[mapserver-users] testing mapserver large files rendering improvements

[Guillaume Sueur]

I'm not sure that "standard" user uses often msDebug().
My point was just to say that if somebody tries and tests MapServer, 
he/she will only read the basic documentation to have a working mapfile 
on a dataset. Once it works, no more reading will be done. And if a 
large shape is tested, the user can then be disappointed.
The thing is that it's more than a specific tuning stuff, maybe more 
than a GIST index on a postGIS database.
I think it should be mentionned at least in the Vector Data Access / 
shapefiles doc of the website. In addition to "Shapefiles are made up of 
a minimum of three similarly named files, with different suffixes..", 
one could write : "having a .qix file made with shptree improves 
considerabily the performances, mostly on large shapefiles"

Beside this, I like the idea of building a MapFile Optimizer too... 
Daniel, what do you have in mind exactly ?


-- 
Guillaume