does using an optimized boinc client to boost your benchmarks so you can get more credit hurt your results? or does rosetta still get the same amount of use out of the results?

As long as you're producing error free WUs, then the credits make no difference to the science.

---

Please refer this thread: WOW! Where can I buy this machine!

And this post on a thread of the board of BOINC Synergy, a team of BOINC crunchers, said that,

using an optomized BOINC without using an optomized project application is merely inflating your credit requeste - a form of credit cheating.

(BOINC Synergy, Optimized version of 5.4.9, John McLeod VII, Posted: Sun Jun 04, 2006 6:46 pm.)

As he is a volunteer developer of the BOINC client, he seems worthwhile to be trusted. So we may say that: without optimised applications of Rosetta, using the optimised client can be regarded as a way of cheating.

But cheating itself don't hurt the results. Develpers can get genuine data. So Rosetta doesn't produce false scientific results from them.

Anyway, haven't developers of this project changed a way of counting credit, not to only FLOPS reported by the client influence the credit, already? Perhaps there was a news on this point on another thread of this project, but I'm not sure. Could someone notice its URL? Thanks.

---

No it does not in any direct way effect the results of the rosetta science appliaction. Credit/work done is just a 'number'.



In an indirect way though it does cause problems.
- No real idea of actual computing power (total FLOPS)
- Causes scheduler problems for yourself until it smooths itself out
- Cannot compare between boinc projects
- Can cause credit fovouring towards specific projects for 'overall Boinc credit hunters'
- Bickering in the Rosetta community ;-)
- ...

But these are mute points for now since
a) they are in widespread use
b) a new system is in the works

---

Team mauisun.org

No it does not in any direct way effect the results of the rosetta science appliaction. Credit/work done is just a 'number'.



In an indirect way though it does cause problems.
- No real idea of actual computing power (total FLOPS)
- Causes scheduler problems for yourself until it smooths itself out
- Cannot compare between boinc projects
- Can cause credit fovouring towards specific projects for 'overall Boinc credit hunters'
- Bickering in the Rosetta community ;-)
- ...

But these are mute points for now since
a) they are in widespread use
b) a new system is in the works

Not so 'mute' what with all the credit bickering.

But they are 'moot' points in regards to the science.

(the word usage and spelling police are leaving now)

---

Proudly Banned from Predictator@Home and now Cosmology@home as well. Added SETI to the list today. Temporary ban only - so need to work harder :)



"You can't fix stupid" (Ron White)

I thought moot was an undecided/unconcluded idea and mute gets used when it's an unimportant wrt the rest of it. (it's quiet therefore gets drowned out).


So with my understanding it is certiainly moot, though that's not was I was meaning. I was meaning unimportant in the larger scheme of things.

Of course I could well be wrong.




<lol, I just puy 'moot mute' in to google, from a scan of the summary some people seem to care a bit to much about language usage, I may read some of them later :-)>

---

Team mauisun.org

First link in the search says my understanding of moot is correct
http://www.worldwidewords.org/qa/qa-moo1.htm
[Q] From Nancy Maclaine: “Did the phrase a moot point originally mean ‘a debatable point’? Nowadays it seems to mean ‘an irrelevant point’ or even ‘a point so irrelevant it’s not worth debating’. Some actually have taken to referring to it as a mute point. What’s the history here?”

[A] Moot point is one of those phrases that once had a firm and well-understood meaning, but no longer does. It was just as you say: a matter that was uncertain or undecided, so open to debate.

It comes from the same source as meet and originally had the same meaning. In England in medieval times it referred specifically to an assembly of people, in particular one that had some sort of judicial function, and was often spelled mot or mote. So you find references to the witenagemot (the assembly of the witan, the national council of Anglo-Saxon times), hundred-mote (where a hundred was an Anglo-Saxon administrative area, part of a county or shire), and many others. So something that was mooted was put up for discussion and decision at a meeting—by definition something not yet decided.

The confusion over the meaning of moot point is modern. It is a misunderstanding of another sense of moot for a discussion forum in which hypothetical cases are argued by law students for practice. Since there is no practical outcome of these sessions, and the cases are invented anyway, people seem to have assumed that a moot point means one of no importance. So we’ve seen a curious shift in which the sense of “open to debate” has become “not worth debating”.

The mute spelling is a development that has come about because moot is now a fossil word, usually encountered only in this phrase; there is an understandable tendency to convert the unknown into the known, and mute seems to fit the new meaning rather better. But it’s wrong.

and err, well I guess I was using mute metaphorically

---

Team mauisun.org

Please refer this thread: WOW! Where can I buy this machine!

And this post on a thread of the board of BOINC Synergy, a team of BOINC crunchers, said that,

using an optomized BOINC without using an optomized project application is merely inflating your credit requeste - a form of credit cheating.

(BOINC Synergy, Optimized version of 5.4.9, John McLeod VII, Posted: Sun Jun 04, 2006 6:46 pm.)

As he is a volunteer developer of the BOINC client, he seems worthwhile to be trusted. So we may say that: without optimised applications of Rosetta, using the optimised client can be regarded as a way of cheating.

But cheating itself don't hurt the results. Develpers can get genuine data. So Rosetta doesn't produce false scientific results from them.

Anyway, haven't developers of this project changed a way of counting credit, not to only FLOPS reported by the client influence the credit, already? Perhaps there was a news on this point on another thread of this project, but I'm not sure. Could someone notice its URL? Thanks.

Well tell him to get with the program and DEVELOP a damn Windows 64 bit Boinc client! Hellllo! 3 years ago 64 bit cpu's came out.. whats the holdup? He brags about being a developer, lets see him develop us a 64 bit client...

---

SETI...Grown Men Looking For ET's That Dont Exist... Do You Dress Up Like Starwars Characters Too?

JM7 is both loved and hated. He wrote the work scheduler that's in everyones version of boinc(boinc versions 4.35 to present). It's the LTD, STD, work fetch, etc part of the program. Rom Walton is in charge of it's overall development (under the direction of Dr. Anderson, ofcourse). It's my understanding that a 64 bit application will work with boinc, even if boinc is only 32 bit. Boinc takes up less than 1% of your cpu as it only manages the project application. Your plea for a 64 bit boinc should be made to Rom.

Well tell him to get with the program and DEVELOP a damn Windows 64 bit Boinc client! Hellllo! 3 years ago 64 bit cpu's came out.. whats the holdup? He brags about being a developer, lets see him develop us a 64 bit client...

Why you wan't a 64 bit version ? - As Tony has written, it did not make any difference, because the boinc client do not any performance intensive tasks.

---

It's my understanding that a 64 bit application will work with boinc, even if boinc is only 32 bit.

Yes, Tony you're right. I have successfull running a boinc 32 bit with a SETI and SIMAP apps in 64 bit on HPUX.

Boinc takes up less than 1% of your cpu as it only manages the project application. Your plea for a 64 bit boinc should be made to Rom.

There will be no great benefit from a 64 bit boinc client. - It could also be that a 64 bit application running slower as an 32 bit one, but it depends on the application.
I think it could be more worthy to port a science app to 64 bit, if the performance in 64 bit will be better.

---

Please refer this thread: WOW! Where can I buy this machine!

And this post on a thread of the board of BOINC Synergy, a team of BOINC crunchers, said that,

using an optomized BOINC without using an optomized project application is merely inflating your credit requeste - a form of credit cheating.

(BOINC Synergy, Optimized version of 5.4.9, John McLeod VII, Posted: Sun Jun 04, 2006 6:46 pm.)

As he is a volunteer developer of the BOINC client, he seems worthwhile to be trusted. So we may say that: without optimised applications of Rosetta, using the optimised client can be regarded as a way of cheating.

But cheating itself don't hurt the results. Develpers can get genuine data. So Rosetta doesn't produce false scientific results from them.

Anyway, haven't developers of this project changed a way of counting credit, not to only FLOPS reported by the client influence the credit, already? Perhaps there was a news on this point on another thread of this project, but I'm not sure. Could someone notice its URL? Thanks.

Well tell him to get with the program and DEVELOP a damn Windows 64 bit Boinc client! Hellllo! 3 years ago 64 bit cpu's came out.. whats the holdup? He brags about being a developer, lets see him develop us a 64 bit client...

Why don't you get the source and compile it (and do the needed alterations) for 64bit ?

I guess isnce there are only 3(4) official platforms Windows 32bit, Linux 32bit (ok so there are now 3 target linux distro) and MacOS-X (Intel and PPC) then the fact this takes the small handful of developers there time, money and getting the compatability issues sorted out, let alone please everybody on these setup (of course don't forget the server side end as well) all their current time and money they could do with the help.

Crunch3r did compile a 64bit BOINC SSE3 client, so it is possible.

---

Team mauisun.org

Is there ANY reason why the BOINC client itself needs to be 64-bit?

As far as I can tell, the only part of BOINC itself that has any effect in the scenario is the benchmark - which is more or less broken for Rosetta anyways...

I haven't looked at the code tho', but from what I understand, all BOINC itself does is some form of managed execution of another application with some [many] argument(s) passed along, including a file that has been downloaded. All of this can be done in 32-bit, even if the corresponding executed application is 64-bit.

The only case I can think of where this is NOT a suitable proposal would be on an architecture where the 64-bit OS isn't supporting 32-bit apps (or are running 32-bit apps EXTREMELY slowly)... Linux & Windows certainly don't suffer from this (on x86), and I beleive all MacOS X apps are written to work on the relevant PPC/x86 archs that the OS supports in some natively supported form - so that shouldn't matter... So can someone please explain what use this would be?

Don't get me wrong, I'm all for getting things to work better - I just fail to see how this is improving ANYTHING - at least with respect to Rosetta.

--
Mats

---