PDA

View Full Version : They have a team problem



rsbriggs
08-15-2003, 07:20 PM
12 Free-DC 8 4 184 759 4,391 85,548
13 Team Victo 3 1 8 54 3,973 83,945
...
49 Free-DC 1 1 0 0 0 750
50 Team Ninja 7 ...


When I joined team Free-DC, they put me in a completely separate team with the same name (# 49), instead of # 12 ...... I tried switching, from Free-DC to Free-DC to see if that would move me to the other....

Darkness Productions
08-15-2003, 10:05 PM
It seems to have worked... as you show up quite nicely right here: http://stats.dp.cx/ecc2/cgi-bin/render-users.cgi?TeamID=17


By the way, thanks for the extra DPs! :D

rsbriggs
08-15-2003, 11:43 PM
THere are more where those came from :D

magnav0x
08-16-2003, 04:54 PM
I still haven't figured out how to set up my account with ECC2. I went to http://ecc2.student.utwente.nl/ecc2/personalpage.php and saw no place to create an account. I went for "forgot username/password" and put my e-mail address in and submited the form. It e-mailed me a password (which didn't work). I got a P4 crunching this right now, but I don't know long I have to wait or how exactly I need to set everything up.

Beyond
08-16-2003, 06:02 PM
Originally posted by magnav0x
I still haven't figured out how to set up my account with ECC2. I went to http://ecc2.student.utwente.nl/ecc2/personalpage.php and saw no place to create an account. I went for "forgot username/password" and put my e-mail address in and submited the form. It e-mailed me a password (which didn't work). I got a P4 crunching this right now, but I don't know long I have to wait or how exactly I need to set everything up.

you have to submit some "units" then it will auto-email you with the address you setup the client to use asking if you wish to join in the stats :rolleyes: after the next stats run. Once you recieve and reply to this email you "should" be able to change your account information.

Hopefully I too will show up in the team's stats soon.........:cool:

magnav0x
08-16-2003, 06:11 PM
Well things is...I submitted work LONG ago and I never got an e-mail. I e-mailed them over and over about it and they would never get back to me so I just deleted the ecc2 software and never looked back. Seems I'll be in the same boat yet again :(

Beyond
08-16-2003, 06:19 PM
Originally posted by magnav0x
Well things is...I submitted work LONG ago and I never got an e-mail. I e-mailed them over and over about it and they would never get back to me so I just deleted the ecc2 software and never looked back. Seems I'll be in the same boat yet again :(

Yeah, that was a problem for many people way back when....I submitted some units Friday evening and recieved the email this morning. So I hope to see some stats after tonights stats run.

magnav0x
08-16-2003, 06:24 PM
I submitted a couple more wu's. If I get an e-mail after the stats run then I'll help out.

Beyond
08-16-2003, 07:02 PM
:thumbs:

rus6
08-18-2003, 04:19 AM
The more the merrier :D we could do with some more crunchers on ecc2

ECL
08-18-2003, 11:58 AM
Does anybody know where the magic number of 20.5 million DPs originated? Everybody seems to think the project is certain to wrap up within a few weeks, but I haven't been able to find any evidence supporting the 20.5 mill number.

I have found this (http://penguin.servehttp.com/feline/stats1.html), which paints an interesting picture of where we are on the probability curve.

Beyond
08-18-2003, 05:58 PM
I "believe" the 20.5 million mark was an estimate. It was bantered about that the project had a 50/50 chance of finding the 2 matching points within the first 20.5M points. How they arrived at that number is beyond me. :scratch:


Of course this brings uup another question, is it worth the risk of building a large cache for "emergencies" at this point or should one just return them asap?

rus6
08-18-2003, 06:14 PM
Originally posted by Beyond
[ Of course this brings uup another question, is it worth the risk of building a large cache for "emergencies" at this point or should one just return them asap? [/B]

Surely you should return them ASAP as you could be sitting on the magic DP :D

Beyond
08-18-2003, 06:28 PM
Originally posted by rus6
Surely you should return them ASAP as you could be sitting on the magic DP :D

Then if I did have the magic dp and I returned it we would be stuck in 12th place :( would we not? :confused:
(not that I have a large cache or anything magically, dp or otherwise)

rus6
08-19-2003, 05:09 AM
True...

If you happened to have found the magic DP, we wouldn't get any higher then 12th...

But if you stock pile DP's and someone else finds the magic DP then aren't all the CPU cycles wasted :cry: and that must be a crime round here ;)

Darkness Productions
08-19-2003, 09:35 AM
Not only would we be stuck at 12th, but there'd be a chance for you to win a chunk of a pretty decent prize....

Might be worth it ;)

wirthi
08-19-2003, 11:09 AM
Beyond, you should upload ASAP.

Two people will "win" this project - those two who find a seperate DP.

So, all the nodes that were uploaded up to now where only uploaded once. Some time in the future, somebody will upload a DP that is similar to one of these (I don't know in which way). Both of them will be "winners".

So, if you upload your work now,

1) you could be the winner right now (if somebody else had already uploaded a similar DP).

2) you could be the winner in the future (when somebody uploads a DP similar to yours).

If you don't upload you just wait for somebody else to find a pair of DPs ...

alpha
08-19-2003, 04:06 PM
Originally posted by ECL
Does anybody know where the magic number of 20.5 million DPs originated? Everybody seems to think the project is certain to wrap up within a few weeks, but I haven't been able to find any evidence supporting the 20.5 mill number.

My understanding of this is that by the 20.5 millionth DP, the chances of two matching DPs having been found is 50%. The only reason why someone should think the project is near finishing is due to the fact that the probability of two DPs clashing is getting closer to 50%. The project is in fact closer to completion after the submission of each single DP.

There is no reason why the first two submitted DPs couldn't match, nor the 26241st and the 1928574th. The probability of two DPs clashing is proportional to the amount of DPs submitted. Once 20.5M DPs have been submitted, the probability of a new DP being submitted which clashes with a previously submitted DP is greater than half (1/2, 50%, etc).

All of the above is how I perceive the problem, I don't claim any of this to be fact :)

Oh, and as tempting as it is to contribute to this project again (new optimized core, new-found Free-DC enthusiasm), I am well and truly finished for good now. Jay Berg couldn't be a worse project manager/admin. I know, I've said I was finished with this project before and still returned - but this time its for good. Have fun kids, kick some anus :smoking:

Beyond
08-19-2003, 04:20 PM
Originally posted by wirthi
Beyond, you should upload ASAP.

Two people will "win" this project - those two who find a seperate DP.

So, all the nodes that were uploaded up to now where only uploaded once. Some time in the future, somebody will upload a DP that is similar to one of these (I don't know in which way). Both of them will be "winners".

So, if you upload your work now,

1) you could be the winner right now (if somebody else had already uploaded a similar DP).

2) you could be the winner in the future (when somebody uploads a DP similar to yours).

If you don't upload you just wait for somebody else to find a pair of DPs ...

Well I hate to disappoint anyone, but the 500+ I dumped earlier was the cache (I did say I didnot have a large cache :) ), and I am uploading as I go, now.

Though this project is not a high prority for me, I do not mind answering the call for help for a short period of time. As stated in a different post/thread I am a little burnt out on OGR and looking to do some more of the lesser reconginzed projects, so one may see me showing up in the strangest of places. :D

wirthi
08-19-2003, 04:45 PM
Originally posted by alpha
There is no reason why the first two submitted DPs couldn't match, nor the 26241st and the 1928574th. The probability of two DPs clashing is proportional to the amount of DPs submitted. Once 20.5M DPs have been submitted, the probability of a new DP being submitted which clashes with a previously submitted DP is greater than half (1/2, 50%, etc).
I guess you mean the right thing, but what you wrote here doesn't seem right to me.

If you have a chance of 50% that the next sent-in DP (for example #25.000.001) is the winner, that chance would be rather high. In fact, it would mean that the project would be solved within - say - 10 DPs ...

IMHO the 20.5M mean the following: when having done 0 DPs up to now, you would have to do another 20.5M DPs to have a 50/50 chance of finding a pair. Or expressed in another way: If you take 20.5M random DPs, the chance that two of them fit together is 50%. Of course, as alpha already said, the project could have been solved with the first to DPs that were sent in, but that chance is somewhere near 0.000000001%.

The probability that the next sent-in DP is a fitting one is constantly rising with every sent-in DP. It is still a very low number, though.

alpha
08-20-2003, 02:09 AM
Originally posted by alpha
Once 20.5M DPs have been submitted, the probability of a new DP being submitted which clashes with a previously submitted DP is greater than half (1/2, 50%, etc).

Indeed wirthi. I withdraw the above sentence because it is total nonsense :)

That should have been:

Once 20.5M DPs have been submitted, the probability of a new DP being submitted which clashes with a previously submitted DP is greater than 50%.

rus6
08-20-2003, 05:09 AM
Originally posted by alpha
Oh, and as tempting as it is to contribute to this project again (new optimized core, new-found Free-DC enthusiasm), I am well and truly finished for good now. Jay Berg couldn't be a worse project manager/admin. I know, I've said I was finished with this project before and still returned - but this time its for good. Have fun kids, kick some anus :smoking: [/B]

Tiss is a shame, I saw some of your "discussions" with Jay on the forums...

alpha
08-23-2003, 04:14 AM
This guy is unbelievable.


It may very well please non-technical people who haven't studied the actual technical issues and refuse to listen to disk drive engineers and power management engineers. But, I frankly don't really care for a simple reason. You don't pay my salary and my house mortgage is more important to support than issues of dubious value!

This was in response to a discussion about frequent disk writes by the official ECC2 wrapper.

I can't understand how he gets away with such an attitude. Perhaps if I cared more for the purpose of the project I would disregard the above and continue contributing, but I know that my CPU cycles are deserved elsewhere.


We do appreciate suggestions for client changes and the track record shows that every team member has implemented many such changes. But this issue is dead! If you don't want SPD writes, my suggestion is to use SETI or some other project. If you don't want to burn up huge amounts of power, turn off your computer.

It's like he couldn't care less. His attitude is "Here is how the client works, if you don't like it go somewhere else or write your own software because we aren't going to change it."

I'm refraining from using expletives here to express my anger and disbelief :D :bang: