[GRASS5] The status of 5.0

[Markus Neteler]

On Thu, Mar 21, 2002 at 04:49:56PM +0100, Bernhard Reiter wrote:
> On Thu, Mar 21, 2002 at 11:40:40AM -0500, Helena Mitasova wrote:
> > So I would just encourage everybody to get
> > grass5.1 to get some feel for it and maybe it would be good time
> > to start moving all GRASS to GRASS5.1 
> 
> Several attempts to make big jumps with GRASS have shown that this
> might not be possible.

Please note that you can link the GRASS 5.0 raster commands into
the 5.1 tree. I have already written a small script for this, but
not yet uploaded. as the GRASS 5.1 writes into a new vector subdirectory
structure, it can be happily used on "old" locations.

Before filling GRASS 5.1 with unchecked code, I definitly recommend
to go this "soft" way of linking into the new code structure and
module-by-module migrating the code with code cleanup.

We need to do heavy cleanup on the GRASS libraries first before
looking at the modules (all related to GRASS 5.1).
 
> So we need to wrap up and release GRASS5.
> I've outlined a good plan and procedure how to get there a while ago.
> 
> Still we need the complete merge of the two branches and a new
> release branch which radically leaves out modules 
> with release critical bugs.
> So the new release branch would not have any module with release
> critical bug and then gets released.

Means: Please make HEAVY use of the bug tracking tool and REPORT
errors. From that we decide for the module candidates of 5.0, I
think.

Bernhard: Can we add something to allow the bug reporting person
to set the "priority" while *submitting* the bug? To have all
on level "30" doesn't make sense if the reporting person already
knows that it is critical (or not). Or do I just have to add something
in the report form?
 
> On the merge:
> Glynn also posted a couple analysis of the differences of the 
> two branches in GRASS5. Glynn: Maybe you can propose how to completly
> discontinue the old release branch. You might also just do it if
> nobody else objects.

Well, that's not so easy. Currently the exp and release are fairly
up-to-date (Glynn and me were taking care for that). There are
differences which may remain (new features which don't go
for 5.0 release).
I don't think that we all should hold our breath until 5.0stable is
out and stop submitting new features - that would be contra-productive.

But I agree that we still have some work for the vector engine in 5.0.
As it is currently implemented, we
 - either have to radically exclude vector modules from the release branch
 - fix at least the import tools and v.support.

David, when do you expect the have the v.in.shape ready? This module
is somewhat crucidal, that's what many users expect from GRASS, to 
read this [...] SHAPE format.

> On the bugtracker: 
> There are many bugs in there. We need people to
> evalutate them, reproduce them, fix or delay them.
Please find these people, especially for fixing... :-)

Markus