[GRASS-dev] [bug #5094] (grass) r.support from the GUI is broken

this bug's URL: http://intevation.de/rt/webrt?serial_num=5094

If you launch r.support (r.support.sh actually) from the gis.m menu,
you get to pick a map and then launch r.support in an xterm. As soon
as you enter "y" or "n" at the prompt the xterm disappears. Replace
"execute r.support.sh" with "term r.support" on line ~421 of menu.tcl
for comparison with what should happen.

After some debugging:

if you answer "Y", it breaks in r.support/front/run.c run_etc_support():
    G_snprintf(buf, sizeof(buf), "%s/etc/support/%s '%s'",
             G_gisbase(), pgm, rast);

if you answer "N", it breaks in r.support/front/check.c check_stats()
    G_snprintf(question, sizeof(question), _("Update the statistics "
             "(histogram, range) for [%s]? "), name);

??

For now I'm setting this to "term" mode in menu.tcl in the 6.2 release
branch while we work out a fix in HEAD.

Hamish

On Thu, 2006-09-28 at 00:38 +1200, Hamish wrote:

> this bug's URL: http://intevation.de/rt/webrt?serial_num=5094

> If you launch r.support (r.support.sh actually) from the gis.m menu,
> you get to pick a map and then launch r.support in an xterm. As soon
> as you enter "y" or "n" at the prompt the xterm disappears. Replace
> "execute r.support.sh" with "term r.support" on line ~421 of menu.tcl
> for comparison with what should happen.

After some debugging:

if you answer "Y", it breaks in r.support/front/run.c run_etc_support():
    G_snprintf(buf, sizeof(buf), "%s/etc/support/%s '%s'",
             G_gisbase(), pgm, rast);

if you answer "N", it breaks in r.support/front/check.c check_stats()
    G_snprintf(question, sizeof(question), _("Update the statistics "
             "(histogram, range) for [%s]? "), name);

??

For now I'm setting this to "term" mode in menu.tcl in the 6.2 release
branch while we work out a fix in HEAD.

I'm about half way through rewriting r.support (been in that state for
awhile, now). I'm splitting up the functionality into individual
modules and removing interactivity. A few of the modules are done, but
I'm waiting for them all to be done before committing to CVS.

--
Brad Douglas <rez touchofmadness com> KB8UYR
Address: 37.493,-121.924 / WGS84 National Map Corps #TNMC-3785

Brad Douglas wrote:

> > If you launch r.support (r.support.sh actually) from the gis.m menu,
> > you get to pick a map and then launch r.support in an xterm. As soon
> > as you enter "y" or "n" at the prompt the xterm disappears. Replace
> > "execute r.support.sh" with "term r.support" on line ~421 of menu.tcl
> > for comparison with what should happen.
>
>
> After some debugging:
>
> if you answer "Y", it breaks in r.support/front/run.c run_etc_support():
> G_snprintf(buf, sizeof(buf), "%s/etc/support/%s '%s'",
> G_gisbase(), pgm, rast);
>
>
> if you answer "N", it breaks in r.support/front/check.c check_stats()
> G_snprintf(question, sizeof(question), _("Update the statistics "
> "(histogram, range) for [%s]? "), name);
>
>
> ??
>
>
> For now I'm setting this to "term" mode in menu.tcl in the 6.2 release
> branch while we work out a fix in HEAD.

I'm about half way through rewriting r.support (been in that state for
awhile, now). I'm splitting up the functionality into individual
modules and removing interactivity. A few of the modules are done, but
I'm waiting for them all to be done before committing to CVS.

I suggest committing them as they are done. Regardless of the presence
or absence of r.support, the individual modules are useful in their
own right.

[Otherwise, r.region wouldn't exist.]

--
Glynn Clements <glynn@gclements.plus.com>