Re: Compiling on RedHat 7.1

[pauljohn]

address@hidden wrote:
> 
> > >>>>> "PJ" == jones peter <address@hidden> writes:
> >
> > PJ> I am attempting to compile the May 8, 2001 test version
> > PJ> of swarm on RedHat 7.1, on a 686. After ./configure and make, I
> > PJ> get the following messages at the end of the make:
> >
> > PJ> activity.h:530:2: invalid preprocessing directive
> >
> > GCC >= 2.96 are not supported for building Swarm.  (They aren't even
> > official releases of GCC.)
> >
> RedHat seems committed to supporting their 2.96. 
Yes, they have fixed some things already in response to bug reports I've
filed. If you file a bug report, the incredible Jakub will likely
answer. Sometimes, he answers before your bug even shows up in their
system.

The last gcc upgrade for RH7.0 fixed all the problems with compiling
swarm, but from what you say they seem to have reintroduced some in RH
7.1. On my RH7.1 system, I'm not able to build rpms now because of
corruption in my rpm database and the RH engineers are working on that
(increasingly widespread) problem.  Its caused me no end of frustration,
actually, and I'm really wishing they would just pick one database
format and stick to it. 

I think your choice is justified, except perhaps not for the reason you
give. The thing about "binary incompatability" is, as far as I know,
mostly confined to c++  programs, where the libraries seem to be in a
constant state of flux.  From what I'm told, programs in C compiled with
gcc-2.95.3 do run on other systems, as you have found with my swarm
RPM.  There's no point in sticking with RedHat's 2.96 on the grounds of
compatability with future RedHat releases because those future RedHat
releases will not be compatable with their current gcc.  See what I
mean? When they get to gcc 3.0, the stuff that is binary-incompatible
with gcc-2.95.3 will also be incompatible with gcc-2.96.  That's part of
the reason the GCC people were pissed off at RedHat, they created an
illusion of continuity.

RedHat's answer seems to be, in essence, "we want to speed development
of GCC so we are distributing this and using our distribution as a
launch pad for bug testing." Don't forget RedHat now owns Cygnus, where
the development on GCC has been since about 2 years ago.  The only
feasible response from the GNU/GCC people is either to complain (that's
the option they have taken) or revoke their blessing from the
Cygnus/egcs team (which they did not take yet, as far as I know).

Of course, all this is filtered through my limited attachment to
reality. 
-- 
Paul E. Johnson                       email: address@hidden
Dept. of Political Science            http://lark.cc.ukans.edu/~pauljohn
University of Kansas                  Office: (785) 864-9086
Lawrence, Kansas 66045                FAX: (785) 864-5700

                  ==================================
   Swarm-Support is for discussion of the technical details of the day
   to day usage of Swarm.  For list administration needs (esp.
   [un]subscribing), please send a message to <address@hidden>
   with "help" in the body of the message.