> Biggest pain IMHO is unpredictability of query times. It turns out that most
> queries are rather fast even on big sites, while some are slow without
> apparent reasons. If we could just limit SQL-side query execution and report
> timeout errors if needed, things would be much simpler. But I do not know how
> to achieve that.
Using ID's for queries would probably simplify both caching and also can be
used to collect statistical information about query execution times. In such
case further queries can be "throttled" ("please wait, site is too busy to
process your query" message) based on statistical information stored in the
queries table.
Also, ID's can be _optionally_ emabled to prevent anonymous from using
Special:Ask.
Just a thought. You may disagree, of course.
Dmitriy
-------------------------------------------------------------------------
This SF.net email is sponsored by: Microsoft
Defy all challenges. Microsoft(R) Visual Studio 2005.
http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/
_______________________________________________
Semediawiki-devel mailing list
Semediawiki-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/semediawiki-devel
