Hi,
I think, that there is already enough new functionality in 5.7
for wich users are waiting a long time. It doesn't make much sense
to postpone 5.7.0 for ever, only because of hazy hope that much
more could be done. I want to move forward and release 5.7.0.
It should be something like 5.0 around 1998, that means something
what can be used, but far from stable release.
Certainly >= 5.7.0 will never contain all the functionality from 5.0.0.
To make 5.7 ready for 5.7.0 release, I want to do this:
- category index for vectors (to minimise later format changes)
- disable direct Shapefile and PostGIS support (currently not finished)
- update some missing modules (not all); I have updated status list of
vector/sites modules (grass51/doc/vector/v.modules.html), my intention
is to update those with N in third column, however I am not sure about
statistical modules for sites (s.normal, s.qcount, s.univar, s.windavg,
s.medp, s.probplt,...), are these modules necessary for 5.7.0?
You can look at the list (especially N/X/?) and send us your comments.
This doesn't mean that features like PostGIS/Shapefile/OGR cannot be
added during > 5.7.0.
Radim
On Wed, Oct 22, 2003 at 11:10:17AM +0200, Radim Blazek wrote:
I think, that there is already enough new functionality in 5.7
for wich users are waiting a long time.
I agree that an experimental release is fine.
The odd minor version number will tell everybody that it is an
experimental release. A release in that 5.7. line certainly
will be a first step towards a stable release then to be called 5.8.x.
Certainly >= 5.7.0 will never contain all the functionality from 5.0.0.
Well it is an experimental development line,
if we clearly say that in the annoucements,
people have no right to be dissappointed.
To make 5.7 ready for 5.7.0 release, I want to do this:
- category index for vectors (to minimise later format changes)
- disable direct Shapefile and PostGIS support (currently not finished)
- update some missing modules (not all); I have updated status list of
vector/sites modules (grass51/doc/vector/v.modules.html), my intention
is to update those with N in third column, however I am not sure about
statistical modules for sites (s.normal, s.qcount, s.univar, s.windavg,
s.medp, s.probplt,...), are these modules necessary for 5.7.0?
You can look at the list (especially N/X/?) and send us your comments.
IMO I don't think src.garden/grass.java/GRASS_libvect.c is worth much effort.
m.in.e00 seems to be worth keeping at a later stage if the effort
of porting is less then adding this to OGR. 
On Wed, Oct 22, 2003 at 11:32:17AM +0200, Bernhard Reiter wrote:
On Wed, Oct 22, 2003 at 11:10:17AM +0200, Radim Blazek wrote:
> I think, that there is already enough new functionality in 5.7
> for wich users are waiting a long time.
I agree that an experimental release is fine.
[...]
> To make 5.7 ready for 5.7.0 release, I want to do this:
Find this table also here:
http://freegis.org/cgi-bin/viewcvs.cgi/~checkout~/grass51/doc/vector/v.modules.html
[...]
m.in.e00 seems to be worth keeping at a later stage if the effort
of porting is less then adding this to OGR.
Compare:
http://bugzilla.remotesensing.org/show_bug.cgi?id=159
Maybe a volunteer will contribute this to OGR one day...
Most of the code seems to be already there.
Markus