On Thu, Jul 20, 2006 at 01:55:47PM -0700, Michael Barton wrote:
Markus,
I think I've done all outstanding tcltk bug fixes AND finished the
georectifier. All are in the cvs in the standard grass6 location if you want
to backport any to 6.1.0.
Great, Michael!
I would suggest to let developers and power users test a
few days before doing the backport. Or do you have
pressing new featues to submit? 
Markus
On Thu, Jul 20, 2006 at 11:01:41PM +0200, Markus Neteler wrote:
On Thu, Jul 20, 2006 at 01:55:47PM -0700, Michael Barton wrote:
> Markus,
>
> I think I've done all outstanding tcltk bug fixes AND finished the
> georectifier. All are in the cvs in the standard grass6 location if you want
> to backport any to 6.1.0.Great, Michael!
I would suggest to let developers and power users test a
few days before doing the backport. Or do you have
pressing new featues to submit?
I have now backported all tcltk fixes into the 6.1.0-branch.
Also added ODBC documentation.
Next week we should get out 6.1.0beta2 or maybe RC1.
Markus
On Sat, Jul 22, 2006 at 12:37:53PM +0200, Markus Neteler wrote:
Next week we should get out 6.1.0beta2 or maybe RC1.
We still have the following open issues:
Showstoppers:
* NVIZ volume broken (IMHO showstopper, since this worked already)
https://intevation.de/rt/webrt?serial_num=4725
* wording of start up GUI text (showstopper)
Not showstoppers:
* NVIZ: show vector map on startup (Bob? Your patch doesn't work for me)
* debian/ directory outdated (who cares)
* i.vpoints status?
* v.in.gpsbabel - db.login error (who cares?)
http://grass.itc.it/pipermail/grass-dev/2006-July/024188.html
* test latest gis.m fixes
* G_make_aspect*_colors() broken
http://grass.itc.it/pipermail/grass-dev/2006-July/024228.html
* sqlite driver memory consumption
http://grass.itc.it/pipermail/grass-dev/2006-July/024232.html
* v.digit: $mapname in the title bar
http://intevation.de/rt/webrt?serial_num=4796
Further discussion (old business):
* The x11-less GRASS can be developed in HEAD, we should
not wait for that. We can have 6.1.1 if desired in near
future. The same applies to the georectifier.
* NVIZ/Mac issues we can port from HEAD once they get fixed.
* snprintf(): now G_snprintf()
* An announcement is drafted at
http://grass.itc.it/announces/announce_grass610.html
Markus
On Jul 22, 2006, at 7:25 AM, Markus Neteler wrote:
On Sat, Jul 22, 2006 at 12:37:53PM +0200, Markus Neteler wrote:
Next week we should get out 6.1.0beta2 or maybe RC1.
We still have the following open issues:
Showstoppers:
* NVIZ volume broken (IMHO showstopper, since this worked already)
https://intevation.de/rt/webrt?serial_num=4725
* wording of start up GUI text (showstopper)
this needs to be changed to what Markus has suggested - see my previous email.
projection location is really confusing
Not showstoppers:
* NVIZ: show vector map on startup (Bob? Your patch doesn't work for me)
* debian/ directory outdated (who cares)
* i.vpoints status?
* v.in.gpsbabel - db.login error (who cares?)
http://grass.itc.it/pipermail/grass-dev/2006-July/024188.html
* test latest gis.m fixes
* G_make_aspect*_colors() broken
http://grass.itc.it/pipermail/grass-dev/2006-July/024228.html
is this really broken? What I said in that email is that it was always grey scale and
that I don't think it should be changed right now as many people are probably used to
this color table rather than the one used in v.surf.rst.
* sqlite driver memory consumption
http://grass.itc.it/pipermail/grass-dev/2006-July/024232.html
* v.digit: $mapname in the title bar
http://intevation.de/rt/webrt?serial_num=4796Further discussion (old business):
* The x11-less GRASS can be developed in HEAD, we should
not wait for that. We can have 6.1.1 if desired in near
future. The same applies to the georectifier.
* NVIZ/Mac issues we can port from HEAD once they get fixed.
* snprintf(): now G_snprintf()
* An announcement is drafted at
http://grass.itc.it/announces/announce_grass610.htmlMarkus
_______________________________________________
grass-dev mailing list
grass-dev@grass.itc.it
http://grass.itc.it/mailman/listinfo/grass-dev
Markus,
On Sat, 2006-07-22 at 13:25 +0200, Markus Neteler wrote:
On Sat, Jul 22, 2006 at 12:37:53PM +0200, Markus Neteler wrote:
> Next week we should get out 6.1.0beta2 or maybe RC1.We still have the following open issues:
Showstoppers:
* NVIZ volume broken (IMHO showstopper, since this worked already)
https://intevation.de/rt/webrt?serial_num=4725
* wording of start up GUI text (showstopper)Not showstoppers:
* NVIZ: show vector map on startup (Bob? Your patch doesn't work for me)
The two potential problems with the patch are the vertical exaggeration,
and the initial line color. If the vertical exaggeration is too high the
vector may be out out view. If you lower the vertical exaggeration does
the vector come into view? If the initial line color (set in map_obj.c)
is white and the background is white the vector will be invisible. Try
changing the color of the vector after start-up to see if the vector
appears.
In the meantime I will keep looking into this.
* debian/ directory outdated (who cares)
* i.vpoints status?
* v.in.gpsbabel - db.login error (who cares?)
http://grass.itc.it/pipermail/grass-dev/2006-July/024188.html
* test latest gis.m fixes
* G_make_aspect*_colors() broken
http://grass.itc.it/pipermail/grass-dev/2006-July/024228.html
* sqlite driver memory consumption
http://grass.itc.it/pipermail/grass-dev/2006-July/024232.html
* v.digit: $mapname in the title bar
http://intevation.de/rt/webrt?serial_num=4796Further discussion (old business):
* The x11-less GRASS can be developed in HEAD, we should
not wait for that. We can have 6.1.1 if desired in near
future. The same applies to the georectifier.
* NVIZ/Mac issues we can port from HEAD once they get fixed.
* snprintf(): now G_snprintf()
* An announcement is drafted at
http://grass.itc.it/announces/announce_grass610.htmlMarkus
_______________________________________________
grass-dev mailing list
grass-dev@grass.itc.it
http://grass.itc.it/mailman/listinfo/grass-dev
--
Bob
I planned to email this after the release, but given the needed modifications
in the startup panel text I am wondering whether the following change
would be useful to make (see the reasoning behind it below).
Change the order of options for defining new location and
modify the text to make it easier to understand
(comments, modifications are welcome, e.g. should we use
coordinate system rather than projection?):
1. Use georeferenced file
2. Select a pre-defined projection (EPSG code)
3. Define parameters of your projection
Helena
--- discussion on how to improve starting GRASS with new location-------
I looks like this may not be as difficult as percieved if we take into account
what the user has before he starts to work with GRASS:
User who wants to start grass may have:
1. georeferenced file with prj info
2. georeferenced file but no prj - should we try to guess the coord. system -
lat/long would be easy, UTM maybe?, state plane would be a lot of work, I don't know
about the rest of the world. I don't think that you can guess a datum?
3. non-georeferenced file
4. nothing - can this happen?
I think that most users will have 1. or 2. (when using official sources
it should really be 1.)
So we can first re-arrange the startup menu to encourage people use 1.
and then make it easier to set it up for case 2:
1. put "Use georeferenced file" option first
this option should take care of both the coordinate system and region (spatial extent)
2. "Use EPSG values" would stay second choice but we can make it
more user friendly:
- please suggest a more intuitive text: for example,
"Select a pre-defined projection (EPSG codes)"
- is it possible to make EPSG list searchable? for example by the state,
name of the projection, datum (I knew what I wanted but it took me forever to find
it in the list)
This option does not create default region, so that is one piece of
the interface that needs to be created. A really nice thing would be a little world
map where you could zoom in to your approximate area and the resolution would
be set automatically so that the no. of rows or cols does not exceed e.g., 1000.
(projecting lat/long into the given projection may be needed, so this may be more
complex than I thought)
The option to type or paste in the coordinates should be there too.
3. "Use projection values" should be last (most difficult) option
(your coordinate system is not in EPSG and your georef file does not have the adequate
projection/datum information or you don't have any georef. file)
- should we change the text to something more intuitive, e.g.,
"Define parameters of your projection"?
- should we have a separate button for starting with non-georeferenced (x,y) location
which is easy to set up?
the rest is pretty complicated and may need a separate discussion.
Markus wrote:
> Next week we should get out 6.1.0beta2 or maybe RC1.
We still have the following open issues:
Showstoppers:
* NVIZ volume broken (IMHO showstopper, since this worked already)
https://intevation.de/rt/webrt?serial_num=4725
it should be possible to do checkout by date
cvs checkout -D 2006-02-28 grass6/
to help narrow down the change that broke it.
maybe try Lorenzo's binary archives or if anyone keeps old linux
binary snapshots?
http://wwwamb.bologna.enea.it/forgrass/files/archives/
Not showstoppers:
..
* debian/ directory outdated (who cares)
Leave to the DebianGIS + UbuntuGIS teams to decide if it's needed?
Replace with links to DebianGIS + UbuntuGIS control files? (Alioth SVN)
(no opinion otherwise)
* i.vpoints status?
working-just-fine in HEAD, I found a bug where if you load a raster
target image before the vector target image, the un-zoomed (top right)
panel doesn't line up correctly. If you load a vector map first it's ok.
not a release-critical bug IMO.
zoomed reverse transform warp overlay (in bottom left panel) remains
TODO.
After overseeing some work with i.points I think it is useful to keep as
a "simple" solution.
* v.digit: $mapname in the title bar
http://intevation.de/rt/webrt?serial_num=4796
suggest leaving as "v.digit - \$mapname" in HEAD; in the release branch
change to either "v.digit" or remove the line so the title is the
mapname.
I'm gone for a month, so I leave it to others to fix my bugs 
Hamish
A big showstopper is also NVIZ being broken on Macs. Until William
Kyngesburye's report of today, the cvs version was broken on all Macs under
all TclTk versions.
I don't know what changed, but William reports that it is now working under
x11 on all Macs. So this is a huge improvement. I'll try to confirm this
next week. If he can get it working under aqua, we'll be set.
Michael
__________________________________________
Michael Barton, Professor of Anthropology
School of Human Evolution & Social Change
Center for Social Dynamics & Complexity
Arizona State University
phone: 480-965-6213
fax: 480-965-7671
www: http://www.public.asu.edu/~cmbarton
From: Bob Covill <bcovill@tekmap.ns.ca>
Date: Sat, 22 Jul 2006 12:23:55 -0300
To: Markus Neteler <neteler@itc.it>
Cc: GRASS developers list <grass-dev@grass.itc.it>
Subject: Re: [GRASS-dev] GRASS 6.1.0 branch statusMarkus,
On Sat, 2006-07-22 at 13:25 +0200, Markus Neteler wrote:
On Sat, Jul 22, 2006 at 12:37:53PM +0200, Markus Neteler wrote:
Next week we should get out 6.1.0beta2 or maybe RC1.
We still have the following open issues:
Showstoppers:
* NVIZ volume broken (IMHO showstopper, since this worked already)
https://intevation.de/rt/webrt?serial_num=4725
* wording of start up GUI text (showstopper)Not showstoppers:
* NVIZ: show vector map on startup (Bob? Your patch doesn't work for me)The two potential problems with the patch are the vertical exaggeration,
and the initial line color. If the vertical exaggeration is too high the
vector may be out out view. If you lower the vertical exaggeration does
the vector come into view? If the initial line color (set in map_obj.c)
is white and the background is white the vector will be invisible. Try
changing the color of the vector after start-up to see if the vector
appears.In the meantime I will keep looking into this.
* debian/ directory outdated (who cares)
* i.vpoints status?
* v.in.gpsbabel - db.login error (who cares?)
http://grass.itc.it/pipermail/grass-dev/2006-July/024188.html
* test latest gis.m fixes
* G_make_aspect*_colors() broken
http://grass.itc.it/pipermail/grass-dev/2006-July/024228.html
* sqlite driver memory consumption
http://grass.itc.it/pipermail/grass-dev/2006-July/024232.html
* v.digit: $mapname in the title bar
http://intevation.de/rt/webrt?serial_num=4796Further discussion (old business):
* The x11-less GRASS can be developed in HEAD, we should
not wait for that. We can have 6.1.1 if desired in near
future. The same applies to the georectifier.
* NVIZ/Mac issues we can port from HEAD once they get fixed.
* snprintf(): now G_snprintf()
* An announcement is drafted at
http://grass.itc.it/announces/announce_grass610.htmlMarkus
_______________________________________________
grass-dev mailing list
grass-dev@grass.itc.it
http://grass.itc.it/mailman/listinfo/grass-dev--
Bob
On Jul 23, 2006, at 12:37 PM, Michael Barton wrote:
A big showstopper is also NVIZ being broken on Macs. Until William
Kyngesburye's report of today, the cvs version was broken on all Macs under
all TclTk versions.I don't know what changed, but William reports that it is now working under
x11 on all Macs. So this is a huge improvement. I'll try to confirm this
next week.
I hope you and others can confirm it's working...
If he can get it working under aqua, we'll be set.
Well, that's starting to get into major C programming and NVIZ/OpenGL stuff. A bit beyond me. I'll probably have to add a bug report for it so someone else can start looking into it.
And it seems to have mutated into two problems: crashing on Intel, and the 'twitch' in the NVIZ file browsers on both PPC and Intel. I need to check the PPC Mac at work where it worked once to see if there is a simple solution to the 'twitch' in the NVIZ browsers.
Further discussion (old business):
* The x11-less GRASS can be developed in HEAD, we should
not wait for that. We can have 6.1.1 if desired in near
future. The same applies to the georectifier.
* NVIZ/Mac issues we can port from HEAD once they get fixed.
-----
William Kyngesburye <kyngchaos@kyngchaos.com>
http://www.kyngchaos.com/
Earth: "Mostly harmless"
- revised entry in the HitchHiker's Guide to the Galaxy
Markus Neteler napisa³(a):
On Sat, Jul 22, 2006 at 12:37:53PM +0200, Markus Neteler wrote:
Next week we should get out 6.1.0beta2 or maybe RC1.
We still have the following open issues:
Showstoppers:
* NVIZ volume broken (IMHO showstopper, since this worked already)
https://intevation.de/rt/webrt?serial_num=4725
* wording of start up GUI text (showstopper)Not showstoppers:
* NVIZ: show vector map on startup (Bob? Your patch doesn't work for me)
Other serious unresolved NVIZ issues:
http://intevation.de/rt/webrt?serial_num=2887
http://intevation.de/rt/webrt?serial_num=2741
http://intevation.de/rt/webrt?serial_num=3041
http://intevation.de/rt/webrt?serial_num=3506
http://intevation.de/rt/webrt?serial_num=4246
http://intevation.de/rt/webrt?serial_num=4556
http://intevation.de/rt/webrt?serial_num=4725
http://intevation.de/rt/webrt?serial_num=4492
* v.digit: $mapname in the title bar
http://intevation.de/rt/webrt?serial_num=4796
other still valid v.digit bugs:
http://intevation.de/rt/webrt?serial_num=2962
http://intevation.de/rt/webrt?serial_num=4110 (the "Could not set Tcl
system encoding to" part of the report)
http://intevation.de/rt/webrt?serial_num=4339 (related to the one above)
http://intevation.de/rt/webrt?serial_num=4439
http://intevation.de/rt/webrt?serial_num=4454
I did my best to pick only those which I know/strongly suspect that are
still valid bugs; there are more grass6 bug reports for NVIZ and v.digit
of unknown status (for me), worthy of having a a look at them though.
Maciek
Michael Barton wrote:
I don't know what changed, but William reports that it is now working under
x11 on all Macs. So this is a huge improvement. I'll try to confirm this
next week. If he can get it working under aqua, we'll be set.
I wouldn't suggest holding up 6.1 for issues issues related to the
Aqua version of NVIZ. Support for non-X11 versions of Tcl/Tk is a new
feature, and it's likely to take a long time to sort out the problems.
AFAICT, the main problem is that, on platforms other than X11, Togl
still needs to use internal Tk headers. These aren't normally
installed as part of the vendor's frameworks, and we don't include
local copies. Also, using internal headers makes the resulting
executable fragile, in that it won't work with versions of Tk which
materially differ from the version for which it was compiled.
--
Glynn Clements <glynn@gclements.plus.com>
Maciej Sieczka wrote:
>> Next week we should get out 6.1.0beta2 or maybe RC1.
>
> We still have the following open issues:
>
> Showstoppers:
> * NVIZ volume broken (IMHO showstopper, since this worked already)
> https://intevation.de/rt/webrt?serial_num=4725
> * wording of start up GUI text (showstopper)
>
> Not showstoppers:
> * NVIZ: show vector map on startup (Bob? Your patch doesn't work for me)Other serious unresolved NVIZ issues:
Regarding this one:
As I have pointed out before, the problems with image dumping are due
to this:
# Set the cancel function for drawing
Nset_cancel_func update
This simply won't work.
Calling "update" can result in arbitrary event handlers being called.
Some of them (i.e. almost anything related to the Togl canvas) will
result in changes to the OpenGL context which will break the rendering
performed by the image dumping code.
Either the cancellation function will have to restore the OpenGL state
(and possibly some NVIZ state) before returning, or the OGSF library
will need to be modified to allow long rendering tasks to be performed
incrementally (i.e. replacing do-it-all with separate
begin/do-one-part/end operations).
Or, alternatively, you'll just have to live with the UI being
unresponsive during image dumping.
--
Glynn Clements <glynn@gclements.plus.com>
On Sun, Jul 23, 2006 at 10:38:20PM +0200, Maciej Sieczka wrote:
Markus Neteler napisa?(a):
> On Sat, Jul 22, 2006 at 12:37:53PM +0200, Markus Neteler wrote:
>> Next week we should get out 6.1.0beta2 or maybe RC1.
>
> We still have the following open issues:
>
> Showstoppers:
> * NVIZ volume broken (IMHO showstopper, since this worked already)
> https://intevation.de/rt/webrt?serial_num=4725
> * wording of start up GUI text (showstopper)
>
> Not showstoppers:
> * NVIZ: show vector map on startup (Bob? Your patch doesn't work for me)Other serious unresolved NVIZ issues:
http://intevation.de/rt/webrt?serial_num=2887
http://intevation.de/rt/webrt?serial_num=2741
http://intevation.de/rt/webrt?serial_num=3041
http://intevation.de/rt/webrt?serial_num=3506
http://intevation.de/rt/webrt?serial_num=4246
http://intevation.de/rt/webrt?serial_num=4556
http://intevation.de/rt/webrt?serial_num=4725
http://intevation.de/rt/webrt?serial_num=4492> * v.digit: $mapname in the title bar
> http://intevation.de/rt/webrt?serial_num=4796other still valid v.digit bugs:
http://intevation.de/rt/webrt?serial_num=2962
http://intevation.de/rt/webrt?serial_num=4110 (the "Could not set Tcl
system encoding to" part of the report)
http://intevation.de/rt/webrt?serial_num=4339 (related to the one above)
http://intevation.de/rt/webrt?serial_num=4439
http://intevation.de/rt/webrt?serial_num=4454I did my best to pick only those which I know/strongly suspect that are
still valid bugs; there are more grass6 bug reports for NVIZ and v.digit
of unknown status (for me), worthy of having a a look at them though.
Please keep in mind that (IMHO)
- we should follow a bit closer the "release often" paradigm
- that we just want to get out 6.1.0 aka unstable
Of course I hope that bugs are fixed...
Markus
On Sat, Jul 22, 2006 at 12:37:53PM +0200, Markus Neteler wrote:
On Thu, Jul 20, 2006 at 11:01:41PM +0200, Markus Neteler wrote:
> On Thu, Jul 20, 2006 at 01:55:47PM -0700, Michael Barton wrote:
> > Markus,
> >
> > I think I've done all outstanding tcltk bug fixes AND finished the
> > georectifier. All are in the cvs in the standard grass6 location if you want
> > to backport any to 6.1.0.
>
> Great, Michael!
>
> I would suggest to let developers and power users test a
> few days before doing the backport. Or do you have
> pressing new featues to submit?
>I have now backported all tcltk fixes into the 6.1.0-branch.
Also added ODBC documentation.Next week we should get out 6.1.0beta2 or maybe RC1.
Status:
- The intro screen is at least improved now.
- some bugs in d.m and gis.m fixed
- some other minor fixes done
open:
- NVIZ:
- volume still broken (a pity!!)
- vector not shown on startup when loaded together with DEM
- Glynn's MingW fixes not ported (not sure if we want that)
- debian control files still outdated (since there is low
interest, just remove them?)
Compare
http://grass.itc.it/pipermail/grass-commit/2006-July/date.html#end
I think that we'll get out 6.1.0beta2 tomorrow or on the day after.
Markus
Markus Neteler wrote:
- debian control files still outdated (since there is low
interest, just remove them?)
What is outdated about them ?
I'd like for them to stay since I use them to recompile the Debian way.
Moritz
On Tue, Aug 01, 2006 at 11:14:01AM +0200, Moritz Lennert wrote:
Markus Neteler wrote:
>- debian control files still outdated (since there is low
> interest, just remove them?)
>What is outdated about them ?
just the versions of suggested libraries and external programs are too
old (tk 8.3 -> 8.4, ...)
jachym
I'd like for them to stay since I use them to recompile the Debian way.
Moritz
_______________________________________________
grass-dev mailing list
grass-dev@grass.itc.it
http://grass.itc.it/mailman/listinfo/grass-dev
--
Jachym Cepicky
e-mail: jachym.cepicky@centrum.cz
URL: http://les-ejk.cz
GPG: http://les-ejk.cz/gnupg_public_key/jachym_cepicky-gpg_public_key.asc
-----------------------------------------
OFFICE:
GDF-Hannover
Mengendamm 16d
30177 Hannover
Germany
e-mail: cepicky@gdf-hannover.de
URL: http://gdf-hannover.de
Tel.: +49 511-39088507
On Tue, Aug 01, 2006 at 11:19:11AM +0200, Jachym Cepicky wrote:
On Tue, Aug 01, 2006 at 11:14:01AM +0200, Moritz Lennert wrote:
> Markus Neteler wrote:
>
> >- debian control files still outdated (since there is low
> > interest, just remove them?)
> >
>
> What is outdated about them ?
>just the versions of suggested libraries and external programs are too
old (tk 8.3 -> 8.4, ...)
Right. But I guess that we want tk >= 8.3 etc here.
The idea is to *not* hardcode the version any longer
if not needed. Since I am no debian user I dunno
how to fix the control file.
Markus
Markus Neteler wrote:
- Glynn's MingW fixes not ported (not sure if we want that)
The changes to the raster library are too invasive for 6.1.0, but the
old version won't compile on MinGW (no socket functions).
Consequently, 6.1.0 won't be particularly usable on MinGW, so there
probably isn't much point in backporting the various minor tweaks
either.
--
Glynn Clements <glynn@gclements.plus.com>
Glynn Clements wrote on 08/01/2006 02:13 PM:
Markus Neteler wrote:
- Glynn's MingW fixes not ported (not sure if we want that)
The changes to the raster library are too invasive for 6.1.0, but the
old version won't compile on MinGW (no socket functions).
Consequently, 6.1.0 won't be particularly usable on MinGW, so there
probably isn't much point in backporting the various minor tweaks
either.
Right. Given this I suggest to get out GRASS 6.1.0 now and then branch
it again later
on or so.
Markus
Markus Neteler napisa?(a):
On Sun, Jul 23, 2006 at 10:38:20PM +0200, Maciej Sieczka wrote:
<snip>
I did my best to pick only those which I know/strongly suspect that are
still valid bugs; there are more grass6 bug reports for NVIZ and v.digit
of unknown status (for me), worthy of having a a look at them though.Please keep in mind that (IMHO)
- we should follow a bit closer the "release often" paradigm
That's OK with me.
- that we just want to get out 6.1.0 aka unstable
Among those bugs listed (and the remaining ones in the BT) there are
some which where reported when 5.7 and 6.0 where unstable, but they
remain not fixed for stable 6.01, 6.02, nor in the 6.0 CVS branch. They
aren't fixed in 6.1 branch too. Let's hope they will be fixed for 6.2
stable.
Of course I hope that bugs are fixed...
I'm using an occasion to remind that there are bugs left. Not everybody
is up to date with the BT; maybe a reminder will trigger developers'
attention to the long forgotten problems.
Maciek