I just noticed another, more severe problem on Windows:
I can only use ESC key sequences in the first interactive text input
window. When I progress to the next one, they don't work
anymore. This means that I cannot create a new location
from scratch under WinGRASS, because I can exit the
first screen (the one asking for names) with ESC+ENTER,
but the next one (the one for setting the default region)
does not interpret EST+ENTER correctly anymore, meaning
I can only quit it with CTRL+C!
This happens with my GRASS binaries and Moritz'
Benjamin
--
Benjamin Ducke, M.A.
Archäoinformatik
(Archaeoinformation Science)
Institut für Ur- und Frühgeschichte
(Inst. of Prehistoric and Historic Archaeology)
Christian-Albrechts-Universität zu Kiel
Johanna-Mestorf-Straße 2-6
D 24098 Kiel
Germany
I just noticed another, more severe problem on Windows:
I can only use ESC key sequences in the first interactive text input
window. When I progress to the next one, they don't work
anymore. This means that I cannot create a new location
from scratch under WinGRASS, because I can exit the
first screen (the one asking for names) with ESC+ENTER,
but the next one (the one for setting the default region)
does not interpret EST+ENTER correctly anymore, meaning
I can only quit it with CTRL+C!
Too bad. The only thing I can think of is to replace the text
interface with a Tcl/Tk wizard. That would also make it much
easier for starters. However, who's going to put all the time
into that with a 6.3.0 release pending?
I made another ugly discovery: cmd.exe's tab completion is
_turned off_ by default on Win >=2000 (unbelievable!).
It could be done by writing a registry value (sic!) into
the user's registry tree.
Should I add that to grass63.bat?
Benjamin
Moritz Lennert wrote:
On 31/10/07 14:25, Benjamin Ducke wrote:
I just noticed another, more severe problem on Windows:
I can only use ESC key sequences in the first interactive text input
window. When I progress to the next one, they don't work
anymore. This means that I cannot create a new location
from scratch under WinGRASS, because I can exit the
first screen (the one asking for names) with ESC+ENTER,
but the next one (the one for setting the default region)
does not interpret EST+ENTER correctly anymore, meaning
I can only quit it with CTRL+C!
I still haven't had the time to look into this, but it is definitely a
show stopper IMHO.
Moritz
--
Benjamin Ducke, M.A.
Archäoinformatik
(Archaeoinformation Science)
Institut für Ur- und Frühgeschichte
(Inst. of Prehistoric and Historic Archaeology)
Christian-Albrechts-Universität zu Kiel
Johanna-Mestorf-Straße 2-6
D 24098 Kiel
Germany
Too bad. The only thing I can think of is to replace the text
interface with a Tcl/Tk wizard. That would also make it much
easier for starters. However, who's going to put all the time
into that with a 6.3.0 release pending?
Noone, not only because of the 6.3.0 release, but also because there is exactly such an interface in wxgrass... This will be one of the next steps, BTW: testing of wxgrass on Windows...
I made another ugly discovery: cmd.exe's tab completion is
_turned off_ by default on Win >=2000 (unbelievable!).
It could be done by writing a registry value (sic!) into
the user's registry tree.
Should I add that to grass63.bat?
Please don't. If people want tab completion, they can set it themselves. This is not related to GRASS.
I just noticed another, more severe problem on Windows:
I can only use ESC key sequences in the first interactive text input
window. When I progress to the next one, they don't work
anymore. This means that I cannot create a new location
from scratch under WinGRASS, because I can exit the
first screen (the one asking for names) with ESC+ENTER,
but the next one (the one for setting the default region)
does not interpret EST+ENTER correctly anymore, meaning
I can only quit it with CTRL+C!
This happens with my GRASS binaries and Moritz'
That would appear to be a problem with the curses implementation.
The Esc-Enter behaviour is part of the vask library (V_call(), in
lib/vask/V_call.c). If it works in some cases and not others, I'm not
sure what can be done.
I still don't have a working Windows system, so I'm unable to debug it
myself.
Too bad. The only thing I can think of is to replace the text
interface with a Tcl/Tk wizard. That would also make it much
easier for starters. However, who's going to put all the time
into that with a 6.3.0 release pending?
Noone, not only because of the 6.3.0 release, but also because there is
exactly such an interface in wxgrass... This will be one of the next
steps, BTW: testing of wxgrass on Windows...
I have been looking into WxGRASS under Windows, so far to no avail.
But I am sure it can be done. The different pieces of information
on the web are just very -- well, complex.
I made another ugly discovery: cmd.exe's tab completion is
_turned off_ by default on Win >=2000 (unbelievable!).
It could be done by writing a registry value (sic!) into
the user's registry tree.
Should I add that to grass63.bat?
Please don't. If people want tab completion, they can set it themselves.
This is not related to GRASS.
Right, I found that setting registry entries from the command line
requires add-on software, anyhow.
Moritz
--
Benjamin Ducke, M.A.
Archäoinformatik
(Archaeoinformation Science)
Institut für Ur- und Frühgeschichte
(Inst. of Prehistoric and Historic Archaeology)
Christian-Albrechts-Universität zu Kiel
Johanna-Mestorf-Straße 2-6
D 24098 Kiel
Germany
> I made another ugly discovery: cmd.exe's tab completion is
> _turned off_ by default on Win >=2000 (unbelievable!).
> It could be done by writing a registry value (sic!) into
> the user's registry tree.
> Should I add that to grass63.bat?
Moritz:
Please don't. If people want tab completion, they can set it themselves.
This is not related to GRASS.
I agree with Moritz, maybe the user set it that way for a reason.
But if you have a quick way of turning it on (ISTR that either the TweakUI
PowerTool or cmd.exe's context menu Properties gave you the option to turn it
on) You could leave it commented out in the .bat file, and/or add a note to the
WinGRASS help page* about how to turn it on and link to the page that talks
about enabling it for map names**.
I think this whole tab completion thing on Win is just too
friggin' complicated and time-consuming for the moment.
Better to get down to resolving the more serious issues
first for the 6.3.0 release.
Benjamin
Hamish wrote:
Benjamin:
I made another ugly discovery: cmd.exe's tab completion is
_turned off_ by default on Win >=2000 (unbelievable!).
It could be done by writing a registry value (sic!) into
the user's registry tree.
Should I add that to grass63.bat?
Moritz:
Please don't. If people want tab completion, they can set it themselves.
This is not related to GRASS.
I agree with Moritz, maybe the user set it that way for a reason.
But if you have a quick way of turning it on (ISTR that either the TweakUI
PowerTool or cmd.exe's context menu Properties gave you the option to turn it
on) You could leave it commented out in the .bat file, and/or add a note to the
WinGRASS help page* about how to turn it on and link to the page that talks
about enabling it for map names**.
__________________________________________________
Do You Yahoo!?
Tired of spam? Yahoo! Mail has the best spam protection around http://mail.yahoo.com
--
Benjamin Ducke, M.A.
Archäoinformatik
(Archaeoinformation Science)
Institut für Ur- und Frühgeschichte
(Inst. of Prehistoric and Historic Archaeology)
Christian-Albrechts-Universität zu Kiel
Johanna-Mestorf-Straße 2-6
D 24098 Kiel
Germany
On Wed, October 31, 2007 16:28, Glynn Clements wrote:
Benjamin Ducke wrote:
I just noticed another, more severe problem on Windows:
I can only use ESC key sequences in the first interactive text input
window. When I progress to the next one, they don't work
anymore. This means that I cannot create a new location
from scratch under WinGRASS, because I can exit the
first screen (the one asking for names) with ESC+ENTER,
but the next one (the one for setting the default region)
does not interpret EST+ENTER correctly anymore, meaning
I can only quit it with CTRL+C!
This happens with my GRASS binaries and Moritz'
That would appear to be a problem with the curses implementation.
The Esc-Enter behaviour is part of the vask library (V_call(), in
lib/vask/V_call.c). If it works in some cases and not others, I'm not
sure what can be done.
I've just had a look at lib/init/set_data.c and don't really see what the
problem is. A clue we have is that ESC-ENTER works in the first screen,
but not in the region definition screen. In terms of code, the only
immediate difference between the two I see is that the first is
constructed directly in set_data.c, including the V_call(), whereas the
second is called with E_edit_cellhd(&window, -1), so the actual screen
construction and V_call() are in lib/edit/edit_cellhd.c. Could that make
any difference ?
I'll try to add some debug code, but first have to understand how to best
do that in the vask environment.
On Thu, November 1, 2007 13:28, Moritz Lennert wrote:
On Wed, October 31, 2007 16:28, Glynn Clements wrote:
Benjamin Ducke wrote:
I just noticed another, more severe problem on Windows:
I can only use ESC key sequences in the first interactive text input
window. When I progress to the next one, they don't work
anymore. This means that I cannot create a new location
from scratch under WinGRASS, because I can exit the
first screen (the one asking for names) with ESC+ENTER,
but the next one (the one for setting the default region)
does not interpret EST+ENTER correctly anymore, meaning
I can only quit it with CTRL+C!
This happens with my GRASS binaries and Moritz'
That would appear to be a problem with the curses implementation.
The Esc-Enter behaviour is part of the vask library (V_call(), in
lib/vask/V_call.c). If it works in some cases and not others, I'm not
sure what can be done.
I've just had a look at lib/init/set_data.c and don't really see what the
problem is. A clue we have is that ESC-ENTER works in the first screen,
but not in the region definition screen. In terms of code, the only
immediate difference between the two I see is that the first is
constructed directly in set_data.c, including the V_call(), whereas the
second is called with E_edit_cellhd(&window, -1), so the actual screen
construction and V_call() are in lib/edit/edit_cellhd.c. Could that make
any difference ?
I'll try to add some debug code, but first have to understand how to best
do that in the vask environment.
A little more info:
When I leave the region definition screen using Ctrl-C, I get back to the
very first screen where you set location, mapset, etc. And this second
time, I cannot use ESC-ENTER either. The first time on that screen (when I
launch grass) it works... Is there something that remains in memory and
which confuses V_call ?
I don't know why but pushing enter gives a CR (carriage return) the very
first time, but from then on always gives NL (newline). But V_call only
checked for the first. Fixed in cvs:
I don't know why but pushing enter gives a CR (carriage return) the very
first time, but from then on always gives NL (newline). But V_call only
checked for the first.
This is supposed to be handled by the nonl() and nl() calls in
V_init() and V_exit() respectively. The input loop in V_call() should
only ever see CR:
The nl and nonl routines control whether the underlying display device
translates the return key into newline on input, and whether it trans-
lates newline into return and line-feed on output (in either case, the
call addch('\n') does the equivalent of return and line feed on the
virtual screen). Initially, these translations do occur. If you dis-
able them using nonl, curses will be able to make better use of the
line-feed capability, resulting in faster cursor motion. Also, curses
will then be able to detect the return key.