[Mapbender-users] Mapstorer Integration

Hr. Johannes Loose loose at lwf.uni-muenchen.de
Tue May 20 04:53:08 EDT 2008

Dear Mapbender users and devs, Dear Mapstorer Community:

Should there be done some work to bring mapbender and mapstorer closer together? 
Here are some of my thoughts:

With mapbender we have an open source WebGIS environment that once installed 
and customized it can be maintained even by not-so-advanced users. After some 
training, an end user with some GIS knowledge, but no deeper computer knowledge 
should be able to use the mapbender GUI and keep GIS projects up to date. But 
mapbender offers no possibility to manage the geodata itself, that´s not what it was 
designed for. Assuming that the UMN Mapserver is the most widely used WMS 
Server, it would be great to have an application to manage Mapfiles with the same 
simpleness. I think of changing colors, symbols, labeling or things like that via a 
web-based user interface.

Well we have an application that can do that, the mapstorer! Unfortunately, the `team 
play´ of mapbender and mapstorer ist not as comfortable as it could be. A mapfile, 
managed by mapstorer has to be exported and imported into mapstorer manually. 
Maybe this mechanism could be simplified by an additional mapbender module and 
some slight changes in the mapstorer export function (Like: export to mapbender, 
autoimport). Dealing with two independent applications, we have a different user 
management, thus forcing two logins. So the second desire for the ,getting together´ 
would be a way to combine the user management, if possible.

Mapstorer in the current development stage is a nice and usable application with 
comprehensive functions but it still has some shortcomings. I´d like to state some 
improvement suggestions at this place, knowing well that it is a lot of work in total 
and that it won´t be solvable over night. Interested users and potential developers 
should also have a look in the mapstorer wiki, where future development is indicated:

* Mapstorer has only a basic user management at this stage, this should be improved 
in order to have multiple independend administrator instances , each with their own 
set of mapfiles. Possibly even more detailed user permissions. At the moment I use 
independend databases for each project, what´s a real dirty workaround.
* Mapstorer currently only supports mapserver 4.x Mapfiles, it would be great to 
have 5.0 syntax to be able to use mapbender with up-to-date mapfiles. The ,big 
solution´ would be to make mapstorer ,speak´ SLD, too.
* If one wants to make the mapbender / mapstorer combination usable worldwide, 
mapstorer would need multi language support. At the moment it´s English only, what 
is ,quite international´.
* Currently the mapfile management is ,mapfile orientated´ not ,layer orientated´. 
Making layers independent form mapfiles, would simplify the management of more 
complex projects and reduce redundancies in the database, if a layer is used more 
than once.

That are only some important points, one could think of thousands of potential 
improvements. But thereover we should not forget what we already have:
*A nice gui, that allows to do small changes in the map layout on-the-fly without 
deeper knowledge about mapserver / mapfile internals, including uncomplex creation 
of mapfiles and the handling of metadata. All parts of a mapfile are neatly arranged 
in a tree structure.
* Mapfiles can be stored, exported, combined and edited, all via user interface. That 
reduces greatly the risk of typos, forgotten lines, quotes or brackets, which would 
prevent a mapfile form working. There is no need to download the mapfile, open it in 
an editor, save it and upload it agein.
*A mapfile import function that works quite well now and does logging, too. One 
can use existing mapfiles, or even convert ArcGIS Projects using AMEIN.
* Easy file based data management with web-based upload.
*A plugin system, with ready to use plugins like the postgis plugin (for data 
management) and the db-backup plugin. This system will also make future 
developments more easy.

So, all in all it is already a usable and very helpful system, but it could be so much 
more. Will it pay of for the users to have if some of the stated points would be 
realized? Would it improve OpenSource WebMapping? What is the effort for the 
devs to get some of the stated improvements done, will it be worth the work? What 
points are easy to put in action, what points require a greater deal of energy?

To make it more easy for you to give a short reply to a long post, I will ask again a 
very general question: Should we bring mapbender and mapstorer closer together?

Please vote +1, 0 or -1 and state your view shortly, what made you decide like that.

Best regards


More information about the Mapbender_users mailing list