PDA

View Full Version : Best RMS in A: 0.000


wirthi
10-23-2002, 10:00 AM
Please, tell me that's no fake. Do we really get a 0.0 for this one? Congrats to stubs25 .. :cheers:
Brian the Fist
10-23-2002, 10:13 AM
No, surely this is a glitch, I'll take care of it and see what happened.
As for the lucky guy, if you're reading this, can you please mail your error.log to trades@mshri.on.ca (if you only have one machine or know which one got the 'lucky' structure? Thanks.
whilden@ars
10-23-2002, 10:35 AM
I have posted this on the Ars Forum as well!

:thumbs:
pointwood
10-23-2002, 10:53 AM
If it wasn't, I would say we should get on with the next protein ;) :D
Halon50
10-23-2002, 01:48 PM
The, er, servers are down. Does this mean the structure was for real? :confused:
wirthi
10-23-2002, 01:53 PM
Originally posted by Halon50
The, er, servers are down. Does this mean the structure was for real? :confused:
No, that was planned long before:
- Just a note that the main server is scheduled for maintenance next Wednesday, Oct. 23 at 11am EDT. This may take several hours and during this time the entire website and data servers will be unavailable. You may leave your clients folding however, and they will upload their data once the server comes back online.
Halon50
10-23-2002, 06:30 PM
Oh yeah, forgot about the maintenance scheduled for today. :rolleyes:
whilden@ars
10-24-2002, 10:06 AM
:cheers:

To MashRinx 0.00

Seems we have a problem! ;)
wirthi
10-24-2002, 10:30 AM
That's just fair. Last time Ars got a 0.00, this time it's FreeDC :D
Too bad that it's an invalid structure again (I guess) :bang:
bwkaz
10-24-2002, 10:36 AM
I can't help recalling Howard's optimism about this structure -- "this one seems to do better than expected" or whatever he said. Indeed it does... ;)

Of course, if that's just because this one breaks the RMS-calculators, then that's a bad thing, but still.
Brian the Fist
10-25-2002, 11:59 AM
Actually, it is interesting because I went back and checked the logs and for the previous proteins a number of structures had NaN (not a number = infinity or sqrt(-1), etc) as the RMSD (which is >99.99 and so wouldn't be noticed by anyone). I know for a fact that the RMSD calculator always works no matter what, so again, it is likely the users submitting these structures have FAULTY RAM. Now yes, I knwo I keep telling people they have bad RAM and they don't want to believe me but 9/10 people Ive told have bad RAM tested it and did! So I strongly reccommend everyone head over to www.memtest86.com which is a superb memory tester and test your RAM right now. It only takes a few hours, and you only have to do it once (until you buy more RAM...) and it could save you months of headaches and problems. I guess with the recent drop in RAM prices has come a drop in quality as well ;( When I have a few minutes I will 'tell' the server to ignore anything with an RMSD of exactly zero so I don't have to keep resetting it by hand..
Darkness Productions
10-25-2002, 03:36 PM
Actually, it only takes about 30-45 minutes on a Duron 800. I ran mine through all the tests, in all memory ranges. 1hr tops. :jester:


Originally posted by Brian the Fist
It only takes a few hours, and you only have to do it once (until you buy more RAM...) and it could save you months of headaches and problems.