When and what ? :-)

[modesti]

  Of course, I haven't been thinking of the Simap come-back.
Probably because they launched some work two days ago :

Server migration to Vienna: first step passed
The server has now been moved and preliminary configured in Vienna. It still uses the old project URL and the old website design. During the next weeks, we will test the technical performance and stability on the new hardware. During this time, smaller and larger chunks of workunits will be distributed. Although these are intended for testing, their results will be used and get credits as usual. If everything works fine, we will launch a renewed project website using a new project URL in December, and then start into production again.

best regards
Thomas 15 Nov 2010 19:04:10 UTC

And there seems to be plenty of it : http://boinc.bio.wzw.tum.de/boincsimap/server_status.php

RESULTS   Approximate #
in database   235,005
unsent   49,377
in progress   99,935

[metepuck]

To get this back on track: Can we agree on Feb 05 to Feb 20? If so, I think we should go on and announce the Charity Crunch beginning of December?

[elgrande71]

It would be great if Michael could reduce the memory consumption of his application.
For me, it will be anyway a pleasure to participate to this charity race.

[Black Hole Sun]

RNA looks like a nice candidate... whereas Rosetta doesn't at all: Charity Race is to help a small bio project.
About SIMAP, I'm fed up with that project disturbing the Boinc sphere for years
Simap fans believe the Boinc world should stop for this project: sorry, SIMAP has no more interest than any other Boinc project out there!
If Simap fans can't understand that, we will race without them. Period.

About the dates, it may be wise to first check the winter scholar holiday period in France (don't know if there's any in Germany).

[modesti]

In France, the "winter holidays" begin on 14th February for zone C and end on 12th March for zone A.

As Simap announced continuous work for about 2-3 months (beginning last night), I think that there's no need to take it into consideration. IMHO, we could easily begin the Charity around the last week-end of January.

[yoyo]

If the goal of the charity race are small projects, than SIMAP is the wrong one. SIMAP has already more than enough user. They have so much user, that they are not able to provide enough work for them. Don't see it wrong, I like SIMAP, but they have already anough crunching power, more than they need.
yoyo

[metepuck]

Right, but it is getting really tough to find a small bio/med project that is capable of handling all that power. I would be fine with RNA but the problem there is the long WU with no checkpointing. This makes it impossible for regular non-24/7 crunchers to ever finish 200+hrs WU, not even mentioning the rare 600+hrs WU.

If checkpointing could be implemented until the CC, the choice would clearly be RNA but if not, the other candidates are SIMAP and Rosetta. And SIMAP is going to win this challenge.

[modesti]

According to the RNA World message board, the project doesn't support checkpointing

[Black Hole Sun]

Well:
SIMAP: OUT! Not a small project, they have enough power.
ROSETTA: OUT! Not a small project, they have enough power.
RNA: OUT! No checkpoints, too long wus.

Next candidates?

[metepuck]

What about ibercivis?

They're not 100% Bio/Med but they are also doing drug research.

Small project, only 15k users, short WU run time, just not sure about the server, we would have to check this first.

[yoyo]

Well:
SIMAP: OUT! Not a small project, they have enough power.
ROSETTA: OUT! Not a small project, they have enough power.
RNA: OUT! No checkpoints, too long wus.

Next candidates?

BTW, the most workunits in RNA World are short. Here you see how long the workunits are really:
http://wuprop.boinc-af.org/results.py?data=duree&projet=RNA%20World&application=cmsearch
yoyo

[metepuck]

Well, even if there are only 1% of them longer than 10 hours, these will shoot out a lot of regular crunchers who are usually switching off their machines in shorter time. Because without checkpointing they will start over this WU again and again and again until it is expired.

I am not saying that it is a bad project, it just does not work for non-24/7 crunchers.

[modesti]

Well, it's almost Christmas and we still do not have any project, nor have we fixed the dates.

As the admin of RNA World seemed to be happy with our Race, maybe we could ask him to try and release only short WUs (i. e. less than 8 hours for example) during the time of the Race. He could leave the "Sumatran Orang Utan" for later in the year (just an example of my experience - it was something like 2-3 days non stop calculation that ended up with "computation error").

We could also ask the members whether they prefer RNA World or Ibercivis (shall we then restrict it to biomedical sub-projects?). But a vote will take time and I think we've already lost plenty of it.

And: we definitely need to fix a time scale! Time is running out of our hands, 2011 is almost at the door...

[metepuck]

You are absolutely right! What about Jan 30, 0:00h until Feb 12, 23:59h?

But I don't know about RNA World. We might be good with a cut to a maximum of 6-8 hours but we definitely have to announce it before. But let's first see if it is possible.

[modesti]

The dates you suggest are fine for me and for the few people who reacted on the AF-Forum since I posted yesterday.

As for the project, Ibercivis seems to get their favour.
If we choose Ibercivis, shall we restrict the Charity Race to the bio-medical sub-projects Docking, Amiloide, Neurosim and Sanidad ? Or shall we take Ibercivis as a whole ?

Any news from RNA World? (Who is in contact with its admin?)