pre rfc draft on rendering

[Stephen Woodbridge]

thomas bonfort wrote:
> hi all,
> 
>>>  - with brushed lines, I'm that brushed-based (as opposed to
> marker-based) lines could be done away with.
> 
> another corrolary question we might like to address is if we want to
> keep the ANTIALIAS keyword and processing with the gd renderer. In my
> opinion it could go away, as it would greatly simplify the
> implementation of the gd rendering backend, and produces rather low
> quality results.
> 
> for those of you knowledgeable of vtable internals, I'd appreciate it
> if you could have a look and comment on the paragraph I added at
> http://trac.osgeo.org/mapserver/wiki/UnifyingTheRenderingInterfaces#ProposedVTableApproach
> .
> 
> one last question: are any of you reluctant or against going this way?
> This does imply some rather substantial changes, so there is a rather
> certain risk of instabilities / bugs inherent to these.

Thomas,

I think code cleanup and simplification are valuable goals to pursue. 
Depending on the amount of instability and the time of the release I 
would be ok with this.

The other concern is the changes to the mapfile. I personal think that 
we have a lot of rendering mapfile options that do not always play well 
with one another and that this adds to the confusion of users. When 
should I use option x? I used x, how come this does not work this way? 
etc. So anything that would rationalize and simplify the definition of 
classes and styles so that they work in a predictable manner regardless 
of the renderer being used is good. We also made a statement that we 
were planning to make additional mapfile changes post 5.o release.

It seems like you are moving in this direction. The key will be that we 
get these changes in early so that there is time for people to work with 
  mapserver SVN to help identify issues and make things stable again 
before we are looking to release.

-Steve W