaining an ASCII document)?  [It is not
clear to me how the receiver of such a tagless document would regenerate
the original instance document.]

Binary Compression:

I am sure that there are lots of tools to binary compress XML
instances.  Does anyone have a summary of such tools?  /Roger