APS@Home gone public

[Tutta55]

http://www.apsathome.org/img/APSathome-logo.png

The APS@Home Boinc project (Atmospheric Process Simular) has gone public end of last week. This means that it is out of the testing phase and account creation is open. See http://www.apsathome.org

The project is now running "real" work units. It has clients for Windows, and Linux 32/64.


EDIT: A few additions

Project mission:
APS@home is a research project that uses Internet-connected computers to do research into the effects of atmospheric dispersion as it relates to the accuracy of measurements used in climate prediction. More details are in the Science message board, and we may be diversifying into other areas of atmospheric science soon.

Daily stats in standard Boinc format are available from http://www.apsathome.org/stats/

[Meadmaker]

Thanks for the update Tutta55. :clap:

Account created successfully. :D
Team created successfully. :D

[russkris]

Thanks for letting us know dude....


Start post your up's and downs and we can get the ball rolling to add it to the Vault

[Meadmaker]

First WU crunched, returned and credited :D

[cswchan]

Nice find... just attached with 2 rigs... going to see how they run... 1st is 5 hrs. & 2nd is 8 hrs. long...

[Tutta55]

I have been involved in alpha testing this project since the very beginning. It has become stable pretty quickly, and is now in production. The awarded credit is well in line with that of most Boinc projects. Many participants appreciate the presence of administrator James in the project forum.

[cswchan]

Hmmm... the project seems to have run out of work since last night...

:(

[Meadmaker]

We've "drained the queue" it seems

[HiveChild]

Thank you for posting the news of this project going live!!!

team has been formed, and I am just waiting to get a workunit :D

[Tutta55]

It is not unusual for the project to run out of work. Work release for this project will be intermittent, in batches.

[cswchan]

Sooo... now you've REALLY done it... even their main page is down...

:rolleyes:

[HiveChild]

Sooo... now you've REALLY done it... even their main page is down...

:rolleyes:


Anyone know what is going on with this project ?

Main site seems to have down now for along time :(

[Tutta55]

APS admin James Dorsey wrote in the BoincStats forum:

Hi again all,

The server is still down, and I'm now not sure when it's up. Our department has started moving into new buildings, so there's nobody about to gain physical access to the server and router. At this point, I can't promise that the system will be back up before I return from Germany on Monday the 30th of July. If you want to detach the project until then to avoid the error messages, I understand. I'll post again when the server is up.

It is still possible that someone may be able to get in to power cycle the system, but it's beyond my control. I'm sorry about this, and I hope you all understand...
Please could you distribute this to your team colleagues? It's not very good news, but there's still a glimmer of hope that it'll be fixed this week.

[Meadmaker]

Thanks for letting us know Tutta. Hopefully it'll be back up again soon.

[HiveChild]

Thank you for the info Tutta :)

[russkris]

Awesome, Hopefully they can get stable soon..

?

[Tutta55]

After solving the router problem, the project has been up again for over a week now. Some new work is just released. Maybe a good moment to start taking it into consideration again?

[russkris]

:nod:

[russkris]

We have the offical nod(kinda) from the site admin to add the project to the Vault.

Post away people's, let us know how it runs etc

Work come in stages, they dont really like the idea of people crunching for the purpose of getting point, they like the idea people crunch fo rthe science...

[russkris]

Anyone?

[vaughan]

This one's as bad as SIMAP for only having work occasionally. Does it qualify for inclusion in the DC-Vault?

[russkris]

Bugga..... Not good news

[vaio]

Work availability does seem to be an issue with this.

[Tutta55]

The project will probably never have a constant work supply. That is a conscient decision of the administrator, who only releases work when there is a scientific need, and does not want to waste people's resources just to keep computers busy. I would say that is a positive. Furthermore, his setup is such that when there is work, it will be distributed over as many participants as possible.

I might be wrong, but it seems to me that the criteria for admission don't state that work supply must be continuous. :confused:

[russkris]

I might be wrong, but it seems to me that the criteria for admission don't state that work supply must be continuous. :confused:

110% correct... I will look into having a chat with the admin

[Tutta55]

There is work from the project right now. Some 5000 work units. Grab a few while you can, if you want a test run.

PS: No, I'm not getting paid by the project ;) I just like it.

[russkris]

Whats the longest period of time you would go without work?

[Tutta55]

Here is what James Dorsey wrote yesterday on the subject of work production:

should add this to the FAQ. The reason that work availability for this project is a bit erratic is that I have to create work interactively in batches. I won't release work just for the sake of having work available, - that would waste your CPU time and just increase everybody's electricity bills for no good reason. It takes me a bit of time to generate work, and I have to put some thought into the parameters set for each batch. At the moment I'm still at the stage of assessing the sensitivity of results to different input parameters (there are lots).

Added to the fact that I need to take some time to consider what work to release, there's still software development work that I need to do to speed up the analysis of results. Add to this the fact that I run this project on my own, and it's something I do additionally to my main responsibilities, and you can probably see why I struggle to keep all of you busy all of the time.

It was never my intention that there would be a constant stream of work on here. It may become that way next year when we're trying to calculate footprints in real time, but for now there will always be gaps between batches of work.

I don't recommend that people run APS@Home as their only project, or even as their main project. Some people here at the university do, but they're not trying to amass huge cross project total scores.

The crunching you do here is currently useful. It would be easy for me to run a work generator, but the results would, by and large, not be scientifically useful. I would like to avoid that situation if possible...

This time we had been out of work since sunday. The period without work usually does not exceed a week in my experience, with a few exceptions. Certainly not months, like some projects.

[vaio]

You going to add this Rusty?
I like the policy of not issuing "make work" just to keep rigs active........at least we know the work we do is useful and not occasional placeholder stuff.

The vault needs some fresh new projects to help keep things interesting :)

[russkris]

Look to be the case mate.. APS will probabyl be added and I am trying to get a few more added

[vaio]

Cool.........aps has been "vaulted".......thanks Rusty :)

[russkris]

No worries mate.. There was a bug, in the adding function, hence why it took longer

[Tutta55]

Hey, great. I will try and keep you updated on APS news.

[vaio]

Feed me :D

[glennpat]

XtremeSystems finally got a work unit turned in :) . I assume we'll now show up in the Vault as having a team on this project.

[fractal]

Over at http://www.apsathome.org/forum_thread.php?id=151 the APS project administrator says

We'll have work before the end of the year - promise!So all is not lost.

[russkris]

Thanks Fractal.....:woot:

[KAMCOBILL]

Cosmology, Enigma, Lattice, Poem, Sudoku and Yoyo seems to have work all the time.