[Mapserver-users] (no subject)

[Kevin Webb]

Dear Mapserver folks,

I’m currently working on a rather larger Mapserver application and have run
into a couple of problems/potential features that I’d like to share and get
your thoughts on.

1)  With PostGIS (or SDE, etc.) connections what is potential of moving the
FILTER attribute to the CLASS – essentially replacing current role of the
EXPRESSION attribute? It seems that given the database server’s power at
indexing and filtering this would be significantly faster than passing all
the data back to Mapserver to have it again reprocess the data in order to
generate different classes. Is this a correct assumption or are there
problems with doing multiple queries to retrieve each class (e.g. is the
query overhead greater than simply doing the filtering inside mapserver)?

I ask because I’m thinking about implementing something similar to ESRI’s
unique and gradient colors features (for example allowing users to select
groups of objects with each layer and apply colors based on attributes – or
quantize attributes and apply a gradient). This would mean lots of
sub-classes in each layer and lots of filtering going on to organize these
classes.  It would be great if I could just pass in a different WHERE clause
for each class.

Does this make sense or are there better ways to accomplish this?


2) This is more of a PostGIS question (which I’m posting on their list as
well), but it has implications for Mapserver as well, so I thought I’d share
it here. What is the current opinion of implementing some sort of line
simplification algorithm (like Douglas-Peucker) on data before it’s sent to
Mapserver. For example, if Mapserver could communicate the current scale
(units/pixel) to the DB server, then it could strip out extraneous
information prior to transmission. This could significantly reduce the
processing time required to parse complex layers at full extent.

I know there’s been some talk about this amongst the PostGIS folks in the
past, but I don’t have a clear picture of where this stands today, or if any
Mapserver users would even find this sort of feature useful.  I think I’ve
come up with a way to implement it, but I’d like to hear input from others
before go too far.

3) Is there any interest or discussion of implementing an opacity setting
for layers (similar to ESRI’s) such that a polygon might be 50% or 75%
opaque, for example? I don’t know much about the implications for this, but
my understanding is that GD already supports this. Correct?

With that said – I think I may be in a position to try implementing theses
features sometime soon. I’d love to hear input from others and I’m going to
start playing with the source on my own soon.

Thanks!
Kevin Webb

--
e: kpw[at]jump9.com
p: (919)260-8375

w: kpw.jump9.com