Hi Larry

It might be worth us setting up a hardware swaps/donations thread. I've got a P2 400 CPU that you can have, but unless you're in the UK then its probably not economical to post it to you - the same money would probably buy you a 1GHz Celeron plus slot1 tualatin adapter fron ebay that would probably get 4x the work done (as the cache is full speed on the P3s, half speed on the P2s).

I guess the server would send you work anyway... I might be wrong, but I'd take it the result would just be longer to crunsh, possibly over the deadline, because your computer would have to swap memory all the time.

I might be wrong though, never tried to add a PC with only 64Mb RAM.

- trib'

LarryM, I finished my tutorial even if it looks ugly.
I'd need a few beta testers to make sure I didn't forget about something crucial before I post it to the general public... Would you like to be one of them?

Anybody else that is interested is welcome, too.
Requisites are: a spare to-be crunshing machine (preferably old crap) with 128Mb RAM, about 1Gb of HDD and a floppy drive. Processor is up to you, my slowest machine is a P2 300Mhz. Oh yeah, and about an hour of your time available. Well... make it two =)


If you're available, send an email to: /tribaal/(AT)/tiscali/(dot)/fr
(remove the / and let (dot) be . and (at) be @)

- trib'

I have just managed to acquire a copy of Windows 98 SE.

...so, if I hear you right, you've got two weeks with which to help a friend get up and running on Rosetta?

This thread is to celebrate milestones in your contribution to Rosetta, to BOINC, or to Distibuted Computing in general; and to congratulate others on theirs.

Doesn't matter what the milestone is, if it gives you a buzz, celebrate it here!

He's added over 150 hosts in a week!

Please, someone just ship this guy their copy of unused win 95/98 that they are no longer using (this way it doesn't violate laws and so forth) ?

I mean, It's 1$ worth of U.S. postage. Make it happen, guys. I've got one copy of 98 and it would violate copyright law to send it to the poor guy.

I agree, a Memtest would be good to run anyway.

With the setup I suggest to use, you should still have about 128Mb RAM available. My smallest machine has around 124Mbs and works like a charm, but then around 120 Mbs are used by Rosetta.

If you're not sure what memtest is or where to get it, try this link. Basically you create a boot floppy with their program on it, that will try to write bits to every sector of your RAM, and reports any errors.

It's good to have in your toolbox anyway =)

- Trib'

Well I realy am not familliar with DSL at all, but I would take it connecting to your provider shouldn't be too hard... In my set-up, the xDSL router is plugged to computers via regular network cable (rj45), and so the only thing linux needs to know is the network card driver (which is really standard stuff, failure is exceptional).

And even though I don't know how AOL works, I take it plenty of linux users use their network, and so they already found a workaround any problem that may arise =)

I'll keep you updated on the tutorial.

- trib'

get yourself a copy of Damn Small Linux - I have it running on 4 of my 6 boxes at home. One of these is a 486DX2 16MB (which is just too old to cut the mustard) the others are a celeron 348mHz a celeron 1.7, and a p4 2G. My PII 233 is about to be converted to linux from win98se - I aborted the last wu as I was testing out a new project.

I'll put up a short tutorial on "how to set up a machine as a 100% roetta crunsher" using linux (my tutorial will cover installation as well, in case you're unconfident).

I used this technique on several machines already, and it works well for casual crunshing (I know, some make diskless computers and whatnot, but I stick with the stable Debian OS =) )

This tutorial should be online soon, I'll post a link in the forums.

- Trib'

Sounds like your hard drive has seen many days of good use. I haven't seen a failure predictor like that before. Perhaps you could post a link to some info? I would have to GUESS that they can't be very accurate. So, maybe a month, maybe a year. If the predictor is based on the MTF (mean time to failure) of the drive, even that has a wide varience I'd think.

I only am running the one computer, and cannot find away to get rid of all the computer ID's.
I read there used to be a merge function, but that it was done away with, so I guess there is no way for me to get rid of all these ID's.

This is one of the things that Rosetta inherits from BOINC. The standard server install that BOINC projects use had a bug with the function to merge hosts together. So, they (Rosetta) just disabled it until BOINC fixes the bug, and then the function will be available again. Your multiple hosts aren't hurting anything. No need to worry about it. When the merge function fix is made, you'll be able to consoladate all of the hosts (and any credits) into one.

However at the end of the day it's more important that the WU's are being done.

HERE HERE!

Thanks again for your help, and I will be staying with the project.

hurray!

Crunch more Rosetta!

Be sure to check out Dr. Baker's Science Journal. He's really doing a great job keeping us informed on progress, and science, and how Rosetta works.

To assure that your BOINC Manager is using the settings you've configured, you'll want to go to the commands pulldown and assure that the "Run based on preferences" is selected. This gives you a quick way to temporarily override those settings directly from your PC, without going through project preferences websites, and updating to the project etc.

I too have been studying my hard drive usage. I'm trying to use the power settings and get the thing to power down, and maybe only write to it every half hour or so. But, mine keeps running too. Never powers down.

What I'm finding is that the cause is:
1) other applications and services running on Windows.
2) page faults, very few of which are caused by Rosetta tasks.

By the way, I'm studying this by looking at the Task Manager of Windows, and then I went in to the VIEW and "select columns" and I added the columns for PF Delta (page faults since last snapshot), I/O Read, I/O Write, I/O Other.

From that, I conclude that if I had more memory, then Rosetta wouldn't be causing page faults. In my case, I've got a dual-core processor, and so two BOINC processes run at the same time, and so even though I have 1.5GB of memory, Rosetta times 2, plus all the other stuff going on is using all of it. I'm even seeing iTunes tasks kick in and generate page faults... and I've not used it since I rebooted my PC.

I even moved all of my BOINC files out to a network drive, and so all the IO happens over the wire... my hard drive still runs constantly. Never powers down.

So, from this, I infer that my hard drive would be running due to page faults from iTunes (and others) regardless of whether Rosetta was running or not. Do you find this to be the case as well?

I see your 4 machines are all 512MB, the recommended minimum. Looks like they are single CPU though, so your's aren't running two at the same time the way mine does. That helps use less memory. (for future readers, if you DO have dual-cores, you might use less memory, and avoid page faults, if you set your general preferences to only use one CPU).

Please avoid sentences like the last of your post. Everyone's going to try to help you out, as soon as they can... with, or without the threat of you leaving. We're just happier to do it, without.

If you still feel you have a specific problem with Rosetta, please provide more information about what you're observing, why you feel it's abnormal, and (if possible) how you feel it might be resolved. That way at least you've contributed some useful information so the project might take steps to avoid upsetting others in your situation in the future.

Absolutely there is.

First you have to go to the BOINC manager and get the protein's identifier... I currently have a unit called

ALL_TOPOLOGY_CODES_1vcc_434_295_0

The protein's identifier is 1vcc. Then you can go to the Protein Data Bank and enter in the ID (in the search at the top).

You'll then get a page (like this) which has the protein's name, what species it's from, what kind of protein it is (mine's a topoisomerase!!) and which paper published the crystal structure (or NMR structure).

Hope that helps!