[Mapbender-dev] How do you spell mapbender or The S in PSC

[Thomas Baschetti]

Hi,

after reading the log of the last MB IRC http://logs.qgis.org/mapbender/%23mapbender.2007-07-23.log
(a little bit slow at the moment)
and this http://www.mapbender.org/index.php/Category_talk:PSC
it seems someone is not so happy at the moment.

Ok, let's change this

First:
i think being active is an important point. No, i'm not active enough but willing to change.
So i first propose to add Arnulfs comments to the procedures for Mapbender PSC.

Second:
i think testing of Mapbender needs some improvement. Trac is ok, but very technical, 
so i don't think many USERS will use this.
Perhaps we should set up some procedures to follow with each RC and every Release:

1. Test database-dumps
  => mysql to new database
  => mysql upgrade from previous version
  => postgresql to new db
  => postgresql upgrade from new version
  
  of course with uft8 and iso...
  Are there any errors? Pitfalls? Is it possible to build
  an mapbender-Installation from scratch with this release?
  
  Is it necessary to test under linux and windows? I don't know

After testing give an ok in the list or (better) mapbender Wiki for this release.

[I'm willing to do this point on a regular basis.]

2. test (and document) new features

i think new features are an important reason for new releases ;-) 
So they have to be tested thoroughly and they have to be documented.
Both test and document from an enduser-point-of-view, the programmer has already tested
this,i hope ;-).

3. test (and document) changed features
  i.e. WFS-Gazetteer
Both test and document from an enduser..(see above)

where to document? Perhaps wiki, at the changelog-pages?

4. document deprecated features
Document which feature is deprecated, which is superceeded by which. 
(wfs_gazetteer=>wfs_gazetteer_client, mod_key.php...)

btw. who decides which feature is deprecated?


I don't think testing should be done by the main programmers, they program
(which is time consuming enough), they know how to use the software, 
they know which adjustment is expected by which feature so many bugs will
not become visible.

Third:
Remind PSC-Members, users, programmers of MB-IRC on a regular basis, perhaps a few hours before meeting starts. I must admit i just forget
the meeting quite often...


Comments and contradiction are welcome...
What do you think?


Bye
Thomas


-- 
Ist Ihr Browser Vista-kompatibel? Jetzt die neuesten 
Browser-Versionen downloaden: http://www.gmx.net/de/go/browser