My Linux boxes say a Good Bye to Einstein@Home.

M. Schmitt
M. Schmitt
Joined: 27 Jun 05
Posts: 478
Credit: 15872262
RAC: 0

Same here, the Win Cliets

Same here, the Win Cliets will stay and Linux will go.

But I also expect a new Linux app that is comparable to the alberts, otherwise I will not come back. And because I am going to swich all my hosts to Linux, Win just for gaming or development if asked for, this can mean a total cut.

But surely we will get a good opimized Linux client too. :)

cu,
Michael

networkman
networkman
Joined: 22 Jan 05
Posts: 98
Credit: 7140649
RAC: 0

RE: RE: Of course, if

Message 27479 in response to message 27475

Quote:
Quote:
Of course, if your philosophy allows for no MS software at all - well, that's an entirely different matter.

It is not a matter of philosophy, it is just a whole bulk of software which is running on the box. I do not want to spend Euros over Euros just to be able to switch from free software to whatever paysoft.

Currently windows boxes do a better job, so let them do the job. Time will come and Bruce will provide a faster Linux client, I have plenty of time to wait.

While I'm not suggesting that you have to switch, my point of view follows closely to the saying that "There's no such thing as a free lunch." I've found that saying to be true throughout my life experience, which is why in my mindset it is a philosophical difference.

Regardless, I wish you happiness and increased productivity wherever you go with your Linux boxes.

"Chance is irrelevant. We will succeed."
- Seven of Nine

B52
B52
Joined: 19 Feb 05
Posts: 45
Credit: 273899
RAC: 0

My few cents worth

My few cents worth here.

If its about credits, then I cannot relate to your objection.
If its about completing the science bit faster, then I cannot relate either.

If its about that windows boxes complete it faster atm, yes then I can relate.

I know, elaboration is needed.

About credits. The optimized crunchers complete a wu much faster, hence claiming FAR FAR less credit. That is unless you cheat, and use something like trux calibrating CC. In that case, an optimized cruncher will claim the same as the standard after adjusting. That could be considered as cheating. But as I complete a wu in xxx secs and the standard CC claims credits for what I actually have spent, and yours do exactly the same. Then I cannot see the difference.

With regards to completing the science bit faster, then I cannot see why anyone should object for any reason what so ever.

If its about the promise that an optimized Linux cruncher should be there shortly, yes then I defo see your disappointment.

Have been rambling enough now.

Gray Handcock
Gray Handcock
Joined: 11 Mar 05
Posts: 211
Credit: 135567
RAC: 0

Yup me too - I went through

Yup me too - I went through the whole compiling thing with Gentoo (around 3 days in total) - (still my favourite Linux due to fiddle-ability) in the hopes of getting an "edge".

At the end of the day, windows xp (for now) runs a faster client thanks to Akosf's efforts (well done, that man !) despite an antivirus and a software firewall (and Skype) in Windows.

I am sure that when the linux app gets optimised, Linux should be faster on the same hardware, but for now I gotta go back to XP :(

Gray

best wishes tae ya all - and a glass of red wine raised to Akosf !

M. Schmitt
M. Schmitt
Joined: 27 Jun 05
Posts: 478
Credit: 15872262
RAC: 0

RE: My few cents worth

Message 27482 in response to message 27480

Quote:

My few cents worth here.

If its about credits, then I cannot relate to your objection.

Win boxes get a lot more credits.

Quote:

If its about completing the science bit faster, then I cannot relate either.

It seems to be a waste of energy and costs, if the same host with a different OS crunches 4 times faster.

Quote:

If its about that windows boxes complete it faster atm, yes then I can relate.

I know, elaboration is needed.

About credits. The optimized crunchers complete a wu much faster, hence claiming FAR FAR less credit. That is unless you cheat, and use something like trux calibrating CC. In that case, an optimized cruncher will claim the same as the standard after adjusting. That could be considered as cheating. But as I complete a wu in xxx secs and the standard CC claims credits for what I actually have spent, and yours do exactly the same. Then I cannot see the difference.

Imho that is absolutely no cheating, but being fair to the other people who do not use a turboalbert. Otherwise, if one regular albert meets two fast ones, the former will get a ulta low credit. Furthermore every WU of the same size should get about the same credits independent of the CPU, the OS and the time it took to finish. It should be no difference between crunching more WUs because of overclocking or using some faster app.

Quote:

With regards to completing the science bit faster, then I cannot see why anyone should object for any reason what so ever.

I see this quite different.

Quote:

If its about the promise that an optimized Linux cruncher should be there shortly, yes then I defo see your disappointment.

Have been rambling enough now.


I think you should think again about your point of view.

cu,
Michael

Wurgl (speak^Wcrunching for Special: Off-Topic)
Wurgl (speak^Wc...
Joined: 11 Feb 05
Posts: 321
Credit: 140550008
RAC: 0

RE: If its about that

Message 27483 in response to message 27480

Quote:
If its about that windows boxes complete it faster atm, yes then I can relate.

Something like that.

My box can help another project more, when I relatively compare it with Windows. In other projects it is a lightning¹ fast machine, in this project it cannot even beat a uniprozessor machine, build in 1999.

¹) Actually one year ago it was :-(

So a WU I crunch here is much more expensive (in terms of electrical bill) compared to the other operating system, whereas it is as expensive in other projects.

archae86
archae86
Joined: 6 Dec 05
Posts: 3145
Credit: 7023504931
RAC: 1809149

RE: Imho that is absolutely

Message 27484 in response to message 27482

Quote:
Imho that is absolutely no cheating, but being fair to the other people who do not use a turboalbert. Otherwise, if one regular albert meets two fast ones, the former will get a ulta low credit.

This case is actually happening now. I've scanned my "results for computer" data and examined at least half a dozen result listings for which the granted credit was outlier low compared to the claimed credit. In every one of those cases, the lowest claimer was running an enhanced ap from akosf with a stock client. Usually the middle ap was also an akosf ap, though a few were Mac's.

I agree with Ziegenmelker that in the real case of Einstein@home and the akosf science aps operating in a mixed environment with the current distributed Albert science ap, use of trux's Calibrating Client contributes to more fairness, and is in no sense whatever cheating.

The counter-productive effects of compensation systems based on effort rather than output are pretty well known in economics.

B52
B52
Joined: 19 Feb 05
Posts: 45
Credit: 273899
RAC: 0

RE: RE: My few cents

Message 27485 in response to message 27482

Quote:
Quote:

My few cents worth here.

If its about credits, then I cannot relate to your objection.

Win boxes get a lot more credits.
Quote:

If its about completing the science bit faster, then I cannot relate either.
It seems to be a waste of energy and costs, if the same host with a different OS crunches 4 times faster.
Quote:

If its about that windows boxes complete it faster atm, yes then I can relate.

I know, elaboration is needed.

About credits. The optimized crunchers complete a wu much faster, hence claiming FAR FAR less credit. That is unless you cheat, and use something like trux calibrating CC. In that case, an optimized cruncher will claim the same as the standard after adjusting. That could be considered as cheating. But as I complete a wu in xxx secs and the standard CC claims credits for what I actually have spent, and yours do exactly the same. Then I cannot see the difference.

Imho that is absolutely no cheating, but being fair to the other people who do not use a turboalbert. Otherwise, if one regular albert meets two fast ones, the former will get a ulta low credit. Furthermore every WU of the same size should get about the same credits independent of the CPU, the OS and the time it took to finish. It should be no difference between crunching more WUs because of overclocking or using some faster app.
Quote:

With regards to completing the science bit faster, then I cannot see why anyone should object for any reason what so ever.

I see this quite different.
Quote:

If its about the promise that an optimized Linux cruncher should be there shortly, yes then I defo see your disappointment.

Have been rambling enough now.


I think you should think again about your point of view.

cu,
Michael

Hi Michael

Did'nt wanna cut in all off this, but as I see your objections, then its not at all about a turboalbert, but about the Boinc CC ??

Meaning that in your opinion, then the CC should give equality to all.

Jim Bailey
Jim Bailey
Joined: 31 Aug 05
Posts: 91
Credit: 1452829
RAC: 0

The credit system is a

The credit system is a steaming pile laying out in the pasture, and always has been. It tells me nothing about what my systems are doing on a daily basis. I'd gladly take one point/credit for each WU that validates, at least that way you can see what kind of actual work per day your systems are doing.

Check my boxes out, before the new app mine claimed high and those numbers got tossed. Now, they claim low, and that gets tossed. Do the numbers, RAC and credit, tell you or me anything about how many WU's my boxes actually do? Nope, not a bit! They are just "feel good" number and have no relationship to what is actually being done.

Anyone who wants my credits or RAC can have them, because they mean nothing. I'm here to crunch as many WU's a day as I can to move the project along. Nothing more and nothing less!

B52
B52
Joined: 19 Feb 05
Posts: 45
Credit: 273899
RAC: 0

RE: The credit system is a

Message 27487 in response to message 27486

Quote:

The credit system is a steaming pile laying out in the pasture, and always has been. It tells me nothing about what my systems are doing on a daily basis. I'd gladly take one point/credit for each WU that validates, at least that way you can see what kind of actual work per day your systems are doing.

Check my boxes out, before the new app mine claimed high and those numbers got tossed. Now, they claim low, and that gets tossed. Do the numbers, RAC and credit, tell you or me anything about how many WU's my boxes actually do? Nope, not a bit! They are just "feel good" number and have no relationship to what is actually being done.

Anyone who wants my credits or RAC can have them, because they mean nothing. I'm here to crunch as many WU's a day as I can to move the project along. Nothing more and nothing less!

Well said m8, and I totally agree. I for one are in these project for the science, NOT the credits.

Comment viewing options

Select your preferred way to display the comments and click "Save settings" to activate your changes.