[sc34wg3] CTM: The arguments for standardization

[Lars Marius Garshol]

* Robert Barta
| 
| I see this pretty relaxed. In AsTMa? I had experimented with
| XTM output. In TMQL itself this comes quite natural:
| 
|      using xtm for http://theuglybeast/xtm/1.1
|      .....
|      return
|         <xtm:topic id="{$a}">
|            ...
|            <xtm:baseNameString>Look ma, XTM</xtm:baseNameString>
|         </xtm:topic>
| 
| So I would assume we can leverage that somehow into the UPDATE
| language.

Yes, it may be that we can do that, in order to support both "camps"
syntactically. Semantically it should be all the same anyhow.
 
| I would also claim that CTM parsers are almost one magnitude faster
| in parsing topic map instance.

I don't think that claim can be made generally. The speed of the CTM
importer relative to the XTM importer in any given topic maps
implementation will depend on the speed of the XML parser relative to
the parser generator used for CTM. 

In the case of the OKS the LTM parser is slower by roughly an order of
magnitude, which, interestingly, is the opposite of your result. The
only thing I would conclude from that is that antlr could do with some
performance optimizations. :)

-- 
Lars Marius Garshol, Ontopian         <URL: http://www.ontopia.net >
GSM: +47 98 21 55 50                  <URL: http://www.garshol.priv.no >