|Downloaded the 4.44a The file appears not to include lattices (no big deal - I downloaded them) but no matter what I do, the client is crashing. This is now a clean install - and still doing the same thing so I'm dead in the water. Am I not holding my mouth right or something?|
|Did you download the full version or just the upgrade version? Just a thought...|
|Tried 'em both. I don't know if the download file changed or what. It doesn't appear to be complete since the lattices are missing - and I'm not sure what else. Whenever it tries to download anything - it craps itself.|
|IIRC the lattices are not included ever. thus when you start a fresh install the program downloads them from the server. Are you running the graphical or commandline version? Is there a point when it stalls or does it just not start up?|
|hmmm... interesting. 444a includes old lattices - but no matter. I don't think the lattices are the issue.|
Tried running both in command line, and background with the same result.
The client starts - then attempts to download samplefiles / and/or lattices and the the MSFT error report dialog pops up whilst the client dies an agonizing death.
Until resolved - I've switched back to 443d and running PhaseRotEby1 so that at least I have some production.
This behavior manifested itself on two different machines in exactly the same way.
|4.44a dies occasionally also for me, always at the end of a simulation. I'm currently investigating, so far it seems to be related to autosave.|
|It looks like v4.44a is not running on older processors. It works fine on my Pentium 4 and Athlon 64, but does not work on my Pentiun III and Athlon (1GHz) processor.|
|I also just found that the -2%/0Mpts bug that I thought I'd fixed has come back in 4.44a, so I'll have another go at fixing it on Monday. Seems to do it after a certain number (11?) of simulations on my machines.|
Machine #1 is an old AMD T-bird 1.2GHz. Machine #2 is a PIII. Therein may lie the rub. I've got another newer machine - I guess I could try it out on that one. Perhaps the mystery will be solved.
|This might be the old "optimise" switch on my compiler that I turned on. I remember previously I had to turn it off to get stable operation.|
|Zounds! Do I sense a 4.44b release?|
|Yes, I'm now paying the price for a year of tweaking on my local machine that wasn't put through its paces by a the hundreds of DPAD users! Until I have a credible 4.44b (early next week probably), I'm leaving the old 4.43d on the page to download.|
This does have the odd effect that 4.43d users can only run PhaseRotEb1 (the -a lattices I restricted to 4.44 because the muon scores evaluate slightly differently).