| Author |
Message |
|---|
|
|
|
I have had a prime in the queue for about 2 months now. I found it right around the SG challenge. Are we behind in processing those or did one (mine) get skipped?
Thanks,
Jim |
|
|
|
|
|
Same here. For more than a month I have SG prime count of 1, but none on my prime list. What gives?
____________
  |
|
|
|
|
|
Same here. I have found 3 primes at SG.
____________
|
|
|
|
|
|
I have two in the queue - guess they'll be attended to some time....!
____________
 |
|
|
|
|
|
See also here:
http://www.primegrid.com/forum_thread.php?id=2553
SGS Primes are appearently held back to be reported..
But no idea for how long ????
|
|
|
|
|
See also here:
http://www.primegrid.com/forum_thread.php?id=2553
SGS Primes are appearently held back to be reported..
But no idea for how long ????
Seems like I saw another thread (can't find it) that they are still working on a whole batch of pps' from prpnet as well pg that had to run the check before they could get back on the sgs or maybe I dreamed that.
____________
@AggieThePew
 |
|
|
|
|
|
Can someone explain, (or point me in the direction of a thread that explains) the process of 'waiting for database space', with regard to reporting/confirming backlogged primes. Its not that I'm impatient (well, perhaps a little) - its more that I would like to understand a bit better what goes on behind the scenes.......
____________
|
|
|
|
|
Can someone explain, (or point me in the direction of a thread that explains) the process of 'waiting for database space', with regard to reporting/confirming backlogged primes. Its not that I'm impatient (well, perhaps a little) - its more that I would like to understand a bit better what goes on behind the scenes.......
Same for me...
I did not find any thread that explains the reporting policy for our found primes within primegrid.com..
I think that all the 'behind the scenes' is managed from the Admin's
(and I also think this is quite good so)
But maybe one time the Admin's will open a new Thread :
"What we working on right now...."
"Behind the scene..what we do with your work done.."
Or something similar ?? |
|
|
John Honorary cruncher
Send message
Joined: 21 Feb 06
Posts: 2875
ID: 2449
Credit: 2,681,934
RAC: 0
                 
|
Can someone explain, (or point me in the direction of a thread that explains) the process of 'waiting for database space', with regard to reporting/confirming backlogged primes.
Although I thought this was discussed before, I'm not able to find the thread. :( Therefore, let's see if I can wrap up all the pieces in this post.
Last year we had two validation issues with SGS primes. The first one resulted in a backlog of over 200 primes (all of which have been released and submitted). The second one resulted in a backlog of about 75 (of which only about 25 have been released and submitted). Both validation issues were resolved last year.
Now the main issue. The PPS (LLR) Prime Search primes are at the bottom edge of the Top 5000 Primes Database...however, they are hanging on by a thread. Last year, we ran into the same situation and were able to "micro-manage" the submissions to make sure all primes made the list...if only for a day.
However, after the second validation issue in December, we have not been able to recover. SGS primes are now released on a "Top 5000" space available basis only. Previously, we could release a handful each day. Now, we are able to release only about a handful a week. With the current trend, it doesn't look like we'll be able to catch up until PPS (LLR) reaches n=666666 (currently at n=612K).
After all PPS primes are submitted, if there is still room between the 5000th prime and the lowest outstanding PPS task, then SGS primes are submitted until that buffer is gone.
SGS primes enter around 2700th position right now. Therefore, all the backlogged primes are completely safe from falling off.
This is certainly NOT an ideal solution but one that we have chosen to keep PPS viable and to allow more users to experience a Top 5000 prime. Once PPS (LLR) reaches n=666666, then all SGS (LLR) backlogged primes will be released. We are hoping to reach that goal within the next few months.
____________
|
|
|
|
|
|
John, thanks for the response.
This is the bit that has me stumped:
SGS primes are now released on a "Top 5000" space available basis only.
If, in general, SGS primes fall in the <3000th position, why is there a need to 'wait for space'. Doesn't everything that is 'numerically' less than a given SGS prime just shuffle along by one - resulting in the 5000th entry being bumped off the list? I understand that the Caldwell database only has a set number of slots for SGS primes, and that these slots will be occupied by the largest SGS primes, but don't SGS primes that do not fall into this (top n ) category still occupy a slot on the basis that they are big enough to gain a position <3000? I must be missing something here!!!!!!
____________
|
|
|
JohnHonorary cruncher
Send message
Joined: 21 Feb 06
Posts: 2875
ID: 2449
Credit: 2,681,934
RAC: 0
|
If, in general, SGS primes fall in the <3000th position, why is there a need to 'wait for space'. Doesn't everything that is 'numerically' less than a given SGS prime just shuffle along by one - resulting in the 5000th entry being bumped off the list?
Yes, this is exactly what happens. But it's not the SGS primes that are being bumped off...it's the PPS primes. If we push the 5000th prime to above the PPS threshold, then none of the PPS primes will make the Top 5000 primes database.
The current method allows for all primes to make the Top 5000 list, with the SGS primes being delayed until there's space.
____________
|
|
|
|
|
|
Understand now, thanks for the this.
____________
|
|
|
mfbabb2Volunteer tester
Send message
Joined: 10 Oct 08
Posts: 510
ID: 30360
Credit: 20,308,107
RAC: 21,442
                   
|
If, in general, SGS primes fall in the <3000th position, why is there a need to 'wait for space'. Doesn't everything that is 'numerically' less than a given SGS prime just shuffle along by one - resulting in the 5000th entry being bumped off the list?
Yes, this is exactly what happens. But it's not the SGS primes that are being bumped off...it's the PPS primes. If we push the 5000th prime to above the PPS threshold, then none of the PPS primes will make the Top 5000 primes database.
The current method allows for all primes to make the Top 5000 list, with the SGS primes being delayed until there's space.
I realize that it would take a lot of extra work, but if the goal is to keep a (somewhat) balance of types, then why not just remove the smallest SGS prime from the Top 5000 for every new SGS added. That would leave the PPS ones intact (for now). If that does not seem fair, then I would say that the PPS probably do not belong in the Top 5000 ... (let the word flinging begin ...)
____________
Murphy (AtP)
 |
|
|
|
|
|
You can't just delete primes from a "largest known" list. ;) |
|
|
|
|
|
I'm not suggesting things change, and John did point out that this wasn't an ideal situation ;-)
Presumably, PrimeGrid wants to find all Proths where n<666666, and it may be difficult to get people to wrap up the hunt if they can't get even a 'transient' listing in the top 5000. I can understand the logic; after all, PrimeGrid doesn't define the structure of the Caldwell listings. Perhaps PrimeGrid has become a victim of its own success.
You can't just delete primes from a "largest known" list. ;)
I don't see why not. If you define the purpose of the list as containing the 'x largest known primes', and a new prime is uncovered and is added to the list by virtue of its magnitude, then by definition the entry at position x is no longer fulfilling the defining criteria of the list and should be removed.
____________
|
|
|
|
|
|
Any chance the Top 5000 could be expanded to the Top 10000?
With the amount of primes being discovered, this seems to be more appropriate.
____________
 |
|
|
mackerelVolunteer tester
Send message
Joined: 2 Oct 08
Posts: 2639
ID: 29980
Credit: 568,393,769
RAC: 6,021
                            
|
|
The top5000 prime list is just that, the top 5000 of all known primes. So you can't just delete primes in the top5000. They do keep lists of top x of specific types even below position 5000, but SGS or PPS isn't particularly interesting in that respect.
Also note the list is external to primegrid. |
|
|
|
|
|
John,
Great explanation and I personally think a great plan.
Jim |
|
|
JohnHonorary cruncher
Send message
Joined: 21 Feb 06
Posts: 2875
ID: 2449
Credit: 2,681,934
RAC: 0
|
They do keep lists of top x of specific types even below position 5000, but SGS or PPS isn't particularly interesting in that respect.
They do keep the Top 20 Sophie Germain Primes...of which we have yet to find one (only "normal" single primes). Additionally, they have the Top 20 Proth Primes...of which PG's Proth Prime Search is much too low to make the list. However, PPS does have the possibility to find a rare Top 20 Prime Fermat Divisor.
Check out Top Twenty's Complete Index for more top 20 lists.
____________
|
|
|
 rroonnaallddVolunteer developer
Volunteer tester
Send message
Joined: 3 Jul 09
Posts: 1213
ID: 42893
Credit: 34,634,263
RAC: 0
             
|
|
12829251133725*2^666666-1 is prime.
____________
Best wishes. Knowledge is power. by jjwhalen
 |
|
|
|
|
|
Had my first one verified yesterday...!
13613082681735*2^666666-1 is Prime!
|
|
|
|
|
|
I continue (patiently) to wait.
____________
Thanks,
Jim
 |
|
|
|
|
[quote]They do keep lists of top x of specific types even below position 5000
PPS does have the possibility to find a rare Top 20 Prime Fermat Divisor.
I'm #10 on that list :) |
|
|
|
|
|
Just noticed that my profile is showing my first ever prime number under SGS, must have gone through either yesterday or the day before. Nothing showing on my primes page yet, but that is to be expected.
Is there still a backlog of SG primes to be released?
I just am dying to know if i was the primary finder or just the double checker!
Peter
____________
My Top 5000 Primes:
110059!+1 is prime! (FPS)
14493618614235*2^666666-1 is prime! (SGS) |
|
|
|
|
Just noticed that my profile is showing my first ever prime number under SGS, must have gone through either yesterday or the day before. Nothing showing on my primes page yet, but that is to be expected.
Is there still a backlog of SG primes to be released?
I just am dying to know if i was the primary finder or just the double checker!
Peter
Congrats on getting a hit and hopefully you are the initial finder :)
They are currently on a holding pattern with the prime validations due to a server upgrade (I think) and Rytis being on vacation.
Rick
Edit: Might have miss-spoke on the holding pattern. That might just be the submission to the top 5000. Anyway here's a link one of the posts about primes and I was thinking there was another one. http://www.primegrid.com/forum_thread.php?id=1035
____________
@AggieThePew
|
|
|
|
|
|
I see that the automatic prime submission is up and running again.
Is there a large backlog of primes to be manually submitted or are SGS primes being held off until PPS reaches n=666666?
Hammeh
____________
My Top 5000 Primes:
110059!+1 is prime! (FPS)
14493618614235*2^666666-1 is prime! (SGS) |
|
|
|
|
I see that the automatic prime submission is up and running again.
Is there a large backlog of primes to be manually submitted or are SGS primes being held off until PPS reaches n=666666?
Hammeh
I was just saying, on one of the "number crunching" threads that I'm
also waiting, for 13 PPS reports and 2 SGS reports. I don't appear to
be anywhere near the top among people running LLR PPS, and Scott
Brown, in particular, has also been running hard on n=666666; and
must have a whole bunch more than that; as well as who-knows-who
else that responded to John's request that 666K be finished before the
April Challenge.
So I'm thinking that holding off SGS isn't the problem, so much as a whole
bunch more of PPS's. -Bruce* |
|
|
|
|
|
well, I just got 2 hits for an SG in less than a 24 hour period. Not sure if I'm the finder or the dc but that's the first time I've ever had hits that close together.
Rick
____________
@AggieThePew
|
|
|
|
|
|
I'm still waiting on my one :(
____________
My Top 5000 Primes:
110059!+1 is prime! (FPS)
14493618614235*2^666666-1 is prime! (SGS) |
|
|
|
|
I'm still waiting on my one :(
Keep after it. It took me quite a long time to ever get my first and then a much longer time to ever get a 2nd.
Happy hunting.
____________
@AggieThePew
|
|
|
|
|
|
Incredibly lucky, I've been running just SGS for the last week, and I have had no hits (but that is to be expected).
____________
 |
|
|
|
|
|
Yep I will agree that it's very lucky... no question about that... LOL I might post to the team recruitment thread... or not :)
____________
@AggieThePew
|
|
|
|
|
|
Got up and in this morning and see that I have another hit on an SG. Still don't know if any of the 3 are initial finder hits or double checker hits.
Go Aggie The Pew
____________
@AggieThePew
|
|
|
mfbabb2Volunteer tester
Send message
Joined: 10 Oct 08
Posts: 510
ID: 30360
Credit: 20,308,107
RAC: 21,442
|
Got up and in this morning and see that I have another hit on an SG. Still don't know if any of the 3 are initial finder hits or double checker hits.
Go Aggie The Pew
I would be happy with either -- my current score is "0" (out of 2960 WUs).
____________
Murphy (AtP)
|
|
|
|
|
Got up and in this morning and see that I have another hit on an SG. Still don't know if any of the 3 are initial finder hits or double checker hits.
Go Aggie The Pew
I would be happy with either -- my current score is "0" (out of 2960 WUs).
I'm willing to bet you and lots of us get hits this weekend. Although, I never got a single hit during the pps challenge.
Good luck man!
____________
@AggieThePew
|
|
|
mfbabb2Volunteer tester
Send message
Joined: 10 Oct 08
Posts: 510
ID: 30360
Credit: 20,308,107
RAC: 21,442
|
|
Thanks. You, too.
____________
Murphy (AtP)
|
|
|
|
|
Got up and in this morning and see that I have another hit on an SG. Still don't know if any of the 3 are initial finder hits or double checker hits.
Go Aggie The Pew
Wow you and Sophie seem to be best friends! Congratulations.
My one SG hit came just a few hours after the start of this year's Tour de Primes, and was followed by a few PPS hits (including 2 in one day) during the Feb challenge. So I had good luck at the start, but the rest of the Tour was barren for me. My two cores are churning away looking for another SG during this challenge, then will return to sieving for a while.
--Gary |
|
|
|
|
Got up and in this morning and see that I have another hit on an SG. Still don't know if any of the 3 are initial finder hits or double checker hits.
Go Aggie The Pew
Wow you and Sophie seem to be best friends! Congratulations.
My one SG hit came just a few hours after the start of this year's Tour de Primes, and was followed by a few PPS hits (including 2 in one day) during the Feb challenge. So I had good luck at the start, but the rest of the Tour was barren for me. My two cores are churning away looking for another SG during this challenge, then will return to sieving for a while.
--Gary
Thanks - however as an update, so far I've had NO hits during the challenge. Hopefully that will change :)
Good hunting!
____________
@AggieThePew
|
|
|
|
|
|
It was a great challenge :) Even got another hit... so now I have to wait to see if my new hits are primes or double checkers.
____________
@AggieThePew
|
|
|
|
|
I see that the automatic prime submission is up and running again.
Is there a large backlog of primes to be manually submitted or are SGS primes being held off until PPS reaches n=666666?
Hammeh
I was just saying, on one of the "number crunching" threads that I'm
also waiting, for 13 PPS reports and 2 SGS reports. I don't appear to
be anywhere near the top among people running LLR PPS, ... -Bruce*
By way of an update, I'm currently waiting for 24 PPS reports and
3 SGPS reports. None during the Challenge, and I was the 7th
individual. Not a complete loss of the weekend, as I got my Gold
badge for Woodall, with the gold for Cullen on the near horizon ...
(uhm, that's c. six more needed, with 13 "in progress", 20 pending;
I'm not too good on patiently waiting, it seems).
Hope those of you with SG reports during the Challege get
credited with the primes; well deserved. -Bruce*
-Bruce* |
|
|
|
|
I was just saying, on one of the "number crunching" threads that I'm
also waiting, for 13 PPS reports and 2 SGS reports. I don't appear to
be anywhere near the top among people running LLR PPS, ... -Bruce*
By way of an update, I'm currently waiting for 24 PPS reports and
3 SGPS reports. None during the Challenge, ... with the gold for Cullen on the near horizon ...
Hope those of you with SG reports during the Challege get
credited with the primes; well deserved. -Bruce*
Once more, if you don't mind too much. Cullen gold came in
quicker than expected (I didn't need the six new tasks, but that's OK).
But on the topic of the thread, I may have spoke too soon
about the SGPS reports from the Challenge; I have a +1 report
this morning. I switched back to PPS LLR shortly after the Challenge,
like 30 minutes after, so either this is a late report from during
the Challenge, or a task most likely sent during the Challenge
but not returned until just after --- in progress and pending had
already cleared last night. -Bruce* |
|
|
|
|
|
My prime report went through yesterday:
14493618614235*2^666666-1 is prime
And i was the inital finder =) so happy!!!
Peter
____________
My Top 5000 Primes:
110059!+1 is prime! (FPS)
14493618614235*2^666666-1 is prime! (SGS) |
|
|
|
|
My prime report went through yesterday:
14493618614235*2^666666-1 is prime
And i was the inital finder =) so happy!!!
Peter
Congratulations. I now have two new Proth's and one new
doublecheck each on Proth and SGPS. Not entirely sure that's
forward progress; I now have 91 reports (primes+doubles)
and far fewer than that credited; like 54, I think. The backlog's
supposed to clear within the week, and I'm hoping to see a
prime above the 200K-digit range from n=666666. Still, another
prime from the Sophie/Twin search would be good too.
-Bruce* |
|
|
|
|
|
Started one of my pc's back on sg last night.. got another hit so now I'm waiting on 5 results to see if the primes are mine or I just checked them.
____________
@AggieThePew
|
|
|
|
|
My prime report went through yesterday:
14493618614235*2^666666-1 is prime
And i was the inital finder =) so happy!!!
Peter
Congratulations. ... I'm hoping to see a
prime above the 200K-digit range from n=666666. Still, another
prime from the Sophie/Twin search ... -Bruce*
Bingo! Initial finder on seven Proths above 200.8K (with the
SG/Twin range in 200.7K). -Bruce* |
|
|
|
|
|
Hello, folks.
I don't understand what's happening with my 1st SGS and 5th PPS hits:
I'd noticed during the Challenge (2nd April) that I received a hit in SGS LLR
After the Challenge (4th April) I'd also noticed that a hit in PPS LLR appeared.
Today is 12 April, but I don't see any changes in my primes list:
http://www.primegrid.com/primes/?section=primelist&userid=64131
I've even checked all pending WU's.
All of them are transffered after the Challenge finish.
Tell me what's happening?
Is there backlog till now? |
|
|
JohnHonorary cruncher
Send message
Joined: 21 Feb 06
Posts: 2875
ID: 2449
Credit: 2,681,934
RAC: 0
|
Is there backlog till now?
Yes, should be cleared within the week.
EDIT: see this post.
____________
|
|
|
|
|
|
I also see that Alexus78, the champion of the last Challenge, has 10 hits in SGS LLR and most of them are received during the Challenge:
You remember he got 800 k cobblestones during the Challenge.
But he has only 2 published primes till now:
http://www.primegrid.com/primes/?section=primelist&userid=77482
Sadly. |
|
|
|
|
|
John, what about a period of storage of the WU's?
Whether we can check the rule of initial finder/doublechecker definition after the 2 weeks for submission? |
|
|
JohnHonorary cruncher
Send message
Joined: 21 Feb 06
Posts: 2875
ID: 2449
Credit: 2,681,934
RAC: 0
|
John, what about a period of storage of the WU's?
Whether we can check the rule of initial finder/doublechecker definition after the 2 weeks for submission?
Tasks remain until purged. Rytis or Lennart can identify the "normal" purge time for each project.
____________
|
|
|
|
|
I also see that Alexus78, the champion of the last Challenge, has 10 hits in SGS LLR and most of them are received during the Challenge:
You remember he got 800 k cobblestones during the Challenge.
But he has only 2 published primes till now:
http://www.primegrid.com/primes/?section=primelist&userid=77482
Sadly.
I'm still waiting for two SGPS reports, one from before the Challenge.
I had 147.3K of those "cobblestones", well above your score, but I'm
not clear that even one of the SGPS reports was from that 147K (that
is, the 2nd one I'm waiting for may have been just after the Challenge;
with none at all from the 147K!). The timespan of 48_hours really isn't
long enough for reliable estimation of what ought to be reasonable, on
these "black swan" observations. I'm sure that your friend/team_mate
will get credit for those other 8 reports; as the claim that the backlog
would be settled within the week is looking a lot more likely, to me,
at least. -bd* |
|
|
|
|
|
Actually I don't exactly understand how it looks like.
The Challenge has finished. Over 120 primes are found, John wrote.
But we don't know till now, who exactly?
First, I didn't know that to identify the initial finder and to report to Top5000 db are the same process.
Second, John already wrote (27 March):
FYI...the validation problem that was occasionally reversing finder/DC'er has been resolved.
Thus I've decided that manual primes submission is not required already. |
|
|
|
|
Actually I don't exactly understand how it looks like.
The Challenge has finished. Over 120 primes are found, John wrote.
But we don't know till now, who exactly?
First, I didn't know that to identify the initial finder and to report to Top5000 db are the same process.
Second, John already wrote (27 March):
FYI...the validation problem that was occasionally reversing finder/DC'er has been resolved.
Thus I've decided that manual primes submission is not required already.
No, the "software issue has been resolved" seems to mean that for the
future; once the backlog has been manually cleared, subsequent primes
will be automatically reported correctly. John's comments were made over
on the "number crunching" subforum, in the sub-subforum
PPS(LLR) n = 666666 Countdown!
What he says was
Prime submission to the Top5000 primes database has been resumed. It will take about a week to clear the backlog. NOTE: All backlogged primes (PPS & SGS) are still safe from falling out of the Top 5000.
If you are showing a prime on your account page, expect it to be submitted within the week.
You were right at the start; we won't know who got which of the
120 primes until all outstanding reports have been (manually) cleared.
That's supposed to happen soon. My backlog of reports that I didn't
know whether I was the 'finder or the doublechecker was as high as
35, but in the past day-or-two it is down to 15, so we are seeing rapid
progress.
Hope this helps clarify matters. The part about having the backlog
cleared within a week is still with Prime Submission being done manually,
if I understood correctly. -Bruce* |
|
|
|
|
... My backlog of reports that I didn't
know whether I was the 'finder or the doublechecker was as high as
35, but in the past day-or-two it is down to 15, so we are seeing rapid
progress. ... -Bruce*
My count of SGP's is now current, including a new report from
the past 24 hours. A bunch of Proths have also been settled,
just waiting for the doublechecks to post (the emails arrive before
the 'finder and doubles show up on the account page listing).
I was the finder of two SGP's and another seven Proths; 9 in
one day's emails! Several of the new Proths are very tasty,
248K-digits.
I'm also within sight of a milestone, at 98 prime reports
(finder + doubles, Proths and SGPs), two more to 100.
-Bruce* |
|
|
|
|
|
Got 4 of 5 primes on the ones I had waiting. Dc on the other. |
|
|
|
|
... My backlog of reports that I didn't
know whether I was the 'finder or the doublechecker was as high as
35, but in the past day-or-two it is down to 15, so we are seeing rapid
progress. ... -Bruce*
My count of SGP's is now current, including a new report from
the past 24 hours. A bunch of Proths have also been settled,
just waiting for the doublechecks to post (the emails arrive before
the 'finder and doubles show up on the account page listing).
I was the finder of two SGP's and another seven Proths; 9 in
one day's emails! Several of the new Proths are very tasty,
248K-digits.
I'm also within sight of a milestone, at 98 prime reports
(finder + doubles, Proths and SGPs), two more to 100.
-Bruce*
Nice one Bruce!
I've still smiling with my 1 SGS prime. How many cores do you currently have running?
Peter
____________
My Top 5000 Primes:
110059!+1 is prime! (FPS)
14493618614235*2^666666-1 is prime! (SGS) |
|
|
|
|
|
So, I'am DC of 7946399780865*2^666667-1:
http://www.primegrid.com/primes/?section=primelist&userid=64131
Today I have no ability to recheck is this a true.
The WU is already purged out.
That I've told about.
I'm not a paranoid, don't think so.
Simply, one occasion in the past leads to a doubt in the future.
EDIT: John, what do think about to add a link to WU for every prime in primelist section?
Second proposal - do not purge WU with prime found. |
|
|
|
|
My count of SGP's is now current, including a new report from
the past 24 hours. A bunch of Proths have also been settled,
just waiting for the doublechecks to post (the emails arrive before
the 'finder and doubles show up on the account page listing).
...
I'm also within sight of a milestone, at 98 prime reports
(finder + doubles, Proths and SGPs), two more to 100.
-Bruce*
Nice one Bruce!
I've still smiling with my 1 SGS prime. How many cores do you currently have running?
Peter
Thanks. The top prime finders list has 97-of-my-98 reports, I'd say
that the backlog's cleared.
Actually, the University is saying that they'd like to have their
blade server back. I had boinc managers open on 39-nodes,
each with 4 dual core xeons; the catch being that they're 32-bit
chips. I thought they'd be good for LLR's, which seems to be the
case. Wonder whether they'd settle for having me cut back to
25-nodes? -Bruce
(We also have a new machine like Alexus78 was using to win
the Challenge, 8-core AMD chips, but it's strictly for jobs run
through the PBS/torque scheduler.)
|
|
|
rroonnaallddVolunteer developer
Volunteer tester
Send message
Joined: 3 Jul 09
Posts: 1213
ID: 42893
Credit: 34,634,263
RAC: 0
|
|
x3mEn wrote: EDIT: John, what do think about to add a link to WU for every prime in primelist section?
Second proposal - do not purge WU with prime found.
The boinc-server-software deletes all units and purges the results from the database after a given time-frame. Each project has it's own preferences for this.
I believe, the compiling of such a dependency was never the plan from DA. One more time a coding like this has to be done by PG itself. But this is no bad decision...
____________
Best wishes. Knowledge is power. by jjwhalen
|
|
|
|
|
x3mEn wrote: EDIT: John, what do think about to add a link to WU for every prime in primelist section?
Second proposal - do not purge WU with prime found.
The boinc-server-software deletes all units and purges the results from the database after a given time-frame. Each project has it's own preferences for this.
I believe, the compiling of such a dependency was never the plan from DA. One more time a coding like this has to be done by PG itself. But this is no bad decision...
Ok, selective deleting is difficult thing.
What about a link to WU for a prime?
While WU exists this link will save a lot of search time.
When WU will be deleted the link might be deleted too or referenced to absent WU.
Now I can type something like http://www.primegrid.com/workunit.php?wuid=167138728 and will receive "the task not found".
Thus I see nothing bad if the link referenced to absent WU.
____________
|
|
|
|
|
|
Got another hit when I checked this morning and I am just working toward my gold badge.. sg is being good to me right now.
____________
@AggieThePew
|
|
|
|
|
|
The hit from yesterday was a prime and I checked this morning and had another hit... guess the challenge cleared out a lot of composites. |
|
|
mfbabb2Volunteer tester
Send message
Joined: 10 Oct 08
Posts: 510
ID: 30360
Credit: 20,308,107
RAC: 21,442
|
The hit from yesterday was a prime and I checked this morning and had another hit... guess the challenge cleared out a lot of composites.
I'm glad at least one of us is getting hits -- I am still 0 for 4798.
____________
Murphy (AtP)
|
|
|
|
|
The hit from yesterday was a prime and I checked this morning and had another hit... guess the challenge cleared out a lot of composites.
I'm glad at least one of us is getting hits -- I am still 0 for 4798.
I'm 13 for 114786. Say I got lucky (I _did_ get lucky, let's be clear), and
I might have only found 11. That would be 1 prime report per 10435 tasks.
Guess that's the wrong way to round (but the division was easy). Actual
number seems to be 1 prime per 8829. There's a bias here, nearly all of my
tasks are since Feb 1, 2011. Ah, but that's OK; we're looking at hits in
April, which is even more recent (the ones done earlier don't count; they've
already missed?).
It's still my impression that I finished in the top10 of the Challenge without
any hit during that 48 hours (there was one that I think was shortly after,
not sure about the date). I also have friends (and a co-author, even) that
mutter about Poisson distribution, and clustering. If I understand correctly,
these independent events tend not to occur uniformly; but rather with
larger empty gaps, mixed with clustering. Someone here undoubtly
understands this better than I do; but a uniform distribution would
suggest that your prime hit isn't later yet, while the non-uniformity says
that we expect large variance from what uniform would suggest.
Again; I'd be overjoyed if someone has a more professional account.
Best Regards, Bruce (... not quite a bus-man's holiday, I'm forbidden (locally)
from offering opinions on statistics.) |
|
|
|
|
The hit from yesterday was a prime and I checked this morning and had another hit... guess the challenge cleared out a lot of composites.
I'm glad at least one of us is getting hits -- I am still 0 for 4798.
I'm 13 for 114786. Say I got lucky (I _did_ get lucky, let's be clear), and
I might have only found 11. That would be 1 prime report per 10435 tasks.
Guess that's the wrong way to round (but the division was easy). Actual
number seems to be 1 prime per 8829. There's a bias here, nearly all of my
tasks are since Feb 1, 2011. Ah, but that's OK; we're looking at hits in
April, which is even more recent (the ones done earlier don't count; they've
already missed?).
It's still my impression that I finished in the top10 of the Challenge without
any hit during that 48 hours (there was one that I think was shortly after,
not sure about the date). I also have friends (and a co-author, even) that
mutter about Poisson distribution, and clustering. If I understand correctly,
these independent events tend not to occur uniformly; but rather with
larger empty gaps, mixed with clustering. Someone here undoubtly
understands this better than I do; but a uniform distribution would
suggest that your prime hit isn't later yet, while the non-uniformity says
that we expect large variance from what uniform would suggest.
Again; I'd be overjoyed if someone has a more professional account.
Best Regards, Bruce (... not quite a bus-man's holiday, I'm forbidden (locally)
from offering opinions on statistics.)
Personally, I think it's a bit of both luck and the cluster affect. Luck that you get the right wu(s) and get them turned in before a dc does and cluster because there does seem to be a bunched group between lots of composites. However, this is just my opinion :)
Oh and did I forget, karma also has to play a role with all the prime gods and their games.
Rick
____________
@AggieThePew
|
|
|
