Took a look over the patch and it looks good to me. Andrea Aime wrote: > Hi, > following up a discussion I had with Arne in Sydney today I > checked how much the compression level settings in PNG affected > the outcome of the FOSS4G benchmark and confirmed the difference > was significant (10-20% on the labelling test). > > In particular on my machine I measured up to 15.2 req/s with > the current compression level, and up to 16.4 res/s with a more > relaxed value that, to my surprise, did not really change much > the size of the output. > > Pity I did not pay attention before the shootout, I was distracted > by unrelated work unfortunately. And yes, next year I'll try to > involve more people into the effort, and earlier, apologies for > mis-managing it. > > Anyways, I've attached to http://jira.codehaus.org/browse/GEOS-3599 > a patch that changes the default value of the compression > to a more relaxed value and also allows the admin to tweak > that value. > > The compression is between 0 and 100, the defaults are 25 > now. Looking at it it seems values between 0 and 20 show > huge changes in the output side, past that the change in > output size is small but the change in effort big. > It is something I'll have to document. > > One thing that left me wondering is where and how set the > defaults. I did so in the WMS facade and in the UI, wondering > if there is a better and central place to enforce defaults thought. > I was thinking persistence (while loading the WMS metadata > map) but I'm not sure where to put my hands. Suggestions? > > Cheers > Andrea >
-- Justin Deoliveira OpenGeo - http://opengeo.org Enterprise support for open source geospatial. ------------------------------------------------------------------------------ Let Crystal Reports handle the reporting - Free Crystal Reports 2008 30-Day trial. Simplify your report design, integration and deployment - and focus on what you do best, core application coding. Discover what's new with Crystal Reports now. http://p.sf.net/sfu/bobj-july _______________________________________________ Geoserver-devel mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/geoserver-devel
