I would like to initiate a discussion about converting GRASS code
to ANSI-C. Should we convert GRASS to ANSI-C? How soon? Is it
even worthwhile spending precious resources to do this? Will
users be able to compile ANSI-C?
Michael
I would like to initiate a discussion about converting GRASS code
to ANSI-C. Should we convert GRASS to ANSI-C? How soon? Is it
even worthwhile spending precious resources to do this? Will
users be able to compile ANSI-C?
Michael
I would like to initiate a discussion about converting GRASS code
to ANSI-C. Should we convert GRASS to ANSI-C? How soon? Is it
even worthwhile spending precious resources to do this? Will
users be able to compile ANSI-C?
TO CONVERT, OR NOT TO CONVERT?
So, what are the advantages to making the conversion? Well, it's
gets GRASS up-to-date. Keeping software (and hardware) updated has
several advantages (should we go into this?).
HOW SOON?
When is 4.1 tentatively scheduled? Is a "4.2" planned? I'm not sure
how close 4.1 is, but I would suggest doing it immediately after
this version. I realize that there are other priorities w.r.t. GRASS
development, but I think we should just take a "time-out" here and
make the conversion.
RESOURCES:
Conversion shouldn't be too big of a problem since there are several proto
converters out there. I've converted statistical and numerical analysis
libraries in a matter of minutes. Of course, testing this out would take
a while (especially for all of the different machines). Is it worthwhile?
Well, it must be done sooner or later, right?
WILL IT WORK FOR USERS:
Will users be able to compile ANSI-C?
Almost everyone on this list has anonymous ftp access,
so there should be little problem getting gcc if they're in a situation
where they would have to purchase an ANSI compiler separate from the
OS (e.g., Sun and 'acc').
If they don't have anonymous ftp access, then they probably get GRASS
from a vendor. If this is the situation, then either
1) vendors can sell pre--compiled versions (and users don't need a
compiler)
2) users can afford software from vendors (and thus should be able
to buy a C compiler), or
3) users can get a copy of GCC through the mail. I don't have any
info on this, but I know that I've seen places that redistribute
emacs, gcc, and other sources.
I guess the biggest question is timing, but I don't know enough about
the goals of OGI to make an informed decision (how pressing are other
things?).
--Darrell McCauley, Purdue Univ