• We’re currently investigating an issue related to the forum theme and styling that is impacting page layout and visual formatting. The problem has been identified, and we are actively working on a resolution. There is no impact to user data or functionality, this is strictly a front-end display issue. We’ll post an update once the fix has been deployed. Thanks for your patience while we get this sorted.

DF (pointless) Self Congrats!

M

m0ti

Senior member
I just got the 10th best fold for the beta algorithm!

W00t! 😎

Not that means much, considering those stats'll be axed at the end of the beta... Oh well!

BTW, the beta algorithm results are really quite impressive. We're doing the same protein as the one we're just finishing up in "standard" DF. We've done 10.5M folds so far and the best RMS is at 6.42. That's quite a bit lower than the 7.10 for 11+ billion folds in "standard" DF. The whole top 10 for the beta is better than the second best fold in the regular folding!

Mind you, the new algorithm produce folds much much slower; but the quality/time invested value definitely went up.
 
WTG m0ti!

Interesting comparison of the two clients. Really let's us see
just how much is being learned about the folding question.

-Sid
 
Originally posted by: gistech1978

when is the protein going to change today??
Click to expand...

a time was never announced
last time was ~4 pm EST , i am guessing the same, but previously, it was typically about noon EST, so who knows

i am very anxious to get the whole project onto the new algorithm, but i of course want thorough testing

i feel like this last (or two) protein is a bit of a waste in regard to the science , and only a stats race. even though i am a moderate stats whore, i would prefer my cycles go to something meaningful. i am not sure the results from the last few proteins before we switch to the new algorithm are going to do much to advance the project. more like marking time (marching in place)

ho hum

good job on helping the beta m0ti

i am going to tear up one lab this afternoon after the protein change and stick all of those boxen pointing to your dfQ as a single point of return, to test it out. the dfQ box will be windows 2000 server and a mix of win95 , win98 and winXP workstations sending to it. the client side can handle both "normal" and service installs of df i assume? i guess i will see 🙂
 
i got the new protein 😛

AA 89

speed is about , D'oh! it crashed!

BRB maybe

or maybe it was only dfGUI that crashed, it was a pop up window error about "" not integer or something, but foldtrajlite.exe is still running @ 99% cpu

i think it is quite fast, but something strange is going on, i can't get dfGUI to run, keeps giving me that errror and when i look at progress.txt , it appears that it is only going to do 500 structures before buffering/sending, this is strange

*EDIT*
:Q

i think somehow i got a copy of the beta client, even though i downloaded "distribfold-current-win9x.zip" :Q
 
Originally posted by: FoBoT
i got the new protein 😛

AA 89

speed is about , D'oh! it crashed!

BRB maybe

or maybe it was only dfGUI that crashed, it was a pop up window error about "" not integer or something, but foldtrajlite.exe is still running @ 99% cpu

i think it is quite fast, but something strange is going on, i can't get dfGUI to run, keeps giving me that errror and when i look at progress.txt , it appears that it is only going to do 500 structures before buffering/sending, this is strange

*EDIT*
:Q

i think somehow i got a copy of the beta client, even though i downloaded "distribfold-current-win9x.zip" :Q
Click to expand...

I was going to say, I got that same error message when I used dfGUI and the beta. hmmmm....
 
if you want to get an "early" copy , you have to ftp into the download site , with an ftp program, not a browser

i am going to redownload, maybe howard was moving stuff around and got something mixed up, i didn't read anything that made it sound like we were switching algorithms yet

it is definetly the beta/new algorithm that i got by downloading "current" , i just pulled it again, and got the same file! :Q

ok , i will now be taking a time out, let this settle out, i am sure it is a mistake , probably on my part
 
While I was reading your post, I see my ZA is lighting up...

probably the new client.... I'll let you know what I get. ASAP

(I do remember that I tried to get the last update using 'upload' on dfGUI and it locked up. I thought that
was your thingy until I read your last post FoBoT.)


Update:
Well, that's weird.... My results just uploaded, but dfGUI says I am still on 96AA??? Progress.txt definitely is not the Beta ???
 
downloading again, it is the same file, 6622177 bytes , gack!

i got the same file from the download link/page on the df web site, it has date/time stamps of Feb 24

i just emailed howard about it to check it out, its probably just me 😱
 
well have one machine running it, and one down, errors out everytime at 0 until next gen! (WIN98SE) Have it on my laptop, Win2k and running!
 
Originally posted by: gistech1978
how do i get the new protein?
none of my clients have updated.
this is my first protein change btw.
thanks
Click to expand...

You don't have to do anything. Everytime the client uploads a data set, it looks for updated version and installs it.

(This assumes you answered Yes to the question of allowing automatic updating when you installed, or are running as a service)

If for some reason, your client does not update automatically, you can just go back to The DF site

and download the latest client file, remove your original installation and then re-install with the new file.


I don't think the protien change has actually happened yet though. My clients have been uploading their results, but I am still showing

(dfGUI) that I am crunching the 96 protien.

-Sid
 
Originally posted by: gistech1978
how do i get the new protein?
none of my clients have updated.
this is my first protein change btw.
thanks
Click to expand...

Me too Me too! do I just download the new one now? Because that's what I'm doing. I thought it would update itself...?

Mine is stuck on checking for newer versions, I'm downloading the new one at a whopping 2.5kbs, I guess the site's getting hammered
 
check this out from the error log!

ERROR: [000.000] {ncbi_socket.c, line 910} [SOCK::s_Connect] Failed pending connect to beta.distributedfolding.org:80 (Unknown)


BETA. DISTRIBUTEDFOLDING.ORG!
 
I'm back up and rolln' now. How do i know what version I'm folding?
This thing is rolln' 2.8fps with a 1600+XP, I was getting about 1.8-2 before 🙂

Edit: Back down to 2fps, but that's still pretty good.
 
Originally posted by: bfour
check this out from the error log!

ERROR: [000.000] {ncbi_socket.c, line 910} [SOCK::s_Connect] Failed pending connect to beta.distributedfolding.org:80 (Unknown)


BETA. DISTRIBUTEDFOLDING.ORG!
Click to expand...
I'm thinkin' Howard did a BooBoo :Q 😀


I was getting this:

ERROR: [777.000] {ncbi_socket.c, line 910} [SOCK::s_Connect] Failed pending connect to bar.csh.rit.edu:80 (Unknown)
ERROR: [777.000] {ncbi_connutil.c, line 526} [URL_Connect] Socket connect to bar.csh.rit.edu:80 failed: Unknown
ERROR: [777.000] {ncbi_http_connector.c, line 117} [HTTP] Retry attempt(s) exhausted, giving up
ERROR: [000.000] {foldtrajlite.c, line 1195} Unable to check update file length
ERROR: [000.000] {foldtrajlite.c, line 1497} Unable to retrieve update file list
ERROR: [010.003] {taskapi.c, line 1199} [ReadServerResponse] Timeout waiting for response, got 0 chars.
ERROR: [000.000] {foldtrajlite.c, line 1604} Unable to retrieve update file

but like I said, my results ARE uploading OK ???
 
the windows manual download defintely contains the beta .exe

Welnic pointed out that the old non-beta .exe can be overwritten into the new protein directory to wipe out the beta .exe with the correct .exe

i just did it and it appears to be working ok

now i just gotta burn my "master" working distribfold directory to CDR and manually update all my boxen
rolleye.gif
 
You must log in or register to reply here.

TRENDING THREADS

Back
Top