Hey Carl - 'Emma', machine #140, has leapt from about 50th place into 1st place on 'recent average credit' since she uploaded her model - she's also leapt from about 50th to 3rd place on 'total credit'. The model appeared to be worth about 5,400 in 'total credit' and she was getting about 295 in 'recent average credit' - methinks something is going wrong with the scoring system when a model is uploaded..?

<a href="http://www.nmvs.dsl.pipex.com/"><img src="http://boinc.mundayweb.com/cpdn/stats.php?userID=6&amp;team=off&amp;trans=off"></a>

HeHe...

I just posted about this same thing in your "successful completion and upload" post. :-)

It appears as though credits are calculated twice or something? After completion of this single WU, my accounts page now shows 12708.83 credits, and 1014.34 RAC. I hate to complain, but this seems excessive. ;-)

It's a bit more than double the credits I've accumulated from the trickles, (so I'm not sure where the extra numbers come from. Perhaps the size of the archives stored on the local machine have something to do with it? I know it's not implemented yet, but I know BOINC is capable of using external hosts for storage, and granting credit for the amount stored.

<a><img src="http://boinc.mundayweb.com/cpdn/stats.php?userID=18"></a>

hehe, I noticed that Heff &amp; went WTF, Good Job On Getting One Done Though ... :)

whoops I think it's back to normal -- apparently I overlooked that boinc will grant the full credit at the end of a run, for which you would have already been granted from the trickles anyway. congrats on getting full runs in, I think these are some of the first? they look like good, stable runs:

<A HREF="http://climateapps2.oucs.ox.ac.uk/cpdnboinc/result_graph.php?result_id=721">heffed</A>

<A HREF="http://climateapps2.oucs.ox.ac.uk/cpdnboinc/result_graph.php?result_id=296">uk_nick</A>

Edited: My Abox uploaded RunID#33375 (035i_40029071)about 12 hours ago. On a whim, I looked at the Machine-ranking page and noticed that Abox is positioned higher than when I checked some days ago. Odd, I thought.

Second run of the day just uploaded from Abox. I thought the first garnered extra credit but none logged for the second....
Another edit: Checked yet again and found what looks like an extra 2500 credits. Mumble, mumble.
Embarrassing.


Ehhh -- all is not completely well with the new runs either: One run with the newly downloaded 3.04 client barfed, then restarted. First I've experienced it.
CPDN Monitor got quit request...
Detaching shared memory...
2004-09-26 23:48:17 [climateprediction.net] Result 2zub_000161172_0 exited with zero status but no 'finished' file
2004-09-26 23:48:17 [climateprediction.net] If this happens repeatedly you may need to resetthe project.
Starting model in /home/jim/CPDNboinc/projects/climateapps2.oucs.ox.ac.uk_cpdnboinc...


BTW, when Abox was finishing its last Model solo, it trickled at five hour (and six second) interval. (Not too shabby, IMO, for a P4 2.8 GHz box I was temporarily using as General Purpose machine.)

&gt; Ehhh -- all is not completely well with the new runs either: One run with the
&gt; newly downloaded 3.04 client barfed, then restarted. First I've experienced
&gt; it.
&gt; CPDN Monitor got quit request...
&gt; Detaching shared memory...
&gt; 2004-09-26 23:48:17 [climateprediction.net] Result 2zub_000161172_0 exited
&gt; with zero status but no 'finished' file
&gt; 2004-09-26 23:48:17 [climateprediction.net] If this happens repeatedly you may
&gt; need to resetthe project.

Hi Jim,

You'll get the last 2 errors when BOINC 4.09 (but not 4.05) detects that an application has exited without creating a <b>boinc_finish_called</b> file in the slots directory it was using. So they're consistent with your CPDN client barfing :)

<a href="http://www.teampicard.net"><img src="http://www.teampicard.net/templates/fisubice/images/phpbb2_logo.jpg"></a><a href="http://climateapps2.oucs.ox.ac.uk/cpdnboinc/team_display.php?teamid=3">Join us here</a>

Thanks, Ian,

The good part is that it auto-restarted; would that other boinc errors auto-recovered as well (or, at least didn't take down all runs on P4's or multi-CPU boxes). A class of comm errors being a case in point.

Well, there's another good part: No repetition of the error. (Yet...?)

As to the number crunching aspect of this thread, of the eight runs completed/replaced this week, six replacements are slower, two slightly faster than the runs replaced on the same four machines. For the four finished yesterday (and wee hours this morning), it may be too soon to make that claim definitively, though. (Luck of the draw on parameter mixes, eh?)

Jim

&gt; As to the number crunching aspect of this thread, of the eight runs
&gt; completed/replaced this week, six replacements are slower, two slightly faster
&gt; than the runs replaced on the same four machines. For the four finished
&gt; yesterday (and wee hours this morning), it may be too soon to make that claim
&gt; definitively, though. (Luck of the draw on parameter mixes, eh?)

It's not necessarily down to luck of the draw, Jim. One of my P4 HT boxes was grinding to its knees last week. At first I thought it was something to do with its parameters, but on Saturday I decided to defrag (even though Windows said it wasn't necessary) and reboot. It's now back up to trickling at its <a href="http://climateapps2.oucs.ox.ac.uk/cpdnboinc/trickle.php?resultid=142009">normal frequency</a>.

<a href="http://www.teampicard.net"><img src="http://www.teampicard.net/templates/fisubice/images/phpbb2_logo.jpg"></a><a href="http://climateapps2.oucs.ox.ac.uk/cpdnboinc/team_display.php?teamid=3">Join us here</a>

I was about to start a thread to the same effect until I saw this here. After my computer uploaded the final results, my credits jumped from ~6000 up to over 14,000. As people have mentioned before, I hate to complain about it, but I'm getting credit for work I didn't do. Something is wrong here...

*AMD Athlon XP 3000+ 2.17 Ghz, crunching @ ~2.5 s/ts
*Intel Pentium III 500 Mhz, crunching @ ~10.3 s/ts

I commented to Carl in an email on September 21st that this 'doubling' was still happening - Carl replied;

"odd, it looks like a leftover problem from beta test machines that went
through the "credit adjustment", so eventually those will all balance out."

<a href="http://www.nmvs.dsl.pipex.com/"><img src="http://boinc.mundayweb.com/cpdn/stats.php?userID=6&amp;team=off&amp;trans=off"></a>

I'm not very clear on the "balance out" thing. Does this mean that nobody's going to change my credits from 14,500 back to where they're supposed to be?

:-D


*AMD Athlon XP 3000+ 2.17 Ghz, crunching @ ~2.5 s/ts
*Intel Pentium III 500 Mhz, crunching @ ~10.3 s/ts

&gt; &gt; As to the number crunching aspect of this thread, of the eight runs
&gt; &gt; completed/replaced this week, six replacements are slower, two slightly
&gt; faster
&gt; &gt; than the runs replaced on the same four machines. For the four finished
&gt; &gt; yesterday (and wee hours this morning), it may be too soon to make that
&gt; claim
&gt; &gt; definitively, though. (Luck of the draw on parameter mixes, eh?)
&gt;
&gt; It's not necessarily down to luck of the draw, Jim. One of my P4 HT boxes was
&gt; grinding to its knees last week. At first I thought it was something to do
&gt; with its parameters, but on Saturday I decided to defrag (even though Windows
&gt; said it wasn't necessary) and reboot. It's now back up to trickling at its <a> href="http://climateapps2.oucs.ox.ac.uk/cpdnboinc/trickle.php?resultid=142009"&gt;normal
&gt; frequency</a>.
&gt;
&gt; <a href="http://www.teampicard.net"><img> src="http://www.teampicard.net/templates/fisubice/images/phpbb2_logo.jpg"&gt;</a><a> href="http://climateapps2.oucs.ox.ac.uk/cpdnboinc/team_display.php?teamid=3"&gt;Join
&gt; us here</a>
&gt;

Greetings from the upper left US coast, Ian, where we are in a series of fine days (it can't last),

Three of my machines run Linux.

My sluggard (WinXP P4 3.0 box) was defragged (a word untranslateable into Linux) and re-booted after TS 288 of the second Model (and at a reasonable compromise with the first Model).

Nonetheless, this slowest machine, the only one running M$, fell from two Models running @ 3.1 sec/TS to two Models running at 3.4 sec/TS. (A lot to blame on parameters, IMO.) I did the numbers on a similar thing weeks ago, in Alpha or Beta, but am too tired to do it now. The deterioration means days of difference, if not a week, in finishing the new pair.

My P4 2.8 box (Abox) in Linux is significantly faster than the 3.0 machine (Cbox) in Windoze. 'Twas true in the first pair of Models, tis also true in the second pair -- even more-so.

Part of it is, of course due to parameter mixes -- but part is due to the wunnerfulness of BillG, I think. (I pronounce it 'bilge' [with long history].)

I've only seen Windoze say defrag was necessary once -- and it was on a friend's XP P4 3.0 machine, dedicated CPDN partition. (Never defragged -- after ~ a year of running CPDN Classic....) Depends on the size of the partition vs. garbage, IMO. ('Defrag' is one of the reasons three of my four machines run Linux -- despite Linux add-ons' serious limitations.)

Best regards,
Jim

Just liked to mention that it happened to me too. My RAC is about twice as high than before and credits also increased enormously after host #26540 uploaded result #210027 an hour ago.
 <br><img src="http://boinc.mundayweb.com/cpdn/stats.php?userID=39&amp;trans=off">

&gt; Just liked to mention that it happened to me too. My RAC is about twice as
&gt; high than before and credits also increased enormously after host #26540
&gt; uploaded result #210027 an hour ago.
&gt;  <br><img> src="http://boinc.mundayweb.com/cpdn/stats.php?userID=39&amp;trans=off"&gt;
&gt;

Greetings, Finla,

I discovered, a week after upgrading my M$ box to 4.09, that it took it upon itself to install the [censored] screensaver version. A day after killing that disgusting feature, the averages are slipping -- ever so slowly.

(Re. unexpected increase in credits, I think most of us on the 'top page' have benefitted to some extent from this anomoly. Would that it were otherwise.)

Please check to see whether the thingee installed the screensaver unbidden on your machine.

Regards.


We have met the enemy and he is us -- Pogo

Thanks for your reply, astroWX. :o)

I also checked my 2 machines - which have both been using 4.09 since about 14 days ago - and everything seems to be okay with them...
Happy crunching!

groan...

I thought this extra-credit horse was dead but apparently not.

My Abox was in third, or fourth, or fifth place last week. Yesterday, it uploaded two models (one of which went cold, with parameters I wouldn't have suspected).

This morning, I looked at Top Machines and found Abox floating on top. This should not be! I don't know how much extra credit it was given. More than enough to make a significant difference, to be sure. Sorry, all.



We have met the enemy and he is us -- Pogo

Apparently I've just been credited with a whole extra run (jumped from 12000+ to 17000+).

yeah... something still isn't right IMHO. Look at <a href="http://macg.no-ip.info:5520/boinc/team_hist_graphs.php?proj=cpdn&amp;teamid=45&amp;pmode=r">this graph</a> on my stats site. You can clearly see several big jumps when people return work units. If the credit that is granted along the way doesn't match up to the final claimed credit, I could see some adjustment going on when you finish but this seems a bit extreme.
<br>
----------------------------
A member of <a href="team_display.php?teamid=45">The Knights Who Say Ni!</a>
Yet another stats page: <a href="http://boinc-kwsn.no-ip.info">http://boinc-kwsn.no-ip.info</a>

About my stats.
Anything to do with <a href="http://climateapps2.oucs.ox.ac.uk/cpdnboinc/workunit.php?wuid=251850">work unit 251850</a> being used five times and the didn't need on the bottom line?

very mysterious...
it seems that there's no influence in which boinc version is used. it also doesn't have something to do with old beta testers. many accounts in our team (hardwarelabs) joined after the official cpdn boinc start and got enormously much credits after uploading their models...