[GRASS-dev] Re: r39856 - Do not fail to start gis.m in XY location due to missing proj info

[Hamish]

> 2009/11/30, Hamish wrote:
> > wrt r39856,
> >
> > a few lines above the new work-around XY locations are specifically
> > tested for & dealt with. the `break` should ensure that XY loc'ns
> > never reach the patched code.
> >
> > is the real problem that the "XY location (unprojected)" text has
> > been translated so the match no longer works?
> >
> > if so, and it is ok to have that be translated (I'm not sure it
> > is), should we change it to run g.region -p and check that the
> > projection code == 0 as a steadier target?
> >
> > or is the XY compare otherwise busted?

Maris wrote:
> in short - Your analysis of code is wrong. Break interupts
> while cycle (@478) and thus $prj is unset but used @490.


ah, you are exactly right. my apologies. I had remembered testing that
XY locations worked with it a while back, and thought they still would
be & was hypothesizing as to why that might be. but it got broken again
since that time for other reasons fixed by your patch.

fix backported to 6.4.


Hamish