On Sat, Aug 04, 2007 at 10:39:48AM +0200, thomas bonfort wrote: > > http://boston.freemap.in/kamap.html?lat=901231.7146&lon=232335.42917&zoom=8&layers=B
> > http://boston.freemap.in/?lat=901231.7146&lon=232335.42917&zoom=8&layers=B > > > > doesn't demonstrate a fairly significant difference in rendering. > > > you're not being very fair here chris ;) > here's the (nearly) beta2 output from mapserver and agg: > http://tinyurl.com/3692ms* The point of the discussion was that there is significant reason to use AGG over GD -- not that there is a fairly significant reason to use Mapnik over MapServer. I'm sorry that wasn't clear. I'd prefer not to use Mapnik. As wonderful as Artem has been in developing it, it's not got nearly the polish, and the manhours of documentation and work that MapServer have had in the user-facing department. (The fact that I field many questions on how to configure Mapnik on the *TileCache* mailing list is simply one example that the project is relatively immature.) The AGG rendering is a huge step up from GD. The rendering engine, and its capabilities, are what forced me away from MapServer in the past. Now I'm close to the point where I could start using it again. I'm glad that MapServer is moving to using the AGG rendering library as an option, because I think it will lead to better maps for users who need them. That's a great thing. It's obvious that the current AGG work is a huge improvement. However, it's even more obvious to me that to make pretty maps, it was absolutely neccesary -- and that's all I was trying to demonstrate: just GD vs. the best you can do with AGG. (I'm not convinced that MapServer can demonstrate 'the best you can do with AGG' until the fonts are rendered with AGG.) Regards, -- Christopher Schmidt MetaCarta
