[GRASS5] Objections to release 5.0.0stable?

[Glynn Clements]

Markus Neteler wrote:

> > > > > are there any objections to release 5.0.0stable?
> > > > > I can't see release critical bugs.
> > > > > 
> > > > > If no objections, we should tar the ball,
> > > > 
> > > > First, we should check whether any important fixes still need to be
> > > > merged into the release branch.
> > > 
> > > Yes, certainly (I was silently including the somewhat common procedure,
> > > silently hoping for your assistance :-)
> > 
> > OK. I've analysed the diffs:
> 
> Thanks for this, Glynn!

> > 1. These are entirely harmless documentation fixes.
> [...] 
> I suggest to submit them to the release branch.

All applied.

> > 
> > 2. These are simple fixes which I'm recommending (i.e. they will be
> > merged unless someone objects).

All applied.

> > 3. I need more information on these, but they seem sufficiently minor
> > to be uncontentious.
> > 
> > 	NEWS.html
> > 		Document fixes since pre5
> > 
> > [Some of the changes which are documented here are up for discussion;
> > the NEWS.html which is supplied with 5.0.0 should only mention changes
> > which actually make it into the release.]

Not applied.

> > 	src.garden/grass.postgresql/tutorial/index.html
> > 		Change "destroydb" to "dropdb"
> 
> This is an HTML file and a documentation fix only (AFAIK 'destroydb'
> doesn't exist any more). So: yes.

Not applied.

Presumably this is related to the PostgreSQL version. I have
PostgreSQL 6.5.3, which has "destroydb" but not "dropdb".

We can't know which version the user has, so the documentation should
probably say something like:

	To destroy a database use: "destroydb humus" (PostgreSQL 6.x)
	or "dropdb humus" (PostgreSQL 7.x).

> > 	html/database.html
> > 		add r.to.pg
> > 		remove v.in.shape.pg
> 
> both yes.

Applied.

> > 	src.garden/grass.postgresql/Gmakefile
> > 		Disable v.in.shape.pg
> 
> also yes (see discussion in ML). To keep this will confuse the user.

Applied.

> > 	src.contrib/GMSL/NVIZ2.2/scripts/panel_cutplane.tcl
> > 		??? log message is:
> > 			Bob Covill: fix for z scale range: bug #1202
> 
> From the tester was reported that the fix fixes the problem partially
> (it's better that before). No recommendation.

Applied.

> > 	html/html/d.dm.html
> > 	src/display/d.dm/d.dm
> > 		add PNG feature
> 
> Looks as not to be dangerous. Did someone test that new button?
> But I'm not shure.

Not applied.

The code does:

	d.mon PNG
	<draw maps>
	d.mon stop=PNG

1. If the PNG driver is already in use, this is bad.

2. Should it be re-selecting the previous monitor when it finishes, or
is that taken care of elsewhere?

One problem is that, at present, you can only have one PNG monitor
active at once. Even if you created multiple monitorcap entries, the
name "map.png" (or "map.gif") is hardcoded.

> > 	src.garden/answers/src.answers/raster/Gmakefile
> > 	src.garden/answers/src.answers/raster/r.answers/inter/Gmakefile
> > 	src.garden/answers/src.answers/raster/r.answers/inter/answers_input.c
> > 	src.garden/answers/src.answers/raster/r.answers/inter/answers_run.c
> > 	src.garden/answers/src.answers/raster/r.answers/inter/bmp_menu.c
> > 	src.garden/answers/src.answers/raster/r.answers/inter/copy_project.c
> > 	src.garden/answers/src.answers/raster/r.answers/inter/edit_region.c
> > 	src.garden/answers/src.answers/raster/r.answers/inter/extract.c
> > 	src.garden/answers/src.answers/raster/r.answers/inter/extract_bmp.c
> > 	src.garden/answers/src.answers/raster/r.answers/inter/extract_chnl_slp.c
> > 	src.garden/answers/src.answers/raster/r.answers/inter/extract_rain.c
> > 	src.garden/answers/src.answers/raster/r.answers/inter/extract_tile.c
> > 	src.garden/answers/src.answers/raster/r.answers/inter/get_bmp.c
> > 	src.garden/answers/src.answers/raster/r.answers/inter/get_channel.c
> > 	src.garden/answers/src.answers/raster/r.answers/inter/get_chnl_slp.c
> > 	src.garden/answers/src.answers/raster/r.answers/inter/get_cover.c
> > 	src.garden/answers/src.answers/raster/r.answers/inter/get_elevation.c
> > 	src.garden/answers/src.answers/raster/r.answers/inter/get_mask.c
> > 	src.garden/answers/src.answers/raster/r.answers/inter/get_outlet.c
> > 	src.garden/answers/src.answers/raster/r.answers/inter/get_rain.c
> > 	src.garden/answers/src.answers/raster/r.answers/inter/get_res.c
> > 	src.garden/answers/src.answers/raster/r.answers/inter/get_soils.c
> > 	src.garden/answers/src.answers/raster/r.answers/inter/get_tile.c
> > 	src.garden/answers/src.answers/raster/r.answers/inter/hit_return.c
> > 	src.garden/answers/src.answers/raster/r.answers/inter/hydro_filter.c
> > 	src.garden/answers/src.answers/raster/r.answers/inter/map_sediment.c
> > 	src.garden/answers/src.answers/raster/r.answers/inter/mk_bmp_tbl.c
> > 	src.garden/answers/src.answers/raster/r.answers/inter/mk_cat_tbl.c
> > 	src.garden/answers/src.answers/raster/r.answers/inter/mk_chnl_tbl.c
> > 	src.garden/answers/src.answers/raster/r.answers/inter/mk_cover_tbl.c
> > 	src.garden/answers/src.answers/raster/r.answers/inter/mk_elev.c
> > 	src.garden/answers/src.answers/raster/r.answers/inter/mk_rain.c
> > 	src.garden/answers/src.answers/raster/r.answers/inter/mk_soil_tbl.c
> > 	src.garden/answers/src.answers/raster/r.answers/inter/read_project.c
> > 	src.garden/answers/src.answers/raster/r.answers/inter/remove_files.c
> > 	src.garden/answers/src.answers/raster/r.answers/inter/set_mask.c
> > 	src.garden/answers/src.answers/raster/r.answers/inter/set_region.c
> > 	src.garden/answers/src.answers/raster/r.answers/inter/show_status.c
> > 	src.garden/answers/src.answers/raster/r.answers/inter/step1.c
> > 	src.garden/answers/src.answers/raster/r.answers/inter/trim_region.c
> > 	src.garden/answers/src.answers/raster/r.answers/inter/user_file.c
> > 	src.garden/answers/src.answers/raster/r.slope/cmd/Gmakefile
> > 	src.garden/answers/src.answers/raster/r.slope/inter/Gmakefile
> > 		Appears to be just s/stdout/stderr/ and Gmakefile fixes
> 
> Exactly. Now usable, before not. Tested. So: yes.

Applied.

> > 4. These seem to be more substantial upgrades, so I'm more sceptical
> > about including them.
> > 
> > 	html/html/r.info.html
> > 	src/raster/r.info/cmd/main.c
> > 		HTML fixes
> > 		-r flag
> 
> a really trivial change (just an if/else thing).

Applied.

Although, src/tcltkgrass/module/r.info should probably be updated
accordingly.

> > 	html/html/m.in.e00.html
> > 	src/misc/m.in.e00/info.c
> > 	src/misc/m.in.e00/main.c
> > 		list= option
> > 		Other fixes ???
> 
> no idea.

Not applied, in light of Roger's comments.

> > 	+ html/html/r.to.pg.html
> > 	src/scripts/contrib/r.to.pg/r.to.pg
> > 		Add HTML file
> > 		Support attribute cats
> 
> r.to.pg is a script. Can anyone confirm that it's working?

Not applied.

> > 	src/display/d.vect/cmd/main.c
> > 	src/display/d.vect/cmd/plot2.c
> > 	src/display/d.vect/cmd/plotCat.c
> > 		???
> 
> was a fix to also draw very small areas. Looks important.

Not applied.

> > 	locale/ru/LC_MESSAGES/d.vect.pg.po
> > 	html/html/d.vect.pg.html
> > 	src.garden/grass.postgresql/d.vect.pg/dbvect.h
> > 	src.garden/grass.postgresql/d.vect.pg/getAllOpts.c
> > 	src.garden/grass.postgresql/d.vect.pg/getSopts.c
> > 	src.garden/grass.postgresql/d.vect.pg/main.c
> > 	src.garden/grass.postgresql/d.vect.pg/runInfxFile.c
> > 		-e flag
> 
> no idea.
> 
> > 	locale/ru/LC_MESSAGES/v.to.pg.po
> > 	src.garden/grass.postgresql/v.to.pg/dbvect.h
> > 	src.garden/grass.postgresql/v.to.pg/getAllOpts.c
> > 	src.garden/grass.postgresql/v.to.pg/getSopts.c
> > 	src.garden/grass.postgresql/v.to.pg/main.c
> > 	src.garden/grass.postgresql/v.to.pg/plotCat.c
> > 	src.garden/grass.postgresql/v.to.pg/runInfxFile.c
> > 		-p flag (PostGIS)
> 
> no idea.
> 
> > 	html/html/d.text.freetype.html
> > 	src/display/d.text.freetype/Gmakefile
> > 	src/display/d.text.freetype/freetypecap
> > 	src/display/d.text.freetype/main.c
> > 	+ src/display/d.text.freetype/test.input
> > 	+ src/display/d.text.freetype/vflib
> > 		new options: align=, rotation=, -r, -p, -s, -c
> > 		command mode
> 
> no idea.
> 
> > 	html/html/ps.map.html
> > 	src/ps.map/ps.map/cmd/Gmakefile
> > 	src/ps.map/ps.map/cmd/colortable.h
> > 	src/ps.map/ps.map/cmd/ctablfile.c
> > 	src/ps.map/ps.map/cmd/eps.c
> > 	src/ps.map/ps.map/cmd/local_proto.h
> > 	src/ps.map/ps.map/cmd/ps_area.c
> > 	src/ps.map/ps.map/cmd/ps_map.c
> > 	src/ps.map/ps.map/cmd/read_cfg.c
> > 	src/ps.map/ps.map/cmd/vectfile.c
> > 	src/ps.map/ps.map/cmd/vector.h
> > 	+ src/ps.map/ps.map/cmd/ps_fclrtbl.c
> > 		Add height option to "colortable" command
> > 		Add pat, pwidth, scale options to "vector" command
> this is pattern filling support.
> 
> > 		Support FP colour tables.
> 
> no idea if it could break anything.

None of the above have been applied.

> > 5. The remainder are I18N fixes to key components (libgis, startup,
> > tcltkgrass). If they break anything, they may well break everything,
> > so I'm opposed to including them in the 5.0.0 release.
> 
> I agree as we didn't receive any comments if this is stable or not.

I'm pretty sure that the libgis part is harmless so long as --with-nls
isn't used (it defaults to "no"). If USE_NLS isn't defined in
config.h, then the "_" (underscore) macro is a no-op, i.e.

	#define _(str) (str)

However, the startup and tcltkgrass aspects aren't conditional; the
changes have effect even if --with-nls isn't used.

-- 
Glynn Clements <glynn.clements at virgin.net>