boinc - TANPAKU

[Brucifer]

Got signed up today and have a couple systems crunching away on it. Overall it doesn't appear to have many people throughout the working on it yet. I saw we had a team on it so I joined that. Since this is another protein effort, it should be a good one. I lived in Japan for eight years. Tokyo University is the #1 University there, and this is a project managed by them. So it should pan out to be a good project. English is lacking yet, but they are working on it. So I'd have to put this effort up there with Predictor and Rosetta. I see there are a couple other team members on this project too.

Oh, and you *have* to use the boinc 4.x software, it isn't ready for 5.2.x yet but will be at some point in the future.

Now I just wait and see how long it takes to crunch out a wu.

[gamer007]

Would be better to post here in the already created thread.

[NeoGen]

So now you're starting with Boinc, heh Brucifer?

Tanpaku is so far a rock solid project. Strangely it is the second boinc project I see that enforces boinc v4. The other one is Cell Computing, also Japanese.
Knowing I that the last versions of the v4 round were very stable and that the switch to all the new v5 features was what brought in so many problems, I'd say they follow the rule "If it's working, don't mess with it"
The japanese people are smarter on that than us western folks that try to "be on the bleeding edge" with the latest version all the time. :roll:

[vaughan]

BOINC - Tanpaku thread created and topic moved here.

[Brucifer]

So now you're starting with Boinc, heh Brucifer?

There are some boinc efforts that are decent. The Xtremlab thing seems to be doing good so far. Maybe they did learn some stuff from the first go-round. The Japanese one I just had to try just because. As I mentioned about living there, I sort of figured that their project should be pretty solid. They normally wouldn't go public with it if there were problems other than language issues.

I may be wrong, but I think this one started using a custom boinc engine during the first part of it. Myself I sure don't see anything wrong from the linux end of things with the 4.x boinc version. It just didn't have all the bells and whistles they tried sticking into the v5 stuff.

And I also don't configure the boinc stuff to run multiple projects under one installation. I make multiple boinc installations with the stuff set up for just that project. May not be the fancy design thing that Berkeley initially designed, but I've encountered fewer hassles with it that way.

But to get back to your original comment about me starting with boinc..... I'm just putting a few systems on some other stuff mainly to do something a little different. I've been totally balls to the wall with the PSP sieving for a while and I get these urges every now and then to alter the routine a bit. We are starting to almost see the light at the end of the tunnel on the "combined goal" effort there, so I thought I could take a few systems off the sieving for a bit. But make no mistake, 90% of my stuff is still sieving, and only the older low power stuff is on the boinc fancies.

[Brucifer]

Sure doesn't seem to be quick on the points... :?:

[NeoGen]

We can't check the results, but they're most likely pending. It happens frequently on new projects. With few people crunching it is harder to validate workunits, if they're in a quorum system.

[Brucifer]

Looks like that part is switched off, as you can't view pending, etc. So it's hard to tell if they just award points a little lower than others.

[NeoGen]

Tanpaku now has english main page!

http://issofty17.is.noda.tus.ac.jp/index_E.php

EDIT: Brucifer, you've just jumped up to the top 10 participants.

[Brucifer]

Man, those points come slow on this project!

What procs do you have running on this one neogen?