converting from Grass 3 to 4.0

[Michael Shapiro]

|
|
|----- Begin Included Message -----
|
||	Derase(D_translate_color("black"));
||	sprintf(name,"%s.pr",border[found].name);
||	setmap = G_find_cell(name,"");
||
|
|This code look ok, so the problem must be happening somewhere else 
|(unless G_find_cell() is the culprit). Can you include more of the
|code. Also "border" is a suspicious variable name. I recall having
|to modify some of the imagery programs to not use this name since it 
|conflicted with some system include files which also used "border".
|(I don't recall the specific details).
|
|
|Michael
|
|
|----- End Included Message -----
|
|The issue of the variable 'border' is that new to 4.0?
|Because this code worked in 3.

Yes, it is new to 4.0. It didn't occur on Sun machines, but on other
platforms.  It isn't so much GRASS as it is that something out there in
Unix changed between releases. But perhaps it isn't the problem in your case.

|
|	If the culprit is G_find_cell(), I will check this afternoon,
| what is different about it. I am programming on the same machine,
| same location and mapset same user and group....
| what else could be different... I have gone back and run the 3 version after
| the 4 has died? It seems odd. Can any of us take a look at the
| 4.0 programmers manual before it is released?
|

The programmer's manual isn't ready yet. I have been dragging my heels
on it.  However, I don't think the 4.0 programmer's manual won't help
you here.  These calls haven't changed, so the 3.0 manual is ok. The
major changes to 4.0 are to the raster color tables, Colors structure
and related subroutine interface, as well as a fairly major rework of
the vector library calls. Can you email me a copy of your code so I can
see if (1) the bug can be reporduced here at CERL and if so (2) see if
I can spot what's wrong.



Michael Shapiro