Member discussion of optimal flag settings., Add your questions and findings here.

[treadlightly]

Thanks Macaholic, more than I wanted to know, but now I'm glad I know it (er...something like that).

[314159]

Ok, I still am sticking with a "Unofficial Statement" that the advmethod slash late Beta flag isn't helping my production, at the moment, on my "Macs."

More news at 11:00...

Ok, It's a bit past 11 PM, at least here on the East U.S. Coast. Your experience is?

I would recommend that anyone posting in this thread, advocating  flag settings other than as "officially" recommended, indicate very specifically the type of Mac that they are addressing: i.e. - Dual G5 2.0 GHz 1 GB RAM dedicated folder, P.F. 0.966627  etc. and the specific time  (GMT) that the WUs were downloaded from the Stanford Work Server.

Like Ken, I experiment with flags and do, in fact, benefit at times in the short run. Like Ken, I have the advantage of typically being available to check the Stanford server situation immediately prior to completion of a WU, and set flags accordingly (it's still a gamble).

I do this on one of my Macs and keep meticulous records. My observation is simply that over almost any "extended" period of time, my average production is best when using the recommended -local -advmethods -forceasm -verbosity 9 arguments with Big WUs off on my Macs (PC's are a different issue). This situation could easily change in the future - seems like it almost always does.

Note that I do not have a G5 in service. These comments apply to G4 and G3 processors only.

Has anyone completed a 600.0 point p114x on a Mac recently? (anger due to losing the total bonus points vs. pro-rata credit)

[ken]

Listen, I'm on the right coast and your on the left... John I'm a third of the way through a "Protein: p1140_RIBO_FSpeptide_EXT_nospring" on a dual 2.o G5... I'll keep you informed, the other procs are happily munching a way on a varied diet of Gros and Dgrow...


All smooth sailing here.   

[Macaholic]

Did someone say Sail?!  

[ken]

But I left the fishing pole, for the folks with WU hooks.  

[314159]

My earlier question about the p114x's was due to the fact that I have not been able to get past 57% on any of my Macs (using "experimental flags").

Readers should not be alarmed! Proper partial (pro-rata) credit was received.

This on 2 G4's vs. your G5. Please post your results.

Box was a 1.25 GHz G4, 1 GB RAM, dedicated machine, P.F. 0.946627. WU downloaded 3 May 01:53:02 GMT.

Switched back to recommended flags at that point.

Ever feel like you were doing beta testing on Mac WUs without having volunteered to do so?

I think that it's time for a fishing road trip! Any takers?

[treadlightly]

I'd love to, but I've been to the eastern time zone once this month already, and I don't think my employer will give me more time off for another.   I went all the way to Ohio for that trip, of course, no one really knows what time zone Indiana is in anyway.

Meet in the middle and hit Lake Michigan for a day?

[utopia-i]

to recommend or not recommend - my experience with trying to get new volunteeres is that they expect/hope that InCrease will have the correct defaults set when installing for the first time, we do recommend that App first don't we,so they update the username part but thats all they require to do - if we ever agree on what they should be.

Can I also be so brave to volunteer more work for InCrease, sorry if this isn't the correct forum to suggest that, that some very simple setup wizzard would help the 'correct/recommended' settings to be configured during initial install process, eg, is you Mac a G3 or G4 processor then check this box and that applies the 'correct' settings .. hope you get the idea, Its always difficult volunteering extra work to a free and great app but just a thought.

I don't really have the time to go chasing double pointers, special extra point Wus, constantly twiddling with settings I'd rather NOT understand.

I expect Stanford to serve the 'correct' units to me, with 2000+ completed WUs so far I hope it know my setups by now :-) It's to their advantage to get the best out of my setup and not waste their time and my time/money - ie the Science is the winner. If they do make 'mistakes' and we are only human, as long as I see progress then they continue to get my free support.

b happy

[goldis]

John, I have completed two p114x's in the last couple days on a G5.

I thought one EE'ed but I can't find it in the logs, so if it was a 114x it was at least a week ago.  I currently have another couple that are still cooking.

[314159]

treadlightly: Lake Michigan it shall be.

utopia-i: Dropped you a pm. Also, it does appear that the Stanford servers have been tweaked a bit for the benefit of Macs. I do not think that they have solved the problem of a Work Server running out of Mac WUs yet (while still having WUs for Win or Linux).  Weekends, when no one was in the Stanford labs, used to be a real "treat" for folders - particularly Mac folks. I dreaded them. Went so far as to bring down some "spares" early each Friday, just in case. (over a year ago)

goldis: I finally got through one of the current batch of p114x's!  I also checked the EE's a bit closer and found that they were very early RCG's. Thanks for the feedback!