[GRASS5] Grass on CYGWIN/WIN32 - IPC Version Status Update

Malcolm_Blue1 · January 4, 2001, 11:34pm

Good news and bad news…

GOOD (for WinNT users):

Looks like I’m over the major hurdles in the IPC code. I have several changes to send to the maintainer, but I am able to get the IPC Message Queues to connect/send/receive/disconnect. (Still more to do: clean up code a bit, increase the number of queues and increase message buffer sizes look like the first priority, but these should be simple changes.)
d.mon programs (start, select, stop) are working for me, as well as a VERY limited set of display commands on WinNT. I still have some minor clean up and testing to do to make sure they are ready for testing by others.
d.erase and d.zoom seem to work. d.erase erased the little that drew with d.rast, and d.zoom did a proper zoom for the little band of pixels that d.rast created (see below), with a cursor, the zoom “rubber band” box and the coordinates and prompts in the command window were in sync with the graphics window operations. Probably a number of other commands will work as well.

In order to run “d.mon select” without code changes, I had to remove the “#ifndef CYGWIN” and “#endif” statements that were in the switch statement in src/display/devices/XDRIVER/XDRIVER24/ipc.new/switcher.c , i.e. sandwiching the FONT and TEXT case blocks. Does anyone know why these cases were being excluded? Is this a relic from the old XFree port? These blocks of code seem to compile fine, and the “case FONT” code block works with the recent XFree release for Cygwin. I think this should be changed for both the original (fifo) version and the ipc version.

BAD:

Limited success with the d.rast command. It only seems to draw the NULL’s. Using the Spearfish sample data, I did:

d.rast soils bg=yellow - it draws a yellow band near the bottom of the X0 monitor window (several pixels wide and the width of the window).
d.rast soils bg=white - it draws a white band where the yellow band was.
d.rast streams bg=red - it draws a red band over the bottom part of the white band

From these and additional tests I am guessing that just the NULL pixels are being drawn. Does this sound right?
BTW: It was based on this limited result that I was able to test d.erase and d.zoom

d.vect, d.sites and d.legend all hang up. They had to be killed from a separate window.
Win98 did not work. The IPC code by itself tested OK, but the d.mon commands were hanging up. WinNT was used for the limited successes above.

I will keep looking at the IPC code and do the minor changes I mentioned above. I also plan to keep on looking into the d.rast problems to see if I can get this working. Suggestions are welcome.

Malcolm

---------------------------------------- If you want to unsubscribe from GRASS Development Team mailing list write to: minordomo@geog.uni-hannover.de with subject ‘unsubscribe grass5’

Malcolm_Blue1 · January 4, 2001, 11:41pm

Oops, I sent that as HTML mail, which some readers may not like.
Sorry.

Malcolm Blue wrote:

Good news and bad news...

GOOD (for WinNT users):

1. Looks like I'm over the major hurdles in the IPC code. I have
several changes to send to the maintainer, but I am able to get the
IPC Message Queues to connect/send/receive/disconnect. (Still more to
do: clean up code a bit, increase the number of queues and increase
message buffer sizes look like the first priority, but these should be
simple changes.)

2. d.mon programs (start, select, stop) are working for me, as well
as a VERY limited set of display commands on WinNT. I still have some
minor clean up and testing to do to make sure they are ready for
testing by others.

3. d.erase and d.zoom seem to work. d.erase erased the little that
drew with d.rast, and d.zoom did a proper zoom for the little band of
pixels that d.rast created (see below), with a cursor, the zoom
"rubber band" box and the coordinates and prompts in the command
window were in sync with the graphics window operations. Probably a
number of other commands will work as well.

In order to run "d.mon select" without code changes, I had to remove
the "#ifndef _CYGWIN_" and "#endif" statements that were in the
switch statement in
src/display/devices/XDRIVER/XDRIVER24/ipc.new/switcher.c , i.e.
sandwiching the FONT and TEXT case blocks. Does anyone know why these
cases were being excluded? Is this a relic from the old XFree port?
These blocks of code seem to compile fine, and the "case FONT" code
block works with the recent XFree release for Cygwin. I think this
should be changed for both the original (fifo) version and the ipc
version.

BAD:

1. Limited success with the d.rast command. It only seems to draw the
NULL's. Using the Spearfish sample data, I did:

     d.rast soils bg=yellow - it draws a yellow band near the
     bottom of the X0 monitor window (several pixels wide and the
     width of the window).
     d.rast soils bg=white - it draws a white band where the
     yellow band was.
     d.rast streams bg=red - it draws a red band over the
     bottom part of the white band

     From these and additional tests I am guessing that just the
     NULL pixels are being drawn. Does this sound right?
     BTW: It was based on this limited result that I was able to
     test d.erase and d.zoom

2. d.vect, d.sites and d.legend all hang up. They had to be killed
from a separate window.

3. Win98 did not work. The IPC code by itself tested OK, but the
d.mon commands were hanging up. WinNT was used for the limited
successes above.

I will keep looking at the IPC code and do the minor changes I
mentioned above. I also plan to keep on looking into the d.rast
problems to see if I can get this working. Suggestions are welcome.

Malcolm

----------------------------------------
If you want to unsubscribe from GRASS Development Team mailing list write to:
minordomo@geog.uni-hannover.de with
subject 'unsubscribe grass5'

rshepard · January 5, 2001, 12:27am

On Thu, 4 Jan 2001, Malcolm Blue wrote:

Oops, I sent that as HTML mail, which some readers may not like.
Sorry.

Not so. It didn't come through that way in pine, and that's _very_
sensitive to html mail.

Rich

Dr. Richard B. Shepard, President

                       Applied Ecosystem Services, Inc. (TM)
              Making environmentally-responsible mining happen. (SM)
                       --------------------------------
            2404 SW 22nd Street | Troutdale, OR 97060-1247 | U.S.A.
+ 1 503-667-4517 (voice) | + 1 503-667-8863 (fax) | rshepard@appl-ecosys.com

----------------------------------------
If you want to unsubscribe from GRASS Development Team mailing list write to:
minordomo@geog.uni-hannover.de with
subject 'unsubscribe grass5'

Eric_G_Miller · January 5, 2001, 2:47am

On Thu, Jan 04, 2001 at 04:27:08PM -0800, Rich Shepard wrote:

On Thu, 4 Jan 2001, Malcolm Blue wrote:

> Oops, I sent that as HTML mail, which some readers may not like.
> Sorry.

Not so. It didn't come through that way in pine, and that's _very_
sensitive to html mail.

Well, it sure as heck shows up as html here.

--
Eric G. Miller <egm2@jps.net>

----------------------------------------
If you want to unsubscribe from GRASS Development Team mailing list write to:
minordomo@geog.uni-hannover.de with
subject 'unsubscribe grass5'

neteler · January 5, 2001, 9:45am

Hey Malcolm,

this really looks like a break-through!! Congrats for this first
approach to get the XDRIVER running.
I keep the fingers crossed for the next time,

Markus

On Thu, Jan 04, 2001 at 07:41:27PM -0400, Malcolm Blue wrote:

Oops, I sent that as HTML mail, which some readers may not like.
Sorry.

Malcolm Blue wrote:

> Good news and bad news...
>
> GOOD (for WinNT users):
>
> 1. Looks like I'm over the major hurdles in the IPC code. I have
> several changes to send to the maintainer, but I am able to get the
> IPC Message Queues to connect/send/receive/disconnect. (Still more to
> do: clean up code a bit, increase the number of queues and increase
> message buffer sizes look like the first priority, but these should be
> simple changes.)
>
>
> 2. d.mon programs (start, select, stop) are working for me, as well
> as a VERY limited set of display commands on WinNT. I still have some
> minor clean up and testing to do to make sure they are ready for
> testing by others.
>
> 3. d.erase and d.zoom seem to work. d.erase erased the little that
> drew with d.rast, and d.zoom did a proper zoom for the little band of
> pixels that d.rast created (see below), with a cursor, the zoom
> "rubber band" box and the coordinates and prompts in the command
> window were in sync with the graphics window operations. Probably a
> number of other commands will work as well.
>
> In order to run "d.mon select" without code changes, I had to remove
> the "#ifndef _CYGWIN_" and "#endif" statements that were in the
> switch statement in
> src/display/devices/XDRIVER/XDRIVER24/ipc.new/switcher.c , i.e.
> sandwiching the FONT and TEXT case blocks. Does anyone know why these
> cases were being excluded? Is this a relic from the old XFree port?
> These blocks of code seem to compile fine, and the "case FONT" code
> block works with the recent XFree release for Cygwin. I think this
> should be changed for both the original (fifo) version and the ipc
> version.
>
>
> BAD:
>
> 1. Limited success with the d.rast command. It only seems to draw the
> NULL's. Using the Spearfish sample data, I did:
>
>
> d.rast soils bg=yellow - it draws a yellow band near the
> bottom of the X0 monitor window (several pixels wide and the
> width of the window).
> d.rast soils bg=white - it draws a white band where the
> yellow band was.
> d.rast streams bg=red - it draws a red band over the
> bottom part of the white band
>
> From these and additional tests I am guessing that just the
> NULL pixels are being drawn. Does this sound right?
> BTW: It was based on this limited result that I was able to
> test d.erase and d.zoom
>
> 2. d.vect, d.sites and d.legend all hang up. They had to be killed
> from a separate window.
>
> 3. Win98 did not work. The IPC code by itself tested OK, but the
> d.mon commands were hanging up. WinNT was used for the limited
> successes above.
>
>
>
> I will keep looking at the IPC code and do the minor changes I
> mentioned above. I also plan to keep on looking into the d.rast
> problems to see if I can get this working. Suggestions are welcome.
>
>
> Malcolm
>
>
>
>

----------------------------------------
If you want to unsubscribe from GRASS Development Team mailing list write to:
minordomo@geog.uni-hannover.de with
subject 'unsubscribe grass5'

--
Markus Neteler * University of Hannover
Institute of Physical Geography and Landscape Ecology
Schneiderberg 50 * D-30167 Hannover * Germany
Tel: ++49-(0)511-762-4494 Fax: -3984

----------------------------------------
If you want to unsubscribe from GRASS Development Team mailing list write to:
minordomo@geog.uni-hannover.de with
subject 'unsubscribe grass5'