Well...

If all around here is like in SETI (credits, outages, etc...), I come back to SETI. Why I must been here if the thing is the same as in a 'real project'?

Depends if listening out for aliens is your thing, or if you get frozen out of SETI ;) The last part of which I've never tried to hide or deny, or give another toss about... , just in case anyone was wondering...

I like more to listennig the aliens...;)

Do you know what your work is doing here...?

Sure. It's all on the MW home page ;)

About MilkyWay@home

The goal of Milkyway@Home is to use the BOINC platform to harness volunteered computing resources in creating a highly accurate three dimensional model of the Milky Way galaxy using data gathered by the Sloan Digital Sky Survey. This project enables research in both astroinformatics and computer science.

In computer science, the project is investigating different optimization methods which are resilient to the fault-prone, heterogeneous and asynchronous nature of Internet computing; such as evolutionary and genetic algorithms, as well as asynchronous newton methods. While in astroinformatics, Milkyway@Home is generating highly accurate three dimensional models of the Sagittarius stream, which provides knowledge about how the Milky Way galaxy was formed and how tidal tails are created when galaxies merge.

MilkyWay@Home is a joint effort between Rensselaer Polytechnic Institute's departments of Computer Science and Physics, Applied Physics and Astronomy. Feel free to contact us via our forums, or email astro [at] cs [dot] rpi [dot] edu.

... and more... ;)

Well...

If all around here is like in SETI (credits, outages, etc...), I come back to SETI. Why I must been here if the thing is the same as in a 'real project'?

Depends if listening out for aliens is your thing, or if you get frozen out of SETI ;) The last part of which I've never tried to hide or deny, or give another toss about... , just in case anyone was wondering...

I'm going to increase the WU queue (I bet that'll make a lot of people happy). The newton methods we're running right now aren't as sensitive to the work queue as the genetic search was.

Thanks, Travis! :-)))

Btw, I don't think this will change the strange work fetch policy of recent BOINC manager versions though. :-/
There is always one last WU crunching while the other core(s) are idling before the darn thing will request more work, grrrr!

Well... Please make your complains here...

I think Logan hates us :(

Looks like I got some Milkyway spillage into SETI.
http://setiathome.berkeley.edu/result.php?resultid=1131907775

Server side it looks like we're crunching more results than before the credit reduction... around 2.9-3.0 a second.

Thanks for the heads up there. I guess if you have the patience to go along with it until the Autumn, so can we. :-)

We've survived this far, I think we can make it :D

If this last application update does what I think it should, the application should be stable for awhile and i'll be able to work on the automatic error checking, which will probably make a few people who don't like frequent updates a bit happier.

Starting a new thread for people to let out all their credit rage :)

Seems to have worked ;-)

Travis you say this is an alpha project and if we can't handle it go elsewhere! I think maybe you need to start looking for some common sense. In alpha there are bound to be problems with the "Official" application amoung platforms, servers running properly, assimilators/validators working correctly and workunit production and distribution. Credit is relative.

While I obviously can't speak for Travis, I know that I'm not talking about the credit level when I talk about people not running this project until it is beta. What I'm seeing are a lot of people who are getting upset at the fact that they're having to update clients frequently or that there are crashes or other performance issues with the newer versions. If one was using only the stock application, there would not be any effort invovled in updating the application, as BOINC would take care of it. It is those of us who read these forums and know how to get an anonymous platform (app_info.xml) version running who would be having to spend extra effort. Does that give us the right to kick and scream when things don't go perfectly? I don't think it does, not when the project has said that they do not have access to a lot of systems, which is why the project is "alpha", so that those of us who test running the software on architectures they don't have can help them get to "beta" and then to a full stable release.

The credit discussion is something that is a no-win, so long as it goes the same way as all the other discussions on the same subject have gone in the past. As I said, the best thing to do is to disassociate projects from each other by getting rid of the shared currency ("cobblestone") and let each project come up with their own scoring. After that, no longer compare between projects unless some independent external entity (not anyone from BOINC / UCB) comes up with a composite index / exchange rate.

I'm done here until there is a level playing field as far as the apps go. The windows users are at a disadvantage as far as I can see. If there is an optimized app for them, then how about a sticky with links to the download site for all the optimized apps for each OS including specific instruction sets.

I've made links sticky when someone has provided an optimized app and posted them on the forums (see the optimized linux apps in the code discussion forum). No one has posted anything for windows so there's nothing for me to sticky and verify.

The solution for everything here is "we're still Alpha". So what in your terms constitutes it as a Beta? The same stock app was used for a couple months with no problems untill the last week or two. Isn't that what Beta is? Running consistently for given periods of time without many problems? Heck that alone beat numerous projects that can't go a week without having errors.

And, to the comparation, in SETI, with AKV8 SSSE3 opti app I can get more than 35 - 40 credits/hour/core with an Q6600. (for to have an easy point of reference)

And with the same sse3 at milkyway you'll receive about 100!
So I would say the credits need to be halved another time!
And they would still be above the seti-level.

Travis must to define openly his credits politics...

Fixed credits...? There are not... were changed every time than an optimizer volunteer gave him a hint about how to improve/fix his faults...

And every time is the half or more... (and the credits core/hour limit, the WUs in cache limit... etc)

Things never saw in anyone project...

He wants to compare his credits with SETI, but forget than SETI don´t have limits. I would like to see what could happen in SETI if they had set this limits...:D :D :D :D

And, to the comparation, in SETI, with AKV8 SSSE3 opti app I was getting more than 35 - 40 credits/hour/core with an Q6600 in my SETI times. (for to have an easy point of reference)

These versions ran, they just had performance fluctuations due to architectural differences (something I can't test very well considering our somewhat limited hardware).

What do you need?

You have an entire group of participants here that can test for you... set up a test environment, and let us test it out before implementing it.

I don't know if they have a "test environment". They might could consider marking versions as "test applications" and then people can go to their BOINC preferences and check the box authorizing / enabling running test applications.

However, like I just said in my other post, if the project is stated as "alpha", then people should be expecting potential problems, some of them severe, and the potential for rapidly changing versions. If anyone is unwilling or unable to deal with that scenario, they should consider waiting until the application / project is moved out of alpha and into beta or to GA.

AND QUIT TESTING on the production side!

That's what the people who came up with the last bunch of opti apps got ripped for not testing. Yet as I mentioned before travis did not test any of the versions before releasing .08-.14 and they kept messing computers up for a couple versions.

These versions ran, they just had performance fluctuations due to architectural differences (something I can't test very well considering our somewhat limited hardware).

No one has posted an optimized windows application for me to approve. I can't do anything in this matter -- it's up to the people who compile the windows apps and make them available. If they post an optimized application on one of our forums I'd be more than happy to verify it and sticky it for everyone.

You guys do all the work for us, we'll just kick back and analize what you've done for us and cut your credits in half for all your great efforts.

I think you may find the numer of participants in your project cut in half as well.

I personally cannot make optimized windows applications because I don't have a windows machine nor a compiler that will do it. I have an OSX machine so I do my best to keep versions for that as optimized as possible.

The fact of the matter is we don't have hardware/compilers for everything out there, so we can't make a bunch of optimized applications for you guys. Thankfully a bunch of our users are generous enough with their time to make optimized applications available for you to use. I don't have any problem putting our seal of approval on them for the rest of you to use. I'm not quite sure what else you want me to do here.

Again, BOINC as a whole is a community, and if any project is generating significantly more credit for the same amount of work than the rest of them (which I think many people would agree the stock milkyway app was doing), this isn't working towards the grander scheme of BOINC at the moment, nor are we being fair to the other projects out there.

Also, arguments that we're just trying to keep Dave Anderson happy are probably directed at the wrong person, because he's on my PhD thesis committee :P