[mapserver-users] Antialiased Lines in mapserver?

[Joe Bussell]

It may be the case that there are enough of us who need a slim,
light-weight GD that is capable of scaling its footprint and selection
of algorithm based on the users needs.  This may require a seperate
development branch specific to our common needs.  Steve's suggestion
regarding updating GD to perform antialiasing adds weight to this choice.

What is the level of cooperation between the developers of GD and the
open GIS community?  Are the folks at Boutell.Com responsive to
development feedback?  Has anyone considered basing a new development
effort to create a light-weight, thread-safe image redering library
based on the GD core which is specifically targeted to web apps?

In any case I am for sacrificing alternative rendering formats and
wiz-bang glitz for raw speed.  I need to generate relatively useful
results for a large data set fast.  Speed is everything.  I would like
my preprocess which creates a set of rasters of the static data to make
use of the best rendering possible.  In that case speed is not as
critical.  For live content the situation is vastly different. 

Cordially,

Joe Bussell
On Time Systems
www.TrafficDodger.com


On Mon, Oct 07, 2002 at 07:44:15PM -0400, woodbri at swoodbridge.com wrote:
> On 7 Oct 2002 at 14:59, Steve Lime wrote:
> 
> > I'm sorta hoping that GD 2.0 will offer anti-aliasing. It's not that
> > hard and there are numerous algorithms out there. Way back I actually
> > looked into using ImageMagick instead of GD. However, the author of IM
> > warned that in tests GD was approximately 6 times faster than IM.
> > That's a hefty price to pay for smooth lines.
> 
> I think that this should be evaluated again in light of the fact that 
> GD 2.0 will be using deeper image depth that will impact performance.
> 
> Which raises another question in my mind. 
> 
> Would it make sense to or is it planned that the output routines will 
> be installable or selectable at runtime (maybe via the map file) so 
> for applications that need the performance but can live with GD 1.x 
> image quality can get it, while other apps can get the features they 
> need?
> 
> Shouldn't it be possible to have output drivers for GD 1.x, GD 2.x, 
> ImageMagick, Postscript, Flash, etc. Assuming that someone implements 
> each of them.
> 
> -Steve W.
> 
> > In my opinion, it might be easier to update GD to do antialiasing than
> > to use IM. The new arch is not well documented yet since a production
> > version has not even been released yet. Basically it would consist of
> > functions to draw shapeObj's a points, lines and polygons given
> > MapServers symbolization scheme.
> > 
> > Steve
> > 
> > Stephen Lime
> > Data & Applications Manager
> > 
> > Minnesota DNR
> > 500 Lafayette Road
> > St. Paul, MN 55155
> > 651-297-2937
> > 
> > >>> David Monterroso Cabello <pyro-x at ixine.com> 10/03/02 09:52AM >>>
> > Daniel Morissette wrote:
> > 
> > >David Monterroso Cabello wrote:
> > >  
> > >
> > >>    And finally,  a question for developers, would it be possible to
> > >>switch to ImageMagick for vector rendering?, that library render
> > fine
> > >>antialiased lines  =), althought i dunno if it support everything
> > >>mapserver needs.
> > >>
> > >>    
> > >>
> > >
> > >I had a quick look at the ImageMagick website and saw that they
> > support
> > >a good set of vector rendering functions, so adding ImageMagick as a
> > new
> > >(optional) output system would be possible in the new architecture
> > >developed in version 3.7.  As usual, we just need someone with time
> > to
> > >do this, or money to fund this work.
> > >  
> > >
> > 
> >     Yes, i thought the same when i saw IM web, i dunno if it is very
> > difficult but if finally there is no other way to do antialiasing  and
> > 
> > you think is posible i would like to try to implement it.  Do you know
> > 
> > about any doc or something i can read to understand the new output
> > architecture?.
> > 
> > Many thx
> > 
> > 
> > 
> 
> 
> 

--