Christian Lynbech on satellite wrote: > Daniel> Requires glibc 2.0? Are you sure? > Unfortunately yes. Trying to run against glibc2.1 makes the program > segfault which I recall as being one of the symptoms of lib mismatches > across those versions (and the recommended linux versions are listed > as R4 or MkLinux).
> Daniel> If so, you should not even be trying to make this run - glibc 2.0 > never > Daniel> really existed on PowerPC, it was a halfbaked port at best. Only 2.1 > Daniel> and soon 2.2 are really an option. R4 and MkLinux were notoriously > Daniel> incompatible with everything else. > Hmm, I was not aware that the situation was that bad. My hopes was > that I could compile a glibc2.0 good enough for the software to run. I'm no expert, but I thought glibc2.1 was supposed to be backward compatible with glibc2.0. I have had problems when I have upgraded my compiler from from 2.7.2 to gcc-2.95.x. The compiler is so good that it optimizes out symbols that the old libraries expect to resolve. The only way (that I know of) to fix this is to recompile the libraries with the new compiler. Brendan Simon.
