3. Added a "cross" option to the "grid" instruction. This will draw
grid intersection crosses at the user defined size, instead of grid
lines. The default is to draw grid lines.
very nice to have... I think the ps.map output is really looking
professional at this point.
4. [...] I also
changed the behaviour of the float colorbar to draw black lines
across
the bar at the annotated position. The lines previously seemed to be
going in the other direction, as oversized tics. I was not sure what
the prefered appearance was?
This looked wrong to me as well. This is the change I used to make it
look a bit better:
ps.map/cmd/ps_fclrtbl.c ~line 133
x1 = l + width;
- x2 = x1 + 0.5 * width;
+ x2 = x1 + 0.37 * width;
I understood the original code to mean the width of the tick mark was
purposefully designed to be half the width of the legend box; not a
directional bug.
I haven't tried with your change yet, or am able to throw up a
screenshot of .5 vs .37 vs your patch for people to look at as I'm out
of town right now. Maybe soon.
(the .37 came from playing around with it until it looked right)
5. Updated the page size routines to correctly change the size if the
"rotate" flag is used. The map scale was unaffected if the rotate
flag
was used.
Just curious for more details on this. Is this a bounding box fix to
Radim's new EPS page size code for 5.1 or the does it have to do with
the -r flag as exists in 5.0.2? For correct auto-scaling when 'scale'
isn't defined? I couldn't figure out which of the CVS changes this was.
I've used the 5.0.2 ps.map quite a bit both with and without the -r
flag, never noticed a serious problem.
?
regards,
Hamish
ps - missing bounding boxes is still an outstanding bug in 5.0.x, not
sure if Radim's new EPS code covers this or not for 5.1.
__________________________________
Do you Yahoo!?
SBC Yahoo! DSL - Now only $29.95 per month!
http://sbc.yahoo.com