Message boards : Number crunching : Decision on points accumulated up to now
Message board moderation

To post messages, you must log in.

Previous · 1 · 2 · 3 · Next

AuthorMessage
EclipseHA

Send message
Joined: 18 Sep 04
Posts: 47
Credit: 1,886,234
RAC: 0
Message 2672 - Posted: 29 Sep 2004, 4:24:18 UTC

Just food for thought...

I joined LHC when the 1000-2000 user slots opened up.. Seemed to be you were doing a "staged opening".. I never saw anything thru the signup process or on the main pages that indicated this was still in beta.

Infact, it took me some time to find a beta reference on any of the main project pages (not the discussion forums) - news that had long since rolled off the main page.

As the user base is now open to 5000 users, do they know this is a beta? I didn't until today! Predictor had a "this is still Alpha and you might lose credits" notice on the main page - No question there - but with LHC, it's not been clear...


Anyway, the food for thought is this - there could be thousands of folks that see their LHC credits wiped and ask "why?". I sure don't think the 1000-5000 chunck of users understood this was beta (look at the forum posts about LHC within other projects and find one reference to 'beta"!)

It might be best for the 75% od LHC crunchers that had no clue this was beta, to just roll the current credits into the "official credits" and forget about beta page of honor, etc....

My 2c
ID: 2672 · Report as offensive     Reply Quote
Profile Doris and Jens

Send message
Joined: 1 Sep 04
Posts: 4
Credit: 498,638
RAC: 0
Message 2674 - Posted: 29 Sep 2004, 4:49:24 UTC - in response to Message 2575.  


> June 22, 2004
> Welcome to the new SETI@home! The project is now officially active. We've
> reset the project by deleting all results and workunits, and setting all
> credits (user, team, host) to zero. Thanks to everyone who participated in the
> alpha and beta tests for their help in debugging SETI@home and BOINC.

There is a little difference to the situation where Berkely reset the credits after the beta project. The credits where wrong there, because they where given when the system was under development. The scale and the formulars changed many times and so the credits didn't really represented the work done for the project. This is different with LHC today.

The BOINC credit system is still not working at the best, but over all reliable today. So - for my feeling - only the wish to start a new race at a cleared point will force to reset the cobblestones.

I personal didn't like the idea of resetting the credits. But I didn't like the idea of this hall of fame too. ;) Go on with or without resetting credits, but use your time for better things. :)


Greetings from Bremen/Germany
Jens Seidler
ID: 2674 · Report as offensive     Reply Quote
Profile Michal Krakowiak

Send message
Joined: 17 Sep 04
Posts: 6
Credit: 289,475
RAC: 0
Message 2724 - Posted: 29 Sep 2004, 13:39:38 UTC
Last modified: 29 Sep 2004, 13:41:54 UTC

I'd like to keep my credits, but also it would be nice having in account inscription "Beta test: 12345 credits" or something like this.
ID: 2724 · Report as offensive     Reply Quote
LP

Send message
Joined: 1 Sep 04
Posts: 39
Credit: 54,460
RAC: 0
Message 2726 - Posted: 29 Sep 2004, 13:53:34 UTC - in response to Message 2674.  


>
> There is a little difference to the situation where Berkely reset the credits
> after the beta project. The credits where wrong there, because they where
> given when the system was under development. The scale and the formulars
> changed many times and so the credits didn't really represented the work done
> for the project. This is different with LHC today.

This also was the case when LHC started.
Users work units were being validated against vastly different cpu's.

It doesn't matter to me one way or the other, although I think that clearing the points board will give a more accurate picture of points from here on.






ID: 2726 · Report as offensive     Reply Quote
Woyteck - Boinc Busters Poland

Send message
Joined: 2 Sep 04
Posts: 9
Credit: 8,705
RAC: 0
Message 2734 - Posted: 29 Sep 2004, 14:24:46 UTC

Before they reset credits, they should deal with every credit raleted problems.

windows - linux validations , overclocked cpu's and some other rather major problems with LHC client...

ID: 2734 · Report as offensive     Reply Quote
Profile sysfried

Send message
Joined: 27 Sep 04
Posts: 282
Credit: 1,415,417
RAC: 0
Message 2747 - Posted: 29 Sep 2004, 16:52:03 UTC

@ LHC Admins!

If you want to reset scores to Zero, you'll have to make sure no client has WU's anymore (by either sending out completely new WU's and renaming them, or any other way you choose). If you just set all scores to zero, any workunit completed at that timepoint but not sent to the server will change the scores at relaunch.

I guess you get my point. :-)

Sincerely,

sysfried
ID: 2747 · Report as offensive     Reply Quote
Profile tiker
Avatar

Send message
Joined: 1 Sep 04
Posts: 34
Credit: 19,096
RAC: 0
Message 2749 - Posted: 29 Sep 2004, 16:59:49 UTC - in response to Message 2747.  

> If you want to reset scores to Zero, you'll have to make sure no client has
> WU's anymore (by either sending out completely new WU's and renaming them, or
> any other way you choose). If you just set all scores to zero, any workunit
> completed at that timepoint but not sent to the server will change the scores
> at relaunch.

Does that mean I can't increase my work unit cache size, download all the work units I can to crunch, and suspend my boinc network access until after the credits are set to 0? :)

If everyone does it, we'll get instant credit afterwards. :)

Disclaimer: Please note that this was a sarcastic post and was posted for the admins to consider. I have no intention to actually do as I have posted above.

---
ID: 2749 · Report as offensive     Reply Quote
Profile sysfried

Send message
Joined: 27 Sep 04
Posts: 282
Credit: 1,415,417
RAC: 0
Message 2751 - Posted: 29 Sep 2004, 17:14:16 UTC - in response to Message 2749.  

> > If you want to reset scores to Zero, you'll have to make sure no client
> has
> > WU's anymore (by either sending out completely new WU's and renaming
> them, or
> > any other way you choose). If you just set all scores to zero, any
> workunit
> > completed at that timepoint but not sent to the server will change the
> scores
> > at relaunch.
>
> Does that mean I can't increase my work unit cache size, download all the work
> units I can to crunch, and suspend my boinc network access until after the
> credits are set to 0? :)
>
> If everyone does it, we'll get instant credit afterwards. :)
>
> Disclaimer: Please note that this was a sarcastic post and was posted for the
> admins to consider. I have no intention to actually do as I have posted
> above.
>
That's exactly my point. And I have my computer on with constant network access, so I'll just wait and see what happens
ID: 2751 · Report as offensive     Reply Quote
Profile Stephen Balch

Send message
Joined: 22 Sep 04
Posts: 42
Credit: 2,956
RAC: 0
Message 2794 - Posted: 29 Sep 2004, 21:42:37 UTC
Last modified: 29 Sep 2004, 21:53:01 UTC

IMHO, go ahead and reset the credits score to zero. Don't forget to flush the "Pending Credits," too. I have some 20 "Pending Credit" results, turned in before 29 Sept 2004 UTC, only one after midnight UTC.

You should be able to determine some period of time, perhaps twice the average calculation duration for project as based on beta data, before/at/after your local midnight for all "beta" results to be completely received, then, using the received time and date for the result records, give us some "production" credit for any results received after that time. I've turned in some 25 or more claimed credits for results since the "Beta Tester's Hall of fame: Total Credits" snapshot was taken, presumably near your local midnight.

Or, you could do something similar with the "sent" dates/times, but if you do this we need some way to flush the WU's that would not be counted as "production." If we don't have some method of flushing the WU's that won't count for "production," if we have to "detach" or "reset" the project, I would imagine there would be a lot of WU's that would need to be resent. Wasting time processing WU's that would not be counted for "production," or having to d/l additional WU's for production" processing because you have had to flush all the WU's in your cache, would probably vex a lot of people.

Something like this might eliminate any problems with WU's already in users caches.

[EDIT] I've just turned in several more results that have been processed just before and after Cerns local midnight. I had to disable BOINCs network access, not to move credits into today, but because my university has had some network problems. Some thing was "beaconing" and they couldn't seem to find out what. It's the third time in two days we've had this problem. [End EDIT]

Just some thoughts.

Cheers,

Stephen
ID: 2794 · Report as offensive     Reply Quote
LP

Send message
Joined: 1 Sep 04
Posts: 39
Credit: 54,460
RAC: 0
Message 2798 - Posted: 29 Sep 2004, 22:23:38 UTC

Well, I have around 450 pending(4702.30)right now, and a large portion of those are from the 28th and 29th. I hope they clear things soon, so I can feel like I'm not "wasting" more wu's and computer time. ;)
ID: 2798 · Report as offensive     Reply Quote
rsbriggs

Send message
Joined: 1 Sep 04
Posts: 14
Credit: 168,153
RAC: 0
Message 2887 - Posted: 30 Sep 2004, 17:15:06 UTC

I see that the scores haven't been reset yet.
Can we assume from this that they will not be reset?

I am easy either way just looking for clarification
ID: 2887 · Report as offensive     Reply Quote
ChinookFoehn

Send message
Joined: 17 Sep 04
Posts: 40
Credit: 293,269
RAC: 0
Message 2891 - Posted: 30 Sep 2004, 17:33:50 UTC - in response to Message 2887.  
Last modified: 18 Dec 2004, 8:19:18 UTC

ID: 2891 · Report as offensive     Reply Quote
Profile bjacke
Avatar

Send message
Joined: 2 Sep 04
Posts: 39
Credit: 4,910
RAC: 0
Message 2893 - Posted: 30 Sep 2004, 17:45:42 UTC

What will happen to the credit if we have send results and it is count to the old credit?


ID: 2893 · Report as offensive     Reply Quote
James R. Davis

Send message
Joined: 17 Sep 04
Posts: 4
Credit: 70,296
RAC: 0
Message 2894 - Posted: 30 Sep 2004, 17:55:19 UTC

Resetting credits would be a serious error on the part of the project administrators.

Impuning the motivations of those who provide idle computer resources in exchange for points instead of some altruistic desire to 'help science' is simply another form of predjudicial thinking.

Credits are like candy for many users, completed work unit results are like candy for the project administrators. They compliment each other and satisfy different needs, but they DO DO THAT - satisfy needs.

Teams exist and are extremely important to many users. Not, I suggest, because teams do 'more science', but because they provide frameworks upon which competition thrives and is measured. And the unit of measurement is CREDITS.

Resetting credits now is a slap in the face of those who have GIVEN their computer resources to the project during the beta test. They took significant RISK in behalf of the project. They also did it in order to earn credits.


ID: 2894 · Report as offensive     Reply Quote
ChinookFoehn

Send message
Joined: 17 Sep 04
Posts: 40
Credit: 293,269
RAC: 0
Message 2895 - Posted: 30 Sep 2004, 18:01:08 UTC - in response to Message 2893.  
Last modified: 18 Dec 2004, 8:19:37 UTC

ID: 2895 · Report as offensive     Reply Quote
Heffed

Send message
Joined: 2 Sep 04
Posts: 71
Credit: 8,657
RAC: 0
Message 2900 - Posted: 30 Sep 2004, 18:47:06 UTC - in response to Message 2894.  

> Resetting credits would be a serious error on the part of the project
> administrators.
>
> Impuning the motivations of those who provide idle computer resources in
> exchange for points instead of some altruistic desire to 'help science' is
> simply another form of predjudicial thinking.
>
> Credits are like candy for many users, completed work unit results are like
> candy for the project administrators. They compliment each other and satisfy
> different needs, but they DO DO THAT - satisfy needs.
>
> Teams exist and are extremely important to many users. Not, I suggest,
> because teams do 'more science', but because they provide frameworks upon
> which competition thrives and is measured. And the unit of measurement is
> CREDITS.
>
> Resetting credits now is a slap in the face of those who have GIVEN their
> computer resources to the project during the beta test. They took significant
> RISK in behalf of the project. They also did it in order to earn credits.

Resetting during beta is a very common occurance. Deal with it...

ID: 2900 · Report as offensive     Reply Quote
Kenneth Larsen

Send message
Joined: 2 Sep 04
Posts: 36
Credit: 90,806
RAC: 0
Message 2909 - Posted: 30 Sep 2004, 20:07:39 UTC - in response to Message 2891.  

> Seeing that a number of the 'leaders' lost thousands of points yesterday;
> could it be the powers-that-be have decided to eliminate the alpha phase
> points only?
>
> If one were one of the original 1,000; it might be possible to determine if
> your first 2 week's credits still are extant.
>
> -H. Richard Utzig
>

I do not think this is what has happened, as I was one of the original 1000 that joined, and I haven't seen any drop in credits.
ID: 2909 · Report as offensive     Reply Quote
LP

Send message
Joined: 1 Sep 04
Posts: 39
Credit: 54,460
RAC: 0
Message 2910 - Posted: 30 Sep 2004, 20:11:09 UTC

I haven't seen a drop either, and I've been here since day 1.
ID: 2910 · Report as offensive     Reply Quote
rsbriggs

Send message
Joined: 1 Sep 04
Posts: 14
Credit: 168,153
RAC: 0
Message 2911 - Posted: 30 Sep 2004, 20:17:49 UTC
Last modified: 30 Sep 2004, 20:18:13 UTC

I was amongst the first 1000 to sign up and as far as I can tell haven't lost any points recently.

Who did lose points ?
ID: 2911 · Report as offensive     Reply Quote
Profile Yeti
Volunteer moderator
Avatar

Send message
Joined: 2 Sep 04
Posts: 453
Credit: 193,369,412
RAC: 10,065
Message 2914 - Posted: 30 Sep 2004, 20:28:15 UTC

I lost some points, but this could be related to me reducing the resource share of LHC with the factor 10



Supporting BOINC, because it is really a great concept !
ID: 2914 · Report as offensive     Reply Quote
Previous · 1 · 2 · 3 · Next

Message boards : Number crunching : Decision on points accumulated up to now


©2024 CERN