[GRASS5] Bug: vector maps shifted (d.area / d.vect)
hmitaso at unity.ncsu.edu
Sun Oct 21 11:49:23 EDT 2001
d.vect should be left alone for GRASS5.0 release - it is a core program
that is being used all the time and it does not have a problem.
d.area seems to me of less importance (I have just tried it out, it is nice,
but I must admit that I did not even know that it existed) because
what it does can be accomplished to certain extent by v.to.rast, d.rast -o,
or is there anything else that d.area does that is irreplaceable and crucial
for work with GRASS? I fully agree with Bernhard that you don't want to change
something that affects lots of code at this point, especially not for a
that is not being used too often (and that is probably why the bug has not
long time ago).
Markus Neteler wrote:
> On Sun, Oct 21, 2001 at 03:58:03PM +0200, Bernhard Reiter wrote:
> > On Sun, Oct 21, 2001 at 01:55:33PM +0200, Markus Neteler wrote:
> > > On Sat, Oct 20, 2001 at 12:44:41PM +0100, Glynn Clements wrote:
> > > > If you remove the +0.5 from fastline(), d.vect and d.area coincide.
> > > > However, G_plot_line() is used in many places, so changing it to suit
> > > > d.vect may break other things.
> > >
> > > Thanks for looking into it. But I wouldn't call it "breaking things":
> > > For 5.0 the display needs to be consistent in my opinion, we should
> > > not accept map shifts. I am willing to search-and-fix, but need a
> > > hint what route to follow.
> > >
> > > Raster and vector overlays should be 100% o.k.
> > Nobody will say no to this.
> > On the other hand Glynn's comment is correct:
> > Changing things in a core function which is used at various places
> > is not a change you want todo between a pre2 and a final release.
> > If this inconsistency was there for a long time other code might
> > work around it in places you cannot easily see. Fixing it in the
> > core fuction is very likely to break things at other places.
> > Therefore you need to test it intensively.
> > Not knowing the details of this case I recommend that adding a
> > workaround to d.vect for the 5.0.0 release should be considered.
> > Bernhard
> ... maybe yes - of course we don't want to break the system now.
> - d.vect is right
> - d.area, v.area are wrong
> according to d.rast (and my eyes).
> So we have to careful where to apply the work-around.
> grass5 mailing list
> grass5 at grass.itc.it
More information about the grass-dev