Well, I'm not really involved in the core development of SMW, so it's
not up to me. But I'd say that, though is a very interesting idea, I'd
argue against turning SMW into an extensible solution. I think one of
the big strengths of SMW is that it's a relatively simple concept -
there's only one way to declare a relation or attribute, and there's
only one place (that is, one page) where you can store any specific
relation or attribute. This ties in to the discussion about arbitrary
triples we were having before - I think it's very nice that, when you
see a semantic relationship on any SMW site, you know exactly where to
go if you want to edit it.

For a similar reason, I'm against allowing arbitary markup schemes -
it would mean people would lose the ability to have consistent editing
across all SMW sites.

In general, I prefer simplicity over flexibility, everything else
being equal. I'm thinking of various applications and languages that
have gotten more bloated and harder-to-use as they've been given more
features and more ways to do any one thing.

My 2 cents,
-Yaron



On 5/9/07, Ittay Dror <[EMAIL PROTECTED]> wrote:
>
>
>
> Hi,
>
>
>
>
> With all the discussion about different syntax, capabilities and views,
> wouldn't it be nice to make semantic mediawiki extensible? (in the same way
> as protege)
>
>
>
>
> What I would suggest (sorry if this is too brief, if hard to understand (and
> you wish to), i can elaborate):
>
> - an independent library for storing (and maybe querying) semantic data -
> best to find something already implemented (in this case, it probably need
> to have pluggable storages, so semantic mediawiki can use the wiki's
> database). this should just provide the API. this is in fact, the basic
> core. other extensions use it.
>
>
> - extension #1 - a parser that extracts all texts that contain semantic
> data. extensions to the parser can do the actual parsing of the data (using
> different syntax) and return the triplets which are then saved. the parser
> then hands off the rendering of the text to another extension. this
> extension is just a nice-to-have, for someone that does want to have a
> different syntax, but doesn't want to learn the ugliness of mediawiki.
>
> - extension #2 - the default relation syntax parser. best if this can be
> extendable also. this can be done by parsing the relation, creating an
> expression tree, possibly with unknown tokens, and then letting extensions
> take over. e.g., using xml.
>
>
> - extension #3 - queries. a base extension that receives the query, passes
> it onto another extension to handle it and, either returns the triplets
> matching the query, or, return some uniform query language to return a
> standard query which the extension then executes to get triplets. the
> extension then proceeds to render the results, again using extensions (e.g.,
> provide a graph rendering of the relations).
>
> - basic rendering extensions - table, list, timeline.
>
> - data type extensions - these are available already, but just make it
> standard interface with the extensions above.
>
> - hooks for different events - relation creation, update, delete. before
> store, after store, before retrieve, after retrieve, etc.
>
>
> - incorporate the extension to design forms that fill in the semantic data
> (posted not long ago)- i really like this as usually relations mean
> structured data, which is more easily handled by a form.
>
> - factbox extension - just to hide the ugliness of mediawiki and allow other
> extensions to actually render the factbox (maybe stack them)
>
> - allow extensions to directly create triplets. this may mean that the
> 'origin' of the triplet may not be a page. this allows inference engines,
> scripting, import relations, etc.
>
>
>
>
>
> please tell me what you think,
>
>
> ittay
>
>
>
>
> --
>  Ittay Dror
>  Chief Architect,
>  R&D, Qlusters Inc.
>  Web: qlusters.com
>  Email: [EMAIL PROTECTED]
>  Phone: +972-3-6081994
>
>  openQRM - Data Center Provisioning
> -------------------------------------------------------------------------
> This SF.net email is sponsored by DB2 Express
> Download DB2 Express C - the FREE version of DB2 express and take
> control of your XML. No limits. Just data. Click to get it now.
> http://sourceforge.net/powerbar/db2/
> _______________________________________________
> Semediawiki-user mailing list
> Semediawiki-user@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/semediawiki-user
>
>

-------------------------------------------------------------------------
This SF.net email is sponsored by DB2 Express
Download DB2 Express C - the FREE version of DB2 express and take
control of your XML. No limits. Just data. Click to get it now.
http://sourceforge.net/powerbar/db2/
_______________________________________________
Semediawiki-user mailing list
Semediawiki-user@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/semediawiki-user

Reply via email to