RE: [mapserver-users] external links in a map

[Ian Turton <ian@geography.leeds.ac.uk>]

At 19:56 04/09/02, Ed McNierney wrote:
> Ian -
>
> While I don't disagree with your general approach, I think this is a very 
> reasonable feature for MapServer, and I've been looking at it a bit 
> myself.  If it's reasonable to let MapServer generate point symbols or 
> polygons on a map image, and it's reasonable for MapServer to have a 
> "query" mode in which you click on those things to produce querymaps of 
> that feature, then I think it's equally reasonable for that server to 
> generate flyover labels, etc. for the same feature.  It happens that HTML 
> requires a completely different set of tags (imagemap) to do such a thing, 
> but that's not MapServer's fault nor is it a good reason to not want 
> MapServer to generate them.
>
> What I do not particularly want is two SEPARATE tools traversing the same 
> data set to generate two related outputs unless there's a deterministic 
> way to ensure they are consistent.  If GeoTools and MapServer don't render 
> a polygon feature exactly the same way, my imagemap won't agree with my 
> image.  That may or may not be important to my application.
I fact I was proposing using MapServer should generate the image 
(preferably in WMS mode) while the client handles the pop ups or hovers 
based on info from the server (possibly acting as a Web Feature Server 
WFS). Certainly for hover/flyover labels you'd want to avoid the over head 
of a HTTP call each time, it's hard to optimize that stuff in a single 
program with out latency in too.

> GeoTools is CERTAINLY worth investigating for this particular problem, but 
> do think that it's a reasonable thing to want MapServer to do.  Another 
> good tool philosophy is that if you've got one tool that already knows how 
> to do 90% of the job, it might be helpful to add the other 10%.  I think 
> they're both useful ways of looking at a problem.
In fact someone else suggested how to do some of it in existing mapscript 
so I'm off to read up on that as I'd previously overlooked templates in 
that regard.


>         - Ed
>
> Ed McNierney


Ian
Ian Turton, Centre for Computational Geography, School of Geography,
University of Leeds, Leeds, LS2 9JT. ***0113 3433392 - New Phone No***
URL: http://www.geog.leeds.ac.uk/staff/i.turton/