- Working with Kevin and others to resolve Qt issues.
- Submitted PMR with IBM regarding usage of Purify and why it's taking so long to run
ucmwb issues
There are a number of issues WRT ucmwb running on Linux:
- Whether to build 32-bit or 64-bit. 32-bit would be LCD yet I don't have a 32-bit Linux machine to build on
- Differences in architectures (i.e. Intel vs. Opteron). I've seen differences WRT this...
- Whether to build static or shared. Static would be better because then we wouldn't have to worry about getting the right shared libraries and the correct LD_LIBRARY_PATH setting
- Connection to Oracle. In order for ucmwb to connect to Oracle directly it needs OCI. In order to have OCI one needs the commercial version of Qt. Got that. But also we need an Oracle Client (Development?) Installation in order to have the appropriate .h and .a files to build. With Oracle on Linux (which doesn't exist) we also need to address the appropriate 64 vs 32 bit'ness, Intel vs. Opteron, static vs. shared WRT Oracle itself...
Purify
Only a little is known at this time but it seems some development group is concerned because running Purify on 7500 had not finished in 10 days of running. Opened IBM/Rational PMR on this. IBM/Rational tells me that using Purify can slow the process by 15-20 times! Makes sense since the instrumented code is doing so much additional work. Need to find out how long the tests would take sans Purify