[GRASS5] color table

Markus Neteler neteler at itc.it
Mon Sep 8 08:27:07 EDT 2003 

On Thu, Sep 04, 2003 at 10:04:54PM +0100, Glynn Clements wrote:
Content-Description: message body and .signature
> 
> Markus Neteler wrote:
> 
> > > > If possible, I would like to release 5.0.3 tomorrow or
> > > > Thursday latest (then I am out of town for more than a week).
> > > 
> > > I'd like more time to look at the changes. Note that nobody has posted
> > > a comprehensive summary of what's actually changed. I'm fairly sure
> > > that some of those changes don't belong in the release branch (e.g. 
> > > Mike's MinGW stuff in src/general/init/chk_dbase.c, possibly
> > > elsewhere).
> > 
> > No problem to wait. I would appreciate if the changes get reviewed.
> 
> A few more issues:
> 
> 1. src/tcltkgrass/module/d.grid has been updated for the bordercolor=
> option and -b/-g flags, but these aren't in the release version of
> d.grid itself. Which one is wrong?

the update was wrong - reverted.
 
> 2. The src/tcltkgrass/module/r.proj update introduced a typo:
> 
>     {checkbox -n {Do not preform region cropping optimization} "" -n}
> I've fixed this one in both head and release branch.

ok
 
> 3. The signal handling changes in
> 
> 	src/general/init/grass.src
> 	src/general/init/init.sh
> 
> prevent trapping SIGKILL (9) and SIGTERM (15). Not trapping SIGKILL is
> sensible enough (IIRC, this will fail on some shells), but why avoid
> trapping SIGTERM?

Seems that I didn't recall correctly the outcome of that discussion.
Readded 15 both in HEAD and release_branch.

> 4. Some of the changes seem rather extensive for something that's
> supposed to be a bug-fix release. Specifically: d.barscale, v.in.dgn,
> m.in.e00, r.los, r.spread.

Concerning r.los:
It was the "trivial" update to FP done by me (based on your
recommendations). Should be ok

> I've attached a summary of the changes.
> 
> -- 
> Glynn Clements <glynn.clements at virgin.net>
> 

Content-Description: Changes between 5.0.2 and release branch
[...]
> src/display/d.barscale/cmd/draw_scale.c
> src/display/d.barscale/cmd/main.c
> src/display/d.barscale/cmd/mouse.c
> src/display/d.barscale/cmd/options.h
> src/tcltkgrass/module/d.barscale
> 	Upgrade

The old version was partially non-functional. It's rather impossible
(for me) to take out only the bug-fix part without added the
small improvement (-t flag).

[...]
> src/general/init/grass.src
> 	Don't trap SIGKILL, SIGTERM

SIGTERM trapped now
 
> src/general/init/init.sh
> 	Don't trap SIGKILL, SIGTERM
SIGTERM trapped now

> 	Add version number
> 	Fix for first-time use
> 	Change test -e -> test -r
 
[...]
> src/libes/proj/do_proj_nad.c
> 	?

[see Paul's email]
 
[...]
> src/mapdev/v.in.dgn/Gmakefile
> src/mapdev/v.in.dgn/README
> src/mapdev/v.in.dgn/main.c
> 	?

Minor include file fixes and
2003-01-07 11:47  radim
        * main.c: stroke arc and curve, read mslink of given entity number

I don't see a problem here for 5.0.3.
 
[...]
> src/mapdev/v.out.e00/v.out.e00.c
> 	???

2003-03-31 09:13  eric

        * v.out.e00.c: Fix vector/mapset lookup.  Add some type casts and
        extra parens to make -Wall happy (oh, and the type casts might be
        necessary on some arches).

Should be fine.
 
[...]
> src/misc/m.in.e00/info.c
> 	???
2002-07-19 22:51  roger
        * info.c, main.c: Added the list= parameter and corrected a format
        that was invalid for large INFO tables

Sounds ok.
 
[...]
> src/raster/r.in.gdal/main.c
> 	???

Relevant are:

2003-07-08 15:38  markus
        * main.c: change to channel numbers instead of color name
        extensions for multiband data in case subsequent channels have
        identical color names (as for SAA AVHRR data)
2003-04-30 14:50  markus
        * main.c: submitted compression change from Eric Miller. Added grey
        color table for GDT_UInt16
2003-03-13 13:49  markus
        * main.c: grey_scale_colors for Byte data, color name support for
        known color names (Red, Green, Blue)
2002-11-07 17:59  markus
        * main.c: use positive xy coordinates for unprojected TIFF instead
        of negative xy coordinates

Here should be checked that the PROJ4 related updates did not reach
the release_branch. AFAIK Paul already verified it.

Maybe some release_branch testers could try r.in.gdal.

> -src/raster/wildfire/src/r.spread/prob_invsqr.c
> +src/raster/wildfire/src/r.spread/local_proto.h
> src/raster/wildfire/src/r.spread/cell_ptrHa.h
> src/raster/wildfire/src/r.spread/cmd_line.h
> src/raster/wildfire/src/r.spread/collect_ori.c
> src/raster/wildfire/src/r.spread/costHa.h
> src/raster/wildfire/src/r.spread/deleteHa.c
> src/raster/wildfire/src/r.spread/display.c
> src/raster/wildfire/src/r.spread/fixHa.c
> src/raster/wildfire/src/r.spread/get_minHa.c
> src/raster/wildfire/src/r.spread/insert2Ha.c
> src/raster/wildfire/src/r.spread/insertHa.c
> src/raster/wildfire/src/r.spread/main.c
> src/raster/wildfire/src/r.spread/pick_dist.c
> src/raster/wildfire/src/r.spread/pick_ignite.c
> src/raster/wildfire/src/r.spread/ram2out.c
> src/raster/wildfire/src/r.spread/replaceHa.c
> src/raster/wildfire/src/r.spread/select_linksB.c
> src/raster/wildfire/src/r.spread/spot.c
> src/raster/wildfire/src/r.spread/spread.c
> 	[Tidying]
> 	Protect headers against repeated inclusion
> 	Move definitions from header files to source files
> 	Use ANSI prototypes
> 	Add local_proto.h
> 	Remove prob_invsqr.c

Motivation:
 http://grass.itc.it/pipermail/grass5/2003-March/005069.html
The old versions seemed not to work, so the fix should go
into 5.0.3.
 
> src/scripts/contrib/d.rast.labels/d.rast.labels
> 	Don't trap SIGKILL
> 
> src/scripts/contrib/d.rast.leg/d.rast.leg
> 	???

minor cosmetics, no problem I think.

[...] 
> src/scripts/contrib/v.line2area/v.area2line
> src/scripts/contrib/v.line2area/v.line2area
> 	Use -eq instead of = (why?)

was crashing recently - I don't know why, but the fix helped.

[...]
> src/tcltkgrass/module/d.grid
> 	*** BOGUS ***
> 	Add bordercolor=, -b, -g

reverted.
 
Also done now:
- update date in 
  src/CMD/VERSION   August -> September
  NEWS.html         August -> September

If there are no objections, I'll package a 5.0.3RC3 today
and produce the md5 checksum as well.

Markus

More information about the grass-dev mailing list