- From: Felix Sasaki <fsasaki@w3.org>
- Date: Thu, 16 Jun 2005 18:55:45 +0900
- To: Andrzej Zydron <azydron@xml-intl.com>
- Cc: public-i18n-its@w3.org
Hi Andrzei, I would propose then that we have a reference to CAM in the paper, that's it. If there is a stable implementation later, we can make more use of CAM in the recommendation. Best, Felix. Andrzej Zydron wrote: > > Hi Felix, > > Felix Sasaki wrote: > >> Hi Andrej, >> >> Thanks for the links, I already had a look in the PDF and the PPT >> before I wrote the mail to you. >> >>> >>> CAM is not just about NOUN definitions. CAM can define >>> BusinessUseContext rules, DataValidations, ContentReference rules as >>> well as Assembly structures. All very powerful and flexible. >> >> >> >> >> I saw the mechanisms you're mentioning, also the list of predicates >> you pointed me to. I'm just not sure of the interplay between all >> these. I also had a look at the CAM specification >> http://www.oasis-open.org/committees/download.php/5914/OASIS-CAM-Specifications-1_0-RC-017C-021904.doc >> >> But it didn't make things clearer to me. > > > > Yes - it is not the most clear documentation in the world! > > >>> >>> My current understanding (still incomplete) of CAM is that it should >>> provide a sufficient vocabulary >> >> >> >> >> Is it a vocabulary? or rather a set of predicates with some >> processing directives? > > > You are correct - it is about predicates and processing instructions > > >>> to be able to describe semantically the requirements for >>> http://esw.w3.org/topic/its0505Translatability, >>> http://esw.w3.org/topic/its0505ReqAttrAndTrans, >>> http://esw.w3.org/topic/its0505WordCount and possibly provide a >>> mechanism for http://esw.w3.org/topic/its0505LimitImpact. It would >>> do this via an external CAM XML definition document that would be >>> applicable to a given DTD/XSD type, and/or individual document >>> instance. This decouples the problem of an ITS tag set for the above >>> topics from embedding and the intendant problems. >>> >> I'm just worried about the mass of problems CAM seems to try to solve >> at the same time. Could you come up with an example, as simple as >> possible? E.g. you have an XHTML document like > > > This may be the main problem with CAM - its implementation is also > predicated on the existence of a CAM processor, which may be the main > mitigating factor against its use. There is a SourceForge site for a > Java implementation of CAM: > http://sourceforge.net/projects/camprocessor but it is still at alpha > stage. Without a standard freely available CAM processor, the > specification is not really usable. > > >> <html xmlns="http://www.w3.org/1999/xhtml" >> xmlns:its="http://www.example.org/its" xml:lang="en" lang="en" >> its:translate="yes"> >> <head><title its:translate="no">World News - 2 May 2005</title> >> <meta http-equiv="Content-Type" content="text/html; charset=utf-8"/> >> </head> >> <body> >> <h1 its:translate="no">World News - 2 May 2005</h1> ... >> <p><its:span lofinfo="this needs special treatment">...</its:span></p> >> </body> >> </html> >> >> and you want to validate it against an DTD for XHTML, and you have a >> XSLT stylesheet which relies on a <p> element without child elements >> from another namespace than "http://www.w3.org/1999/xhtml". How would >> you describe validation and XSLT transformation processes with CAM? >> It would be great to have such an example, so that we also could use >> if for the EXTREME article. > > > Unfortunately my knowledge of CAM is woefully insufficient to be able > to provide an example! > > Best Regards, > > AZ > >> Best, Felix. >> >>> I need to spend a lot more time with CAM before I could say for >>> certain whether it would be a viable solution, but there appear to >>> be some very good things within the CAM specification that are worth >>> investigating further. >>> >>> Best Regards, >>> >>> AZ >>> >>> Felix Sasaki wrote: >>> >>>> >>>> Hi Andrzej, >>>> >>>> I had a look at CAM [1] which you talked about yesterday at the >>>> telecon. It seems to be that it CAM is made for the augmentation of >>>> schemas with additional information. One kind of information can be >>>> used to describe the relations of element and attribute names from >>>> different namespaces to an external knowledge base of names, in the >>>> CAM terminology "nouns". I'm not sure if that solves our problems, >>>> for example to avoid impact of ITS on existing markup schemes and >>>> documents, e.g. problems with XPath etc. What we need is a way to >>>> describe the relations of markup A from namespace to markup from >>>> namespace B, to be able to say s.t. "For this processing step, >>>> IMAGINE that <html:span> is equal to <its:span>." With CAM, this >>>> seems to be possible through the "nouns" knowledge base, and there >>>> seems to be no mechanism to process "IMAGINE". Also it seems that >>>> there is not very much happening in the TC, the spec. was from >>>> 2004, and i didn't found many implemenations / applications. >>>> >>>> I just had a quick look into this, so maybe my impression is wrong >>>> that this doesn't solve our problems. What do you think? Do you >>>> have more material? >>>> >>>> Best, Felix. >>>> >>>> [1] /www.*oasis*-open.org/committees/tc_home.php?wg_abbrev=*cam*/ >>>> >>>> >>>> >>>> >>>> >>> >>> >> >> >> >> >> > >
Received on Thursday, 16 June 2005 09:55:52 UTC