[GRASS5] Re: GRASS 5.3 release

GRASS GRASS developer
1

I agreed to get the scripts g.parser compliant, but of course will welcome any help offered. A month back, I went through and updated as many 5.3 scripts as I could get to work to 5.7 and submitted them to Markus. This is about 90% of the work needed. They need to be tested by others too of course, but should work pretty much as is with 5.3.

The biggest change will be in scripts that switched from unnamed command-line arguments (e.g., script [argument 1$] [argument 2@] ... ) to the 'option=argument' format (e.g., script [option1=value1] [option2=value2]...). What needs to be done is 1) change the tcltk menu modules to properly use all new option=value arguments and 2) change the man pages to also indicate the proper command-line use. This is not an insignificant number of scripts, but it is not the majority either. There are also some scripts that don't seem to work in 5.7 and may not work in 5.3 either. We should probably either fix or deprecate them.

I was going to call the updated version of tcltkgrass 4.1 to indicate these changes. I doubt if I can get it done by the time you all want to release 5.3.0, and I DON'T want to hold up this release in any way. Can I do this to get into a 5.3.x or 5.4 release? How should I number tcltkgrass? Does it matter?

Michael Barton
______________________________
Michael Barton, Professor & Curator
School of Human Origins, Cultures, & Societies
Arizona State University
Tempe, AZ 85287-2402
USA

voice: 480-965-6262; fax: 480-965-7671
www: http://www.public.asu.edu/~cmbarton
On May 7, 2004, at 2:27 AM, grass5-request@grass.itc.it wrote:

there are alternatives like r.in.tiff

all scripts -> g.parser ?
missing man pages (Debian bug #229086, 242436) no
Remove any remaining GD,GDBM mentions ?
init.sh/set_data race ??? not sure what this was about
NVIZ updates done
add r.series, r.grow2, d.info **TODO**
add r.[in|out].mat, r.univar(2) **TODO**
OSX TclTk -> Aqua so close
TrueColor for core display modules C done; man pages?
shared libs working on all but Cygwin ?

2

Hello Michael

On Fri, 7 May 2004, Michael Barton wrote:

The biggest change will be in scripts that switched from unnamed
command-line arguments (e.g., script [argument 1$] [argument 2@] ... )
to the 'option=argument' format (e.g., script [option1=value1]
[option2=value2]...). What needs to be done is 1) change the tcltk menu
modules to properly use all new option=value arguments and 2) change
the man pages to also indicate the proper command-line use.

If as I understand it this means custom scripts etc. in existing installations
will not work due to the new syntax, IMHO I don't think this change should
be made so close to the release of 5.3 and should be left to 5.7.

Paul

3

If you mean the scripts that are in the normal GRASS distribution, a week ago, I would have said yes, this is a problem. However, I was wrong about this. AFAICT, all scripts in 5.7 should work with 5.3--except those that call special 5.7 vector features. Currently, these are very few. The only thing that needs to be changed is how some (a miniority) are referenced in the tcltkgrass GUI and the man pages for the same scripts. Markus suggested that it would help 5.7 development it all essential scripts shared by 5.3 and 5.7 be located in the 5.7 CVS branch and represented by links in the 5.3 branch (I hope I have the terminology +/- correct here). I could try to get some idea of how many scripts need tcltkgrass updates to use the 5.7 version over the next week if that is helpful.

For other scripts--i.e., those that individual GRASS users have developed but are not part of the regular release--some that may work in 5.3 may not work in 5.7 without some work but usually not a lot of work. This is not an issue for the release. IMHO, it would be a nice idea to have a script-writing tips page on the GRASS www site to lay out a set of basic GRASS script standards--e.g., bash scripting, use of g.parser headers, etc.--to get all scripts to a 5.7/5.3 compatible format. Perhaps I could try my hand at that this summer, along with the other related things.

Finally, there are some 5.3 scripts that don't yet work in 5.7 because they call GRASS modules that don't (yet) exist in 5.7. A good example is r.fill.nulls. It requires s.surf.rst. Because of the demise of the GRASS sites format, s.surf.rst does not exist in 5.7. And, unless something has been done in the last couple weeks while I've been focused on the end of semester chaos, there has yet to be a program using the vector-format points to take its place. Somehow, there needs to be a way to sort through these and say which are worth updating to 5.7 and which are not. Again, however, this could be treated as a 'bug fix' later and shouldn't have any effect on the 5.3 release.

Michael
____________________
C. Michael Barton, Professor
School of Human Origins, Cultures, & Societies
PO Box 872402
Arizona State University
Tempe, AZ 85287-2402
USA

Phone: 480-965-6262
Fax: 480-965-7671
www: <www.public.asu.edu/~cmbarton>
On May 8, 2004, at 9:04 AM, Paul Kelly wrote:

If as I understand it this means custom scripts etc. in existing installations
will not work due to the new syntax, IMHO I don't think this change should
be made so close to the release of 5.3 and should be left to 5.7.

Paul

4

On Sat, May 08, 2004 at 09:43:48AM -0700, Michael Barton wrote:

If you mean the scripts that are in the normal GRASS distribution, a
week ago, I would have said yes, this is a problem. However, I was
wrong about this. AFAICT, all scripts in 5.7 should work with
5.3--except those that call special 5.7 vector features.

Right.

Currently,
these are very few. The only thing that needs to be changed is how some
(a miniority) are referenced in the tcltkgrass GUI and the man pages
for the same scripts. Markus suggested that it would help 5.7
development it all essential scripts shared by 5.3 and 5.7 be located
in the 5.7 CVS branch and represented by links in the 5.3 branch (I
hope I have the terminology +/- correct here).

No, I suggested the opposite (maybe I was not clear in that
personal mail). With one limitation:
But I also suggested to stop 5.3 development and to go for 5.7.

I could try to get some
idea of how many scripts need tcltkgrass updates to use the 5.7 version
over the next week if that is helpful.

For other scripts--i.e., those that individual GRASS users have
developed but are not part of the regular release--some that may work
in 5.3 may not work in 5.7 without some work but usually not a lot of
work. This is not an issue for the release. IMHO, it would be a nice
idea to have a script-writing tips page on the GRASS www site to lay
out a set of basic GRASS script standards--e.g., bash scripting, use of
g.parser headers, etc.--to get all scripts to a 5.7/5.3 compatible
format. Perhaps I could try my hand at that this summer, along with the
other related things.

Yes, a sort of automated script repository in the web might be quite
useful.

Finally, there are some 5.3 scripts that don't yet work in 5.7 because
they call GRASS modules that don't (yet) exist in 5.7. A good example
is r.fill.nulls. It requires s.surf.rst. Because of the demise of the
GRASS sites format, s.surf.rst does not exist in 5.7.

It's in fact much simpler:
Once we have fixed the RST LatLong bug (the next days, already
identified), I'll submit r.fill.nulls in 5.7 with the small change
that s.surf.rst is named v.surf.rst in 5.7.

And, unless
something has been done in the last couple weeks while I've been
focused on the end of semester chaos, there has yet to be a program
using the vector-format points to take its place. Somehow, there needs
to be a way to sort through these and say which are worth updating to
5.7 and which are not. Again, however, this could be treated as a 'bug
fix' later and shouldn't have any effect on the 5.3 release.

Michael

Markus

5

On Fri, May 07, 2004 at 09:17:34AM -0700, Michael Barton wrote:

I agreed to get the scripts g.parser compliant, but of course will
welcome any help offered. A month back, I went through and updated as
many 5.3 scripts as I could get to work to 5.7 and submitted them to
Markus. This is about 90% of the work needed. They need to be tested by
others too of course, but should work pretty much as is with 5.3.

Michael,

could you store the updated scripts onto your web site
for others to test?

Markus