Message boards :
Number crunching :
New computer on the way
Message board moderation
Previous · 1 . . . 14 · 15 · 16 · 17 · 18 · 19 · Next
Author | Message |
---|---|
Wiggo Send message Joined: 24 Jan 00 Posts: 35109 Credit: 261,360,520 RAC: 489 |
..and also loosing the gauge to judge performance by:-(...We lost that ability years ago with the intro of CreditScrew. :-( Cheers. |
Kevin Olley Send message Joined: 3 Aug 99 Posts: 906 Credit: 261,085,289 RAC: 572 |
..and also loosing the gauge to judge performance by:-(...We lost that ability years ago with the intro of CreditScrew. :-( I know its totally useless between different projects, but most of the credit systems are. But it seems to be getting even worse here which could make it even harder to judge between performance enhancements etc. Kevin |
Wiggo Send message Joined: 24 Jan 00 Posts: 35109 Credit: 261,360,520 RAC: 489 |
The thing is that before CreditScrew Einstein credits wern't that much above what we were getting here, but since CreditScrew they have grown further apart and that difference is accelerating by the day. Cheers. |
Grant (SSSF) Send message Joined: 19 Aug 99 Posts: 13765 Credit: 208,696,464 RAC: 304 |
In the most recent threads about Credit I put forward a proposal that resolves all of the current issues with Credit New, but it's been pretty much ignored by every one. So for all the noise in this thread about the poor payout of Credit, It can't be that much of a concern if no one was prepared to point out any faults with my proposal, or support it's implementation. Grant Darwin NT |
Sirius B Send message Joined: 26 Dec 00 Posts: 24882 Credit: 3,081,182 RAC: 7 |
CreditNew is like England in the World Cup - won it once 52 years ago & getting screwed ever since :-( |
Keith Myers Send message Joined: 29 Apr 01 Posts: 13164 Credit: 1,160,866,277 RAC: 1,873 |
In the most recent threads about Credit I put forward a proposal that resolves all of the current issues with Credit New, but it's been pretty much ignored by every one. I posted to the thread about your proposal and that I considered it well thought out and easy to implement. But the reality is that we have no control over CreditScrew. That is totally under the control of the administrators and they have been absolutely silent on the matter any time the credit mechanism has been brought up. So, I don't know what you really expect in response to your proposal. I see fat chance of it ever being implemented until key project personnel are gone and new people are brought on board with a more open attitude towards change. Seti@Home classic workunits:20,676 CPU time:74,226 hours A proud member of the OFA (Old Farts Association) |
Grant (SSSF) Send message Joined: 19 Aug 99 Posts: 13765 Credit: 208,696,464 RAC: 304 |
So, I don't know what you really expect in response to your proposal. We've complained about how stuffed Credit New is for years, but have never offered an alternative (other than going back to the previous FLOPS counting version which we've been told just isn't going to happen). What I've proposed doesn't require that. It addresses all of the issues we have with Credit New. It addresses all of the stated goals of Credit new (bar one- Cherry picking penalisation which is better done another way) . It requires no more of a project starting up (or an existing project) to determine the Credit to be paid out for work done than is presently required. It also makes it possible for future versions of the BOINC manager to re-allocated work between GPU, CPU, eQPU (external Quantum Processor, or whatever the next big processing in thing is) after it's been downloaded & still maintain caches & not affect runtime estimates, credit claims & allocations (and whatever else is affected by rescheduling at this time). What i'd hoped for was for those that have issues with the current system would have pointed out any short comings so they could be addressed, or agree that as it stands it is the system we should have. Then instead of people just complaining abut Credit New, they could also point out that there is an alternative that not only would work, but also addresses all (bar one) of Credit News's stated goals. If people feel it is workable, and will do what it should, then someone could present it to the BOINC development group. They're the ones to convince, but it's better to give them something that has been checked over by many people with different perspectives, than just one or 2. Grant Darwin NT |
TBar Send message Joined: 22 May 99 Posts: 5204 Credit: 840,779,836 RAC: 2,768 |
Actually, there is a very simple way to adjust the BLC credit to reflect Arecibo scores. By now it is apparent the difference is about 100%, you score twice as much credit when running non-vlar Arecibo tasks. So, when the task begins with blc... simply add a x2 to the end of the credit equation. You still get all the warts associated with CreditFewer, but, the different tasks will award credit much more in line with the task run-times. The way it is now when Arecibo tasks are available it raises the APR very slowly, by the time the RAC starts to rise the tasks are finished and the APR starts to fall again. Receiving the same credit for both tasks will stabilize RAC a little, and result in a much higher RAC. It would be much like it was when running just Arecibo tasks, and most people will be happier. All it would take is adding a simple x2 to the equation. Naw...to simple. |
betreger Send message Joined: 29 Jun 99 Posts: 11366 Credit: 29,581,041 RAC: 66 |
One idea I have not seen floated is 2 separate credit schemes, such as credit new and run simultaneously another credit system. That way you can pick whatever floats your boat. |
Kevin Olley Send message Joined: 3 Aug 99 Posts: 906 Credit: 261,085,289 RAC: 572 |
In the most recent threads about Credit I put forward a proposal that resolves all of the current issues with Credit New, but it's been pretty much ignored by every one. Hang on a minute, I cannot speak for the others here but you are talking to a lorry driver who has absolutely no IT skills or understanding of what was being said in that thread, Yes I know there is a problem with creditnew but I could not fault or agree with your proposal because I would not do that to something that I do not understand. Some people are good at some things and not so good at others, and I can tell you that there are a lot of people that I would not like to see trying to do the work that I do:-) Kevin |
Grant (SSSF) Send message Joined: 19 Aug 99 Posts: 13765 Credit: 208,696,464 RAC: 304 |
All it would take is adding a simple x2 to the equation. And it doesn't address the low amount of Credit awarded for Arecibo tasks, and it doesn't address the variability of Credit awarded depending on hardware, OS, driver & application, and doesn't address cross project comparisons. Instead of implementing another kludge, that is just a patch & doesn't actually fix anything, why not fix the whole problem? Especially when it's as simple as paying out credit in accordance with the stated definition of the Cobblestone? That's all that is required to fix everything, not just one of the many issues. Grant Darwin NT |
Grant (SSSF) Send message Joined: 19 Aug 99 Posts: 13765 Credit: 208,696,464 RAC: 304 |
Hang on a minute, I cannot speak for the others here but you are talking to a lorry driver who has absolutely no IT skills or understanding of what was being said in that thread, Yes I know there is a problem with creditnew but I could not fault or agree with your proposal because I would not do that to something that I do not understand. And i'm just a Postie & barely could wrap my had around much of it, and a lot of it I couldn't (the way a Cobblestone is defined really does make it hard to figure things out- at least for me). But by going back to the very basics (extremely vague memories of high school physics), and looking at the Credit system that gave us the least grief compared to the current one I managed to figure out what was meant to be happening, and how it could be done, in the simplest possible way (because that was the only way I could figure it out- simply). Grant Darwin NT |
Grant (SSSF) Send message Joined: 19 Aug 99 Posts: 13765 Credit: 208,696,464 RAC: 304 |
One idea I have not seen floated is 2 separate credit schemes, such as credit new and run simultaneously another credit system. That way you can pick whatever floats your boat. Because things are already complicated enough, why screw them up even further? Why not just fix the current system, when it's (relatively) an easy & straight forward thing to do? Grant Darwin NT |
TBar Send message Joined: 22 May 99 Posts: 5204 Credit: 840,779,836 RAC: 2,768 |
As I said, you still get the warts. Good luck getting them to use a completely different credit system, we've seen how well that works. A simple x2 to is very easy to apply to the current system, and is just as easy to remove. I would suggest doing it without any announcement, and certainly well before the WOW event...which isn't that far away. |
Kevin Olley Send message Joined: 3 Aug 99 Posts: 906 Credit: 261,085,289 RAC: 572 |
But the reality is that we have no control over CreditScrew. That is totally under the control of the administrators and they have been absolutely silent on the matter any time the credit mechanism has been brought up. That's probably more the problem, If they even bother to view the forums they see the thread title and move onto the next one:-( Kevin |
Keith Myers Send message Joined: 29 Apr 01 Posts: 13164 Credit: 1,160,866,277 RAC: 1,873 |
They in fact don't bother to read the forums. The only time they post to the forums is in announcing some promotional fluff about the project. And the comment about promoting the proposed CreditNew scheme to the BOINC steering committee would fall on deaf ears anyway. All that committee seems to be concerned with is getting the BOINC platform onto as many devices as possible. So development is only on BOINC hardware compatibility. Nothing on the credit mechanism. That code is locked away in the vault never to be brought out again to the light of day. Seti@Home classic workunits:20,676 CPU time:74,226 hours A proud member of the OFA (Old Farts Association) |
tullio Send message Joined: 9 Apr 04 Posts: 8797 Credit: 2,930,782 RAC: 1 |
I happen also to run GPUGRID tasks. GPU tasks get huge credits, CPU tasks, which run a longer time, a few bucks. Every project has its own rules. Tullio |
Wiggo Send message Joined: 24 Jan 00 Posts: 35109 Credit: 261,360,520 RAC: 489 |
..A simple x2 to is very easy to apply to the current system, and is just as easy to remove...It's not that simple because you're not addressing the factor that causes the constant ongoing devaluation that's been happening. All you're suggesting is a quick and dirty band-aid job. ;-) Cheers. |
TBar Send message Joined: 22 May 99 Posts: 5204 Credit: 840,779,836 RAC: 2,768 |
What I'm suggesting is People stop pounding their heads against the brick wall. Instead of trying to change the D.A.'s pet App, just convince the Project Scientist to make a Project tweak to the Pet. One simple x2 at the end of the equation and the Project is back to the same credit awards it had before the introduction of the BLC tasks. That solution is about twice as good as the current situation, and actually might be possible. Or, you can just keep pounding your head(s). |
Stargate (SA) Send message Joined: 4 Mar 10 Posts: 1854 Credit: 2,258,721 RAC: 0 |
I've put pillows on my walls not so bad now :) |
©2024 University of California
SETI@home and Astropulse are funded by grants from the National Science Foundation, NASA, and donations from SETI@home volunteers. AstroPulse is funded in part by the NSF through grant AST-0307956.