[GRASS5] NULLs+GRASS5

[Helena Mitasova - staff]

>I'm seriously considering reverting to version
>4.x, as the ease of use greatly surpasses the additional functionality
>of version 5.0 for me. Will these issues be addressed in the "stable"
>release, or are they inherent to the product?

Malcolm, I can understand your troubles when switching from
integers only/no NULLs GRASS to GRASS5, especially if you
have lots of old data with masks and you are used to have data
with no zero values. While it was easy to use for maps
which used only few integer/no zero categories it was a 
nightmare when you needed to work with real values which could
be zeroes or anything else. So if you really plan to work with
GRASS in the "integer style" mode it may be easier for you 
to use the old GRASS, however, if you plan to move ahead and
do some more complex modeling and visualization I suggest
that you use GRASS5 and help us to identify where NULLs do
not work so that it can be fixed. While floating point seems to
be working well, NULLs haven't been tested as much  
so it would be great if you could identify with
as much detail as possible what is wrong and send it to Markus -
and even better - if you have any programmers around, fix it and
submit the fix to Markus. As far as I know, nobody is working on
NULLs - is that right? 
So far, when I had a null related problem I ran r.support and
set the null file to whatever I need it to do - I believe that there 
are 2 options (create a NULL file or consider zeroes as NULLs).
There also should be a program (r.null?) which you can run in a script
(there may even be that script available somewhere) which allows you
to update your entire mapset to support NULLs properly (as considering
zeroes as NULLs as default isn't such a great idea, but that is 
what the users of GRASS voted for many years ago).
So I would suggest - take advantage of the floating point and
all the other new stuff and you will see that many things which 
were tedious to do are much easier now. And yes, there are still
bugs and unfinished programs but hopefully the stable release will
have most of them fixed, and whatever cannot be fixed in a reasonable
time, as I understand it,  will go into the development version.

I am sending this to the developers list because I am concerned
that NULLs were not tested adequatelly and it would be useful to
get some feedback from others

Helena


---------------------------------------- 
If you want to unsubscribe from GRASS Development Team mailing list write to:
minordomo at geog.uni-hannover.de with
subject 'unsubscribe grass5'