gump-general mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Nicola Ken Barozzi <nicola...@apache.org>
Subject [PATCH][gump.py] Going forward
Date Tue, 22 Apr 2003 09:27:04 GMT

Attached is an enhanced gumpview that also lists project build list 
(there is a bug that does not clear something on subsequent calls, 
didn't track it own), a mrege.py that dumps the workspace (redirect to a 
file if needed; it chokes somewhere on a descriptor characther, dunno0 
why), renamed sort to dependencies, and cleaned up some stuff.

I'm now in the process of making it useful for me, and that means making 
it build (surprise surprise!).

The point is that I want to start from a every-day-use scenario where 
users *don't* want to download half a RAID just to build a project. That 
means using Ruper to get dependencies.

Here are different build scenarios:

1 - build projects in same module and download others' jars
2 - build projects I have already from CVS and download others' jars
3 - build projects with CVS images of only the directly dependant
     projects and download others' jars
4 - use all from CVS

I need scenario 1 first, and then 2. Usual Gump usecase is 4. 3 can be 
avoided IMHO.

I would do separate scripts for these scenarios (renaming gump.py to 
engine.py or somethig like that):

1 - build.py
2 - xbuild.py
3 - X
4 - gump.py

After, I would also like to make Gump use the local descriptors instead 
of the cached ones if present on disk.

-- 
Nicola Ken Barozzi                   nicolaken@apache.org
             - verba volant, scripta manent -
    (discussions get forgotten, just code remains)
---------------------------------------------------------------------

Mime
View raw message