Justin Deoliveira ha scritto:
> <snip>
  > I think I agree with Gabriel in that adding a derived property which
> just used other properties is one thing, but adding a method that does a 
> full blown resource lookup is another, and something I would like to 
> avoid on model objects themselves.
> 
> As for the case described in GEOS-3049, I think just a better check on 
> startup will fix that, since it involves physically removing jars from 
> geoserver and restarting. I have not been able to reproduce a similar 
> issue by simply brining a database down in between two requests. The 
> behavior in that case seems reasonable, unless i am failing to reproduce 
> the real issue.

+1 on adding better startup checks

A very similar stack trace materializes also when a feature type fails
to compute, see http://jira.codehaus.org/browse/GEOS-3099. It would
seem the only safe approach is really to guard the xml parser
configuration code with enough try/catch to skip each type that
fails to compute, which in the particular case might be patch n1
attached to GEOS-3049

Cheers
Andrea

-- 
Andrea Aime
OpenGeo - http://opengeo.org
Expert service straight from the developers.

------------------------------------------------------------------------------
Crystal Reports - New Free Runtime and 30 Day Trial
Check out the new simplified licensing option that enables unlimited
royalty-free distribution of the report engine for externally facing 
server and web deployment.
http://p.sf.net/sfu/businessobjects
_______________________________________________
Geoserver-devel mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/geoserver-devel

Reply via email to