I just had a lengthy journal entry about how LAM 6.3.3b52 is officially dubbed "release candidate 1". Happiness all around.
Unfortunately, I hit ctrl-c at the jjc prompt, and all was lost. Doh. Gotta put something in jjc to prevent that from happening in the future. :-(
Suffice it to say that we're starting the formal LAM release process. I put some way-cool centralized error reporting stuff in the lamtests module (there was a lengthy explanation of it in the Journal Entry That Is Now Deceased; it's too late to re-type it all now), and generally expanded the testing base. This actually resulted in finding a few more bugs and minor memory leaks for obscure cases in LAM (which is a good thing -- yay for testing!).
I will, however, re-print an excerpt from a LAM user that I got today:
"I wrote you a while ago regarding C++ extensions for MPICH. By now we've switched to LAM. Feature availability convinced us to do so... :)"
I replied to her that all Right Thinking people use LAM. Resistance is futile.
Here's a cute one -- kudos to anyone who can decipher it:
10001001101111010000011110011101111111010101000001100110 11001011100101110110001000001100010110010111101001110100 11001011110010010000011011101101111111011101111110000000
It's in Darrell's .sig.
Now that Brian and Ron will run with LAM's release process, I'll head back to poggenc... By the looks of vorbis-dev, Ogg/Vorbis beta 4 is pretty close. There's still some broken things in terms of building in non-gcc/non-Linux/non-shared-library environments, so I'll keep bitching about those. :-)
There are currently 413 xmms instances running on my machine out of a total of 497 processes. 83% of the jobs on queeg are xmms.