@currawong ... (i) Apple has about a dozen crash report logs from me, all correlated with a unique string that I included in my problem report. While we can take stabs in the dark by looking at the info that the crash logger is sending to the mothership, and we can poke around in the .diag files in the DiagnosticReports directory, only Apple staff have the full context (ie source code) and tools to do a proper analysis (once escalated from Support into the core Engineering team). If you can share what we should look for, please do. Recall, the crash reporter does not always appear -- about 50/50 for me.
FYI, I've changed my uptime logging script to capture the first ten lines of 'top' in an attempt to get more info on the state of the machine before it hangs/crashes. Because it sure looks like something is leaking memory quite badly (not the screensaver; I turned that off long ago). Ok, maybe not a memory leak, but something is bogging the machine down over time. Here's my (hacky but might be helpful) script:
#!/bin/zsh
cd /Users/KTyyz
echo >> uptimelog.txt
echo "----- ----- ----- ---- ----- ----- ----- ----- -----" >> uptimelog.txt
echo >> uptimelog.txt
while true; do
# uptime >> uptimelog.txt
top -l 1 | head -10 >> uptimelog.txt
echo >> uptimelog.txt
sleep 60
done
Before I switched from a simple uptime to top, it logged this just before a crash/hang:
17:20 up 40 mins, 4 users, load averages: 1.83 1.77 1.86
17:21 up 41 mins, 4 users, load averages: 1.76 1.79 1.86
17:22 up 42 mins, 4 users, load averages: 1.81 1.80 1.85
17:23 up 43 mins, 4 users, load averages: 1.55 1.71 1.81
17:24 up 44 mins, 4 users, load averages: 1.62 1.72 1.81
17:33 up 53 mins, 4 users, load averages: 1.75 1.70 1.80
17:55 up 1:15, 4 users, load averages: 2.19 1.84 1.84
18:04 up 1:25, 4 users, load averages: 3.49 2.27 2.00
9 then 12 then 9-10 minutes to execute "sleep 60" ??? Something's "not right". This script was the only user-level process happening at the time.