Re: FW: Swarm2.1.1 source code compiling error (explow.c) on Linux Re d Hat 7.2

[Paul E Johnson]

Dear Tina:

If I debug your program, are you going to have ChevronTexaco make a 
contribution to the Swarm Development Group?  If you are asking me to do 
your assignment, don't you suppose there should be some quid pro quo?

The dynamicGraph app you sent me does not have an author or a date, and 
I can see it is basically a combination of the DynamicGraph program that 
runs against the GraphLib for Swarm that is available elsewhere.  If it 
is really important to run that particular program, we can figure it 
out, I'm sure, or I can tell you how to do it correctly with the Graph 
Library.  I've looked at it a while and I think I see what is wrong, but 
before I go to the trouble of fixing it, I have to know if it is really 
really important to run that particular program, because it does not run 
AS IS on my system. I don't believe this program ever did run, actually, 
so the problem for me is not well formulated.

IN the Swarm FTP there is a a working copy of the original DynamicGraph 
program here:

ftp://ftp.swarm.org/pub/swarm/apps/objc/contrib/DynamicGraph-2.1.tar.gz

I've just tested it and it does run against Swarm-2.1.140 with the 
GraphLib installed in my system.

The problem with the dynamicGraph you sent me is that the GraphLib has 
been hacked and stuck into the middle of the program, and its code is 
not all completely correct.  For example, if you are using the graph 
library from your current directory, the include statements should refer 
to "graph.h" and not <graph/graph.h>. See?

I've just verified that I can rebuild/install the Graph Lib I 
distributed a while ago on a current system:

http://lark.cc.ku.edu/~pauljohn/Swarm/miscSupportSoftware/graphLib-20010423-1.i386.rpm
http://lark.cc.ku.edu/~pauljohn/Swarm/miscSupportSoftware/graphLib-20010423-1.src.rpm

I suggest to test your Swarm install with the "approved" Swarm apps in 
this package:

ftp://ftp.swarm.org/pub/swarm/apps/objc/sdg/swarmapps-2.1.1.tar.gz

These are more or less guaranteed to go, mostly, on a 2.2 Swarm. At 
least, the problems are confined to a few syntax changes.

Now, if we really need to fix your program, I've hacked on it for 45 
minutes to get a clean compile without warnings and then I get a runtime 
crash:

$ ./dynamicGraph
after create swarm.
after create buildObjects.
after create DiGraph.
before create PMMLCGgen.
*** event raised for error: InvalidArgument
*** function: _i_Object_s__doesNotRecognize_(), file: 
/home/pauljohn/LinuxDownloads/redhat/BUILD/swarm-2.1.140.20020514/src/defobj/DefObject.m, 
line: 827
DiGraph does not recognize allocIVars:
*** execution terminating due to error
/home/pauljohn/LinuxDownloads/redhat/BUILD/swarm-2.1.140.20020514/src/defobj/Symbol.m:187 
-[Error(c) _raiseEvent:]
Aborted

That happens because you try to use "self" as a zone for objects that 
are not Swarms.  Code like this is wrong:

uRandPosition = [UniformInteger createBegin: self];

After fixing those, I get this error

*** event raised for error: InvalidCombination
PMMLCG initState: superclass method invoked - Yell at Sven!
*** execution terminating due to error
/home/pauljohn/LinuxDownloads/redhat/BUILD/swarm-2.1.140.20020514/src/defobj/Symbol.m:187 
-[Error(c) _raiseEvent:]


I don't find it necessary to Yell at Sven, just change the syntax of the 
line that creates the generator to:

randGPosition = [PMMLCGgen create: [self getZone] setStateFromSeed: seed];

Then I get this:

$ ./dynamicGraph
after create swarm.
after create buildObjects.
after create DiGraph.
before create PMMLCGgen.
before create randGPosition.
after create randGPosition.
error: PMMLCGgen (class)
PMMLCGgen does not recognize getUnsignedMax
Aborted

This one has me puzzled at the moment. getUnsignedMax is in the 
documentation and the source code of Swarm.  I would think this error 
indicates something wrong with Swarm itself, except for the fact that I 
know the original DynamicGraph and GraphLib programs do run.


So, as you can see, this program requires a lot if fixing, and I'm tired 
of doing it.

If you are just using it to test your Swarm installation, there are lots 
better ways.

--
Paul E. Johnson                       email: address@hidden
Dept. of Political Science            http://lark.cc.ku.edu/~pauljohn
1541 Lilac Lane, Rm 504
University of Kansas                  Office: (785) 864-9086
Lawrence, Kansas 66044-3177           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.