[GRASS-dev] grass7 on WinGrass Multi.make deps

[Hamish]

Colin wrote:
> I've been in the field and computerless for the last month

hope you had fun.

> but I'm back on task now. The work around was actually designed by
> Glynn last January at my request, but you seem to be right that it is
> no longer needed as I just had a successful compile without it. After
> looking at the message again it may have been a mistake to include it
> since the real problem was a version issue with MSys' make (3.79.1 vs
> 3.81).
> Either way I'll remove it from the wiki since it is no longer needed.

ok, thanks.

> Thanks for all the work including those bugfixes Hamish. My intention
> by documenting them all was that someone would do that.

They were easy to deal with once found - an associated bug ticket gets
it looked at sooner by folks not interested in windows though :)

> the real goal for me was that if I can compile grass on windows, than
> I can develop/compile my own grass modules for my research, on windows.
> It also had the excellent side benefit of providing wingrass bugfixing.

the best motivation indeed.

 
> On that, the latest trunk update gave the following errors when
> compiling, but otherwise seems to be running fine:
> 
> /src/trunk/scripts/r.colors.stddev
> make: *** No rule to make target
> `/src/trunk/dist.i686-pc-mingw32/scripts/r.colors.stddev.py',
> needed by `script'.  Stop.
> 
> /src/trunk/scripts/v.colors
> /src/trunk/scripts/v.in.garmin
> /src/trunk/scripts/v.in.gpsbabel
> /src/trunk/scripts/v.out.gps
> /src/trunk/visualization/nviz

with the exception of nviz those are all shell scripts which I guess
are my responsibility to port to python, but I haven't found the time to
rewrite them yet. I won't have any time to do anything about that for
another month or so. I was considering not porting v.in.garmin and
redoing v.in.gpsbabel from scratch using v.in.ogr's XML driver. Decision
on what to do with v.in.garmin partly depends on how well the new
v.in.gps python module works. i.e. sometimes gpsbabel can be lossy not
passing on some garmin specific data fields.

grass7 no longer ships the batch file 'sh' launching work-around which
would enable them to run.


Hamish