I think it's to do with teams. I don't think it's a coincidence that the team recruiting slogan on the download site matches the team that KevinT is a member of. I guess (and it is only a guess) that at least one team has been promoting - heavily promoting - a client which is deprecated by the project developers. In which case, I don't have much sympathy: I think any complaints should be directed at the team promoters.

Here is a result of that is using 5.10.13 - official release of Berkeley -
still has validate errors -

Sorry to burst your bubble.

http://setiathome.berkeley.edu/result.php?resultid=648270957
http://setiathome.berkeley.edu/result.php?resultid=647043352

---

Dave,

It would be interesting if you were to look at your message log for this morning, and see if that particular upload/report pair respected your 0.1 day connect interval, or whether - like my Beta log earlier - a work fetch request meant that your result was actually reported much quicker than normal.

Here's from around that time, one of the validate errors in question was 11fe07ac.23804.13569.6.6.70

...
11/4/2007 5:23:54 AM|SETI@home|[file_xfer] Started upload of file 11fe07ac.23804.13569.6.6.70_0_0
11/4/2007 5:23:58 AM|SETI@home|[file_xfer] Finished upload of file 11fe07ac.23804.13569.6.6.70_0_0
11/4/2007 5:23:58 AM|SETI@home|[file_xfer] Throughput 26347 bytes/sec
11/4/2007 5:23:58 AM|SETI@home|Computation for task 11fe07ac.23924.91123.9.6.75_0 finished
11/4/2007 5:23:58 AM|SETI@home|Starting 11fe07ac.23839.92350.7.6.107_1
11/4/2007 5:23:58 AM|SETI@home|Starting task 11fe07ac.23839.92350.7.6.107_1 using setiathome_enhanced version 528
11/4/2007 5:24:00 AM|SETI@home|[file_xfer] Started upload of file 11fe07ac.23924.91123.9.6.75_0_0
11/4/2007 5:24:01 AM|SETI@home|Sending scheduler request: To fetch work
11/4/2007 5:24:01 AM|SETI@home|Requesting 2550 seconds of new work, and reporting 1 completed tasks
...

-Dave

Bingo! 3 seconds.

I think it's to do with teams. I don't think it's a coincidence that the team recruiting slogan on the download site matches the team that KevinT is a member of. I guess (and it is only a guess) that at least one team has been promoting - heavily promoting - a client which is deprecated by the project developers. In which case, I don't have much sympathy: I think any complaints should be directed at the team promoters.

Here is a result of that is using 5.10.13 - official release of Berkeley -
still has validate errors -

Sorry to burst your bubble.

http://setiathome.berkeley.edu/result.php?resultid=648270957
http://setiathome.berkeley.edu/result.php?resultid=647043352

Please check your logs, as Dave Nolan has done.

I didn't mean to imply that v5.10.13 was immune from errors - heck, I posted the log from one of my own earlier in the thread. But I do think that v5.9.0 almost guarantees them.

Wasn't BOINC v5.10.13 the same release that if you set your Connect To interval to 0 (zero) it was the same as RRI?

---

Wasn't BOINC v5.10.13 the same release that if you set your Connect To interval to 0 (zero) it was the same as RRI?

yes

---

"Living by the wisdom of computer science doesn't sound so bad after all. And unlike most advice, it's backed up by proofs." -- Algorithms to live by: The computer science of human decisions.

Wasn't BOINC v5.10.13 the same release that if you set your Connect To interval to 0 (zero) it was the same as RRI?

Yes, and for that reason people on this board soon started advising the use of a C.I. that was slightly above zero. The advantage of v5.10.13 is that you can do this sort of fine-tuning (mine is at 0.01) - v5.9.0 is uncontrollable.

I think it's to do with teams. I don't think it's a coincidence that the team recruiting slogan on the download site matches the team that KevinT is a member of. I guess (and it is only a guess) that at least one team has been promoting - heavily promoting - a client which is deprecated by the project developers. In which case, I don't have much sympathy: I think any complaints should be directed at the team promoters.

Here is a result of that is using 5.10.13 - official release of Berkeley -
still has validate errors -

Sorry to burst your bubble.

http://setiathome.berkeley.edu/result.php?resultid=648270957
http://setiathome.berkeley.edu/result.php?resultid=647043352

Please check your logs, as Dave Nolan has done.

I didn't mean to imply that v5.10.13 was immune from errors - heck, I posted the log from one of my own earlier in the thread. But I do think that v5.9.0 almost guarantees them.

So it seems, but this has not been true except in the last few days, I think.
I just posted all my validation errors in the Validate Errors II thread, and it was a more lengthy list than I had thought, no wonder my RAC is having troubles.
(There must be a bbcode thing I am spacing out on right now, I thought those would all be live links).

EDIT....
BTW, crunch3r is currently working on an updated client with a reporting time delay to try to get around the current problem.

---

"Freedom is just Chaos, with better lighting." Alan Dean Foster

(There must be a bbcode thing I am spacing out on right now, I thought those would all be live links).

I think you're missing the close ] at the end of your url= statements and
some kind of label for the link, ie.

Fixed

---

Yes, and for that reason people on this board soon started advising the use of a C.I. that was slightly above zero. The advantage of v5.10.13 is that you can do this sort of fine-tuning (mine is at 0.01) - v5.9.0 is uncontrollable.

It possibly is a good place to add that newer clients have this setting too (e.g. 5.10.20 that i use) however will save up reports up until your next work fetch, up to 24 hours. I gather that might effectively halve server connections...

That is possibly why I've had good fortune this round ... they never seem to wait a full day, but an hour or two is common, and uploads coinciding with reports seems to not be happening to me.

Jason

[ CI = 0.2, 3 day cache ]

---

"Living by the wisdom of computer science doesn't sound so bad after all. And unlike most advice, it's backed up by proofs." -- Algorithms to live by: The computer science of human decisions.

So it seems, but this has not been true except in the last few days, I think.
....

I agree, just as I agree with your earlier comment that the SETI/BOINC servers should be robust enough to accept immediate reporting.

But the fact is that they aren't - not all the time, anyway. And nothing short of a hefty donation of new hardware and liveware (staff) is going to make them that way.

Think of it like your (always entertaining!) experiments with overclocking. We, the crunchers, are doing our best to 'overclock' the servers. Sometimes we stay within the range of stability, and get away with it. But sometimes we push the 'overclock' a bit too far, and the whole thing blows up in our face. Under those circumstances, it's better and more productive to back off a few cycles, rather than keep pushing to the limit and end up having to wait all next week for Fed-Ex to deliver the credits.

(There must be a bbcode thing I am spacing out on right now, I thought those would all be live links).

I think you're missing the close ] at the end of your url= statements and
some kind of label for the link, ie.

Fixed

I'll fix it tonight, it took me half an hour to enter it all, and I've gotta run right now.

---

"Freedom is just Chaos, with better lighting." Alan Dean Foster

So it seems, but this has not been true except in the last few days, I think.
....

I agree, just as I agree with your earlier comment that the SETI/BOINC servers should be robust enough to accept immediate reporting.

But the fact is that they aren't - not all the time, anyway. And nothing short of a hefty donation of new hardware and liveware (staff) is going to make them that way.

Think of it like your (always entertaining!) experiments with overclocking. We, the crunchers, are doing our best to 'overclock' the servers. Sometimes we stay within the range of stability, and get away with it. But sometimes we push the 'overclock' a bit too far, and the whole thing blows up in our face. Under those circumstances, it's better and more productive to back off a few cycles, rather than keep pushing to the limit and end up having to wait all next week for Fed-Ex to deliver the credits.

Totally agree Richard. This is a fair analysis of the situation.

Fixed Boinc to 5.10.28(Benchmarks better actually), Flu not so bad today, So far. :D

Later.

---

The T1 Trust, PRR T1 Class 4-4-4-4 #5550, 1 of America's First HST's

(There must be a bbcode thing I am spacing out on right now, I thought those would all be live links).

I think you're missing the close ] at the end of your url= statements and
some kind of label for the link, ie.

Fixed

Fixed my Validation Errors II thread posting....it's quite a list...many credits have been blown away by the invalidator ray.

---

"Freedom is just Chaos, with better lighting." Alan Dean Foster

So it seems, but this has not been true except in the last few days, I think.
....

I agree, just as I agree with your earlier comment that the SETI/BOINC servers should be robust enough to accept immediate reporting.

But the fact is that they aren't - not all the time, anyway. And nothing short of a hefty donation of new hardware and liveware (staff) is going to make them that way.

Think of it like your (always entertaining!) experiments with overclocking. We, the crunchers, are doing our best to 'overclock' the servers. Sometimes we stay within the range of stability, and get away with it. But sometimes we push the 'overclock' a bit too far, and the whole thing blows up in our face. Under those circumstances, it's better and more productive to back off a few cycles, rather than keep pushing to the limit and end up having to wait all next week for Fed-Ex to deliver the credits.

Totally agree Richard. This is a fair analysis of the situation.

As do I Richard. Although I should not have to have to change my Boinc client to have my work credited properly, I will have to stand down on this one until Eric can analyze the situation to see if there is some way to shuffle the server config or whatever to fix it.
I will probably update the Boinc client on all of my rigs tonight when I get home.

Oh, and Richard...I am glad you get some enjoyment out of my overclocking escapades. At least I am nothing if not persistant, eh? LOL.

---

"Freedom is just Chaos, with better lighting." Alan Dean Foster

I noted yesterday that after Eric made his "fix" that may or may not be the "real fix" then this page stopped working. This might be part or all of what he changed possibly to reduce data base activity. All of this is just my opinion and not known to be fact.

I noted yesterday that after Eric made is "fix" that may or may not be the "real fix" then this page stopped working. This might be part or all of what he changed possibly to reduce data base activity. All of this is just my opinion and not known to be fact.

[conspiracy mode on]
So that we can't see that they're blocking v5.9.0?
[conspiracy mode off]

I noted yesterday that after Eric made is "fix" that may or may not be the "real fix" then this page stopped working. This might be part or all of what he changed possibly to reduce data base activity. All of this is just my opinion and not known to be fact.

[conspiracy mode on]
So that we can't see that they're blocking v5.9.0?
[conspiracy mode off]

No I think just to reduce data base activity on the berkeley servers somewhat.

I'm using BOINC 5.10.28 and i have validate errors too. (only two, by the moment...)

The last is http://setiathome.berkeley.edu/workunit.php?wuid=172286435

Was reported 16 seconds after upload:

04/11/2007 21:21:13|SETI@home|Started upload of 09fe07ac.30195.25017.7.6.140_0_0
04/11/2007 21:21:17|SETI@home|Scheduler request succeeded: got 1 new tasks
04/11/2007 21:21:19|SETI@home|Finished upload of 09fe07ac.30195.25017.7.6.140_0_0
04/11/2007 21:21:19|SETI@home|Started download of 22mr07ac.8557.98648.8.6.196
04/11/2007 21:21:23|SETI@home|Finished download of 22mr07ac.8557.98648.8.6.196
04/11/2007 21:21:33|SETI@home|Sending scheduler request: To fetch work. Requesting 3156 seconds of work, reporting 1 completed tasks
04/11/2007 21:21:38|SETI@home|Scheduler request succeeded: got 1 new tasks
04/11/2007 21:21:40|SETI@home|Started download of 22mr07ac.8253.102738.3.6.49
04/11/2007 21:21:44|SETI@home|Finished download of 22mr07ac.8253.102738.3.6.49

---

Logan.

BOINC FAQ Service (Ahora, también disponible en Español/Now available in Spanish)

I noted yesterday that after Eric made his "fix" that may or may not be the "real fix" then this page stopped working. This might be part or all of what he changed possibly to reduce data base activity. All of this is just my opinion and not known to be fact.

I noticed that that page stopped working on Tuesday after the normal outage.