[mapserver-users] Meeting

[Matt.Wilkie]

Hi David,

Yes Mapserver is free. Yes bugfixes are often available within a
day of being discovered. Yes Mapserver is an amazing package and 
well worth spending the extra agonizing and frustrating hours to 
make it work - something Mike's original post noted:

MN> Thanks for listening. :-) I do love Mapserver. It's something 
MN> I've wanted for a long time. 

However these facts do not negate Mike's observations or the validity of his
complaints. Mapserver -is- a bear to setup. It
-is- frustrating that you cannot display even a simple map "out
of the box". The poor state of the documentation and examples 
-is- worth addressing.

I don't object to what you said, just the tone and target. Yes,
everybody should keep in mind that Mapserver is -free- software, 
and understand that -we- are it's community, and that it is up 
to -everybody- to address any lacks that are encountered along 
the way. The original post was not a whine directed solely at 
the developers, rather an impassioned plea to all who have made 
progress with Mapserver to share their gains:

MN> So, I implore the Mapserver community: Contribute every 
MN> Mapscript and .map sample file you can find and *put it 
MN> in the distribution*. Or at least make another distribution 
MN> (and ms_demo doesn't cut it) that has dozens of examples.

Your fire is appreciated David, I just think there are more
applicable (and toastable) targets. :)

cheers,

-matt

-----------------------------------------------------------------
Matt Wilkie * Yukon Renewable Resources GIS
http://renres.gov.yk.ca/pubs/rrgis/
-----------------------------------------------------------------




> -----Original Message-----
> From: Fawcett, David [mailto:david.fawcett at moea.state.mn.us]
> Sent: 2001 July 25 13:01
> To: mapserver-users at lists.gis.umn.edu; 'Mike Neuman'
> Subject: [mapserver-users] RE: [MapServer-users] Meeting
> 
> 
> 
> Yeah, I can see your point.  20 whole hours to get a 
> powerful, flexible, continually updated, did I mention FREE 
> software program to work.  You could just spring $10,000 for 
> ARC IMS and have your own map up and running in 10 hours.
> 
> The documentation is getting better and better.  The 
> resources that could have been put into documentation in the 
> past, have instead been put into new development, added 
> functionality, and bug-fixes.  The people developing 
> MapServer, Mapscript, PHP Mapscript, etc. have put a lot of 
> unpaid hours into improving the product, and they share their 
> apps and code. 
> 
> Have you noticed that bug-fixes are often committed to the 
> CVS or PHP modules within a day of notification?  Noticed how 
> willing people are willing to help when you post problems on 
> the listserv?
> 
> Yeah, I think I will go out and buy ARC IMS, SDE, and Oracle 
> spatial.  Should be much less painful.
> 
> </sarcasm>
> </rant>
> 
> David.
> > ----------
> > From: 	Mike Neuman[SMTP:mcn at engarde.com]
> > Sent: 	Tuesday, July 24, 2001 6:08 PM
> > To: 	mapserver-users at lists.gis.umn.edu
> > Subject: 	RE: [mapserver-users] Meeting
> > 
> > > MapServer is not just a success, it is a raving success.
> > 
> >   I'm new to the Mapserver community. I would agree with 
> this statement from
> > a *technical* point of view, but strongly disagree from an 
> approchability
> > point of view. There are virtually NO examples of how to 
> create maps, do
> > things in mapscript, etc. It took at least 20 hours of me 
> fighting with
> > Mapscript and a .map file before I could get a relatively 
> simple map to
> > display. Then another 8 hours to get the projection right. 
> If I didn't hate
> > my GIS program so much, I probably would have given it up 
> after the first
> > few hours.
> > 
> >   The documentation has several major errors (queryByPoint 
> still appears on
> > the web site, when queryUsingPoint is correct), PHP 
> Mapscript crashes my
> > webserver when executing simple (and correct, according to 
> the docs) code,
> > and Perl Mapscript coredumps whenever you create a new 
> class in an existing
> > layer or a new layer in an existing map. To get the 
> projection stuff to work
> > as you might expect, I had to upgrade to 3.5. My existing 
> Mapscript code
> > stopped working entirely (with only minor changes made to 
> fix the semantic
> > changes of msSaveImage) by spitting out a blank map, so I 
> had to downgrade
> > to 3.3. Finally, in attempt to debug several things on my 
> own, I downloaded
> > and installed SWIG. Unfortunately, the .i files are no 
> longer compatible
> > with the current version of SWIG. I could not build 
> mapscript from the CVS
> > checkout, and I could not overwrite any of the _wrap files 
> on the nightly
> > builds.
> > 
> >   All of these problems have workarounds, but without 
> examples, it's an
> > *extremely* painful process to figure them out.
> > 
> >   So, I implore the Mapserver community: Contribute every 
> Mapscript and .map
> > sample file you can find and *put it in the distribution*. 
> Or at least make
> > another distribution (and ms_demo doesn't cut it) that has dozens of
> > examples. The Mapscript examples IN the mapscript distro 
> are worthless--they
> > don't even draw a map!
> > 
> >   Thanks for listening. :-) I do love Mapserver. It's 
> something I've wanted
> > for a long time. I can't imagine how fast Mapserver's use 
> would spread if it
> > were approachable. That's all I ask.
> > 
> > -Mike
> > 
>