Saturday 02/11/02

The new Gromac protein pAla7-2 sure has lifted our output over the past few days going from 5300 to 7300 points a day. We need everyone to change over to the Gromacs to give the team a really big boost and try and keep our nose in front of the oc.com team on the weekly stats.

To make sure you get the new protein you will need the latest version of the client (3.14) start the client with these switches -advmethods and -forceasm. To do this make a shortcut and add the switches to the target like this...

"C:\Program Files\Folding@Home2\FAH3Console.exe" -advmethods -forceasm

If you don't get a gromacs WU after the current WU has finished you may have to delete the queue.dat file and the work folder and then restart the client again. Also make sure in your client.cfg you have disableassembly=no, I had this set to yes on one of my machines and it overrode the manual switch above and the protein takes three times as long to process. The thread in the forums about this subject is still going strong so if you have any questions have a look in there.

The 5 million challenge looks likely to happen now with the boost from the pAla protein. Yesterday the 3 teams produced 25,520 points so the target will definitely be reached if this protein is around for a few more weeks.

TheWeatherMan has been busy again and has updated his EM3 program, this time a major change to the different views that are available.

Electron Microscope III v2.1.3 (Update)

Fixed Frames progress problem.
Fixed Language label issue.
Fixed extra launch command issue.
Added several new views (3,4,13).
Added ability to select and scroll largest view.
EM changes size to fit view selection.
More Linux issues resolved.

Those who have gotten one of the new 200 frame work units may have noticed that the frames progress bar would show 100% when it was only half way finished. This issue has been resolved.

When using the language files the HTML page had the labels for the drop down selections for proteins and boxes reversed. Fixed this issue.

When adding extra launch commands, after selecting to use it, EM would sometimes clear the option when restarting on some systems. This has been fixed.

I have been adding a lot of changes to the way the EM program can be displayed and manipulated using the skin selection menu. Along with being able to change most item images, the image and button positions can be changed. Because many people wanted to resize EM or display more clients at once, I have added and enhanced the viewing capabilities. Now, even if you are only watching one client you have all the different views available to you. What this means is EM can now change in size as it shows different views. The current view selections now available are, 1, 2, 3, 4, 5, 7, and 13. In the multi-view selections (2 or more) EM will size the window to show only what you have available to display up to a maximum of its original default height. When EM reaches its maximum height and you have more clients to display in that particular view, EM will display the scroll bar so you can move through the clients.

I expect that many of you will like this new feature. While I have tested everything pretty well, this new ability may have some issues I did not catch. Please let me know if you see anything weird with the display. If you find a problem, please remember how you got to the problem so I can find it easily.

While most Linux issues have been resolved, the queue.dat file display is not working perfectly. I should have that fixed in the next update.

Enjoy!

Get it here: http://www.em-dc.com
 

Sunday 03/11/02

The xylus 500 is up.

Wednesday 06/11/02

The party is over and the pAla7-2 Gromac proteins have stopped being issued, the server is still running so that you can return any outstanding WU's. It looks like we all had a bit of fun with these super fast point scorers, just look at the list of names in the Folding Club promotions above. Our weekly production has increased by over 13,000 points and made the 5 million by New Year challenge a bit more obtainable. Time will tell if we will see this type of point scoring typical of future Gromac proteins.

A couple of comments from Vijay about the next Gromac project and server 117...

"we're retooling that proj. Already got a lot of data. Expect a new project there soon."

"Likely next week -- we are redoing some aspects of the planned new Gro WUs due to a science issue."

Last night there was an outage that was beyond the control of the Folding@Home group, apparently a switch cause the problem on the Stanford network.

"Looks like there are some issues with the Stanford network -- I can't even get in remotely to see what's going on. It's a spotty outage, so many things are up, and some things are inaccessible."

and an answer to a question of some missing stats...

 "we might need to input some of the stats manually by script for the WUs received during the switch outage. I'll ask Chris to do this. He'll likely get it done in 1-2 days or so."

Buzzard has been busy will his Stats Grabber and has released version 0.82 hot on the heals of version 0.8. Its a very handy program and now can save your stats in CSV format on a regular basis or at a specific time.

A new version of KDfold is out version 5.8c. Thanks to Robert for this program that can just about do everything when it come to monitoring your DC projects.

LinKDFold (Linux) V5.8c F@H 1,2,3 G@H 1,2(F@H3) DF & F-A-D
WinKDFold (Windows) V5.8c F@H 1,2,3 G@H 1,2(F@H3) DF & F-A-D
Release 5.8c 11/03/2002
Monitor Multiple Projects and Clients from one application instance-
KDFold Currently Supports- Folding@home 1,2,3 & G@H 1,2(F@H3) F@H & G@H Autoswitching in newer clients & Distributed Folding & Find-A-Drug

1.FAD-Fixed Client Log to show proper log for each process.
2.FAD-Fixed a % complete bug if percent complete was under 10%.
3.FAD-Fixed the machine name population visual error so when chosen it populates immediately.
4.F@H, G@H tweaked some of the date handling routines in the main check routines.

Known issue: LinKDFold: Some of the themes for KDE or Gnome cause a window resize. I am working on this issue.

Check the forums for all of the changes and keep current on KDFold News.

Winkdfoldv5.8c(Windows)-

Download WinKDFold v5.8c(Windows)-
http://www.kdfold.com/winkdfoldv58c.zip

Download LinKDFold v5.8c(Linux)-
http://www.kdfold.com/linkdfoldv58c.tgz

Download LinKDFold(Linux) install script-
The script is inside the .tgz file but this may help some
http://www.kdfold.com/install.sh

Install Example-
./install.sh linkdfoldv58c.tgz

Home Page-
http://www.kdfold.com

Support Forum-
http://www.plasticstudios.org/forums/index.php

Have fun,
Robert-

TheWeatherMan has also updated EM3 as well, grab it here

One of our folding members petercr and his wife have just had a new arrival in the form of a baby daughter named Joanne, on behalf of the Folding team congratulations Peter and Lindy. Drop by and wish them both well.

"Joanna is the reason my wife and I have been Folding & continue to do so..."

Monday 11/11/02

Useless fact No 3456: If you type in folding@home in to Google this very page comes up 9th of 29,000 web pages returned, interesting hey.

Here is an interesting article from The Tech Report about what effects Folding@home has on your system performance. An interesting read and confirming what we already knew....

The results of our testing couldn't be clearer, at least for the systems we've tested today. Quite simply, the impact of running Folding@Home in the background is negligible.

Read the report here

The xylus 500 is up.

It looks like Statsman has erased all the data except for the past 5 updates. This means that the PPW numbers for the Folding Clubs are going to be out for the next 6 days. Also I have just seen that he has cut back the number of names to 1000 so that means you need to be in the top 1000 to be included in the hunt for the Folding Clubs or at least 150 points at the moment. This has certainly put a spanner in the works for the newer members.

New versions of EM3 (v2.1.8) and KDFold (V5.8e) have been released.

One of our members, Plext moved to England a few months ago to take up a new position and promised to bring online a few machines to help the cause. After eagerly looking at the Plext's stats for the past few weeks I noticed a significant jump in his output which has signaled the start of his test phase before he unleashes the rest of his fleet. His current CPU count stands at 94 so if this is only a sample we will look forward to what you have in store once the "test phase" has been completed. Welcome to the big time Plext...

Cheers people

I'm finding that there is at least one advantage to living in (to me anyway) this sub-arctic country..........cpu temp

Seriously though, I'm hoping for some good results next week.

This is the test phase, and hopefully only the tip of the iceberg.

The Server Status page has had a few new features added to it over the past week or so. Here is a list of them and what they mean...

WUS AVAIL: shows how many WUs are available for the server to give out. If the server runs out of WUs to give out, it will tell the AS and the AS will stop assigning new clients to that server.

WUS WAITING: shows how many WUs have been assigned by the data server. Note: this can include WUs which are beyond the deadline, so it may give an exaggerated count.

WUS RCVD: shows how many WUs have been received since the last time the servers WUs were updated into the stats. This shows the relative number of WUs being received on the different servers (if all is fine) or which servers are not being inputted into the stats db if there is some problem.

WT: reflects the server weighting -- how frequently the AS will assign clients to that particular server. Higher numbers means more frequently. Zero means never.

MIN SPEED: is set for certain servers with projects for fast machines only. The speed is determined by the client's benchmark. Future client versions will print the speed to the logs so one can know the relative speed.

Plus a few others with no explanation like "DL" and "HDF". I would assume DL means the number of WU's that have been downloaded since the page was last updated.

The "MIN SPEED" feature looks like a good one and when a new client has been released it will print out what your machine has been benchmarked at. Therefore it looks like they maybe finally getting around to issuing WU's in relation to how fast your computer is. All this is dependant on and what Stanford considers a fast computer is as well.

A few folks are complaining about not being credited with stats from last weeks outage after Vijay manually credited the missing stats. In a thread in our forums and on the Folding-Community forums a few folks have pointed out that they are still down a fair few points from last week. I checked my stats and what EM3 has said I have done compared to what I have recorded at Stanford and I seem to be missing quite a few pAla's. Also at the time of the correction the "anonymous" username had a big jump in its stats indication perhaps Vijay lumped all the missing stats against that name instead of the individuals that did the work. It will be interesting to see what becomes of this after the strong wording in post in the Folding-Community thread.

The Google army have now claimed the overclockers.com team and are in 4th place. We are next to get invaded by the ever growing mob of faceless folders. At the rate they are producing we have less than two weeks before we will be overtaken.

Since my last news update our No1 folder DGROMS.com (otherwise known as David or FB) has been busy and has the honour of passing Norgs in the "Points Per Week" based Folding Clubs and has slipping in to the 6000 Club. Which is a great achievement helped along by the points friendly pAla protein, never the less it is a grand effort and one we are all in awe of well done DG. Also DG has amassed over 60,000 points which is a testimony to his dedication to the cause and staying power, no doubt with his future potential output we will see DG the No1 folder in the world soon.

Also thanks to Dean (Darkness') for whipping up a nice shiny new (chrome or is that silver plated) Folding Club badge.

Tuesday 12/11/02

Only 1 million more points to go and the challenge to reach 5 million by the [H]ardOCP, overclockers.com and OCAU teams will be complete. We should reach the target with a few days to spare at current rates of production.

Well it looks like Plext has done enough testing and more and more of the iceberg is on show, he has so far got 232 284 active CPU's and is really starting to crank out the points. But Plext has let a few more clues out about what he has planned.

This is still the testing.
Expect the 'berg sometime around christmas.

at this

Mostly P4s@ 2.4
By christmas there should be about 2000
By late next year upwards of 10k
 

Now that's looking good for an OCAU Folding team Christmas present.

The daily stats were a bit low this morning due to the Stanford servers being overloaded. This message was on our stats page

The Folding@Home server is currently under a high load, please check back in 2 hours (message cached Mon Nov 11 14:38:29 PST 2002 ).

Maybe Plext had something to do with it.

Saturday 16/11/02

There is a poll running in the forums on the issue of keeping or removing the Aggregate teams (Default and Google) from Statsman's page, go have your vote say on the subject.

The Plext factor is still on the increase and he has now past the 500 active CPU mark.

Tomorrow's  10am (1pm ESST) should see the stats on Statsman's pages complete a full weeks worth of updates so we should start to see a lot more Folding Club promotions in the next couple of days (besides Plexts' daily club promotion).

Sunday 17/11/02

The xylus Top 500 is up.

Monday 18/11/02

Congratulations to the overclockers.com team on passing the 1 million milestone, welcome to the club.

We have been Googled on Statsman but if this poll is anything to go by we should be back in second place soon.

Useless fact No 3457:Did you know we have 106 members in the Top 1000, [H] have 124 and oc.com has 87.

The Folding club is only 10 away from 1000 members, which we should pass in the next day or two.

Tuesday 19/11/02

For you adventurous types there is now a new beta out (3.23) with plenty of new features to try out and still a few more extras to come before the official release.

Windows console 3.23beta
Windows graphical client 3.23beta
Linux console 3.23beta

The next client, v. 3.23 is now available in beta form for Windows and Linux donors. You can get it at http://folding.stanford.edu/beta/beta.html (note that this is a different setup than before). It incorporates much of the To Do list items pertaining to the client that was prepared on the Forum, and more will be coming. The Macintosh version lag time should not be long this time.
Thanks,
Guha

Here are the changes:
- If a work unit is rejected for any reason, server now returns reason
to client, which prints it out. (In this client, and servers will transition
to the new code as projects come online.) If work unit is fatally flawed,
then deleted from queue.
- "Could not transmit unit" statement now also includes date unit was
completed ("Could not transmit unit x (Completed xxxx)").
- Checks for bad URL, port on reception of work unit, deleting it before
even starting work if bad.
- Time stamp on every line of FAHlog.
- Arguments to client printed to log.
- Some log messages modified (like the requested, and deserved, "Thank you"
added!). Some further spacing changes may be desired.
- File "unitinfo.txt" printed and updated, with information and progress
for current work unit. (Had been requested for use by 3rd party people,
but keep in mind that format may remain fluid until release.)
- Benchmark done before every new work unit request; result printed (on verbosity 6).

Console (Linux/Windows) specific-
- Answers to Configuration questions now validated. One letter answers (i.e.
"y" or "Y" instead of "yes") allowed.

Windows (all versions) specific-
- Client FAHlog output overhauled to use Windows API functions. (Interested
in hearing if you see any change in performance, particularly from those
Win95/98 users who said log froze sometimes.) Log file is locked for
writing by other programs while client running.
- Networking overhaul to use whatever proxy settings are used by Internet
Explorer (default setting).

Windows GUI specific-
- Bug fixed where if click off systray menu, menu did not disappear.
- Tool tip on systray icon now gives program status (including work
progress if currently working)
- When move mouse over hyperlink portions of display, mouse icon becomes a
hand.
- Bug fixed where if move display, then the links at the bottom of the
display (ticker tape ones) did not always work.
- Menu item ("Status") added to give easy access to Team Stats page, User
Stats page, and Logfile.

Planned to be ready for release 3.23 or next beta 3.23-
- Access to queueinfo information from the GUI menu.
- Requests donors make that are simple -- other requests kept in mind for
next revision.

Statsman on the Google issue...

First of all thanks for everyone's input and the relative civility that was shown in the posts.

The sentiments from [H]ard, OCAU and Overclockers.com are heavily in favor of removing the aggregate groups from the StatsMan stats pages.

I think I am going to try and implement a compromise that I saw suggested on one of the forums (or maybe this is a combination of a couple, sorry for not being able to give credit where due).

I am going to remove the aggregate groups from the graphs and the current Total Teams table. I am going to create a seperate table (similar to how Stanford seperates them) for the aggregate groups, but there will not be individual group pages for them. I really don't want to have to track their individual users, just the group totals. This way those that wish to see how these groups are progressing in total and the contribution to the project that they represent will have that information. For the rest of us who don't care, you can just ignore the additional table.

Wednesday 20/11/02

TheWeatherMan has a new protein file available for download for Electron Microscope III and also a note of thanks for all those that have supported him for providing a great program.

I just posted a new protein data file. You can pick it up in the download area, or you can let EM just pick it up automatically.

Here's a list of the new proteins:
P186_1fsvunf280
P187_1FSVnat
P188_1fsvunf300
P189_1fsvunf320
P190_1fsvunf340

They're all worth 6.4 points... haven't seen any of them yet, hope they are fast points.

EM checks on new stuff once a day, midnight to be exact, and will autodownload the latest protein file and notify you if there is a new updated EM program. You have to be connected to the internet for this to work. EM does not dial out and try to get this information. I've got some really cool stuff coming in the next update. I'm looking forward to finishing the beta tests.

Thanks to everyone who has helped out by donating to PayPal or used the ad links to purchase stuff. Everything goes to expanding and supporting the web site.

There is a new Linux beta out because of a bug with the Log file in the original release.

Friday 22/11/02

If you are using EM3 there is a new Protein Data file available for download which includes the new proteins 612 to 616.