stephenbrooks.orgForumMuon1GeneralReduction in badgers?
Username: Password:
Search site:
Subscribe to thread via RSS
Dave Peachey
2006-02-19 06:22:21
Stephen,

Any reason for the reduction in my number of submitted WUs/Mpt score today (Sunday 19th)?  I noticed it had gone down by 240-odd results/110k Mpts in the Sunday morning update.

Is this a general results tidy-up or something wrong with me in particular?

Dave
K`Tetch
2006-02-19 22:51:16
how does it compare with the sendresults.  Could have been depue-deletion, if the same stuff was submitted to two servers.

Not sure what yours means though - if you're talking about decayRot, and how your mpts per result are lower, then its probably because you're not generating very many high yield results.  lots of small simulations, and you'll probably only get 30MPTS per sim unless its a local high.
Stephen Brooks
2006-02-20 03:45:21
I'm thinking the / just means a slash and not a ratio.

The stats server is currently acting like an incompetant, though honest, debit card company.  It sometimes puts things on twice and then has to remove one copy.  This multiple sending is mainly because I'm currently using Windows' godawful FTP program and I'd be much better to write my own client.  I've already almost written an HTTP/1.1 client that will be used in v4.43. The FTP may take a version or two more...

The best way is to check historically and normally you have a "doubled" production a few days before you find points removal.
Dave Peachey
2006-02-20 03:55:53
Stephen,

Ah, b*ggered stats server - what a surprise Frown Oh well, I'll just carry on regardless - provided my scores continue their general upwards trend (week on week) then it'll all work out in the end!

Cheers
Dave
Stephen Brooks
2006-02-20 05:58:36
In fact, your scores ought to end up correct, just getting there by an unusual route.
tomaz
2006-02-20 06:16:10
Can only a part of a larger file be lost ?  Due to windows ftp problems I manualy collect results.txt files, make one file and manualsend it.  I send one at Friday, it has about 600 kb.  There were also few DecayRot results among others.  They appeared later in stats.  But what catch my eye was that among result there was also high PhaseRotB result (about 4.2422..) which never appeared in stats.  On the other hand I think there was everything OK with general Mpts update.  Later that day I also send another file which contained another high PhaseRotB result (4.2421..) which also did't appeared in stats.  Was something wrong with that results and you removed them or there is something elese wrong?
Any point in sending that files again ?  (I make backup of every file send).
Regards
Tomaz
Stephen Brooks
2006-02-20 13:33:16
I changed the way it looks for maximum results just recently - it now only counts the "re-checked" ones, which usually happens when you generate one that is the highest so far in your results.dat file.  But I might move it back to the old way of doing it because those of you who have the samplefile in results.dat will then not get very many actual re-checks.  (I'll then delete the rogue result that should have been re-checked and wasn't, that caused me to make that change to begin with).
Fozzie
2006-02-21 01:11:15
I check all my uploads with Muon Cockpit now before they go and it looks like I am owed 750 thousand credit.

I hope this turns up soon, I would hate to have to take that sort of hit.
Stephen Brooks
2006-02-21 02:43:06
Do you know which upload it was that was lost?  (and which server it went to)

[edit] Actually, I touched one of your files (on the just-retired optimisation PhaseRotC_bigS1 - a 167MB file) so the server is now giving it a full re-count.
Fozzie
2006-02-21 12:31:40
I have alot of offline boxen.

Is PhaseRotC_bigS1 now defunct and garnering zero credit?

If so I'll need to speedily get round my offline boxen and sort the lattices and sample files.
Stephen Brooks
2006-02-23 03:33:47
bigS1 is still being credited but you are encouraged to move to the new lattices.
Fozzie
2006-02-24 08:00:12
I have pushed out the new files.

How do i ensure BigS1's aren't done from now on?
Stephen Brooks
2006-02-24 08:59:48
The client should update itself from the web and remove PhaseRotC_bigS1.txt from the lattices directory, replacing it with the new one (DecayRot.txt).

If you're offline, you can delete and update the file manually, just download the txt file from the links here:
http://stephenbrooks.org/muon1/lattices/
K`Tetch
2006-02-25 19:29:13
to clarify, if hat stephen says isn't exactly clear to you (as it wasn't to me after only one reading)

1) go to the lattices directory
2) delete phaserotc_bigs1
3) place decayrot lattice file in directory.

9it can be downloaded from the link stephen gave - just 'save target' on the lattice name.
Stephen Brooks
2006-02-27 03:15:47
quote:
if what stephen says isn't exactly clear to you (as it wasn't to me after only one reading)
Insert part B into knob 4, stand on bottom.
BOHICASETI
2006-02-27 08:04:51
You forgot to glue it counterclockwise.
HaloJones
2006-03-02 00:29:54
quote:
Originally posted by Stephen Brooks:
I've already almost written an HTTP/1.1 client that will be used in v4.43. The FTP may take a version or two more...


An HTTP client would be very nice and would be likely to get my farm back on-line.
: contact : - - -
E-mail: sbstrudel characterstephenbrooks.orgTwitter: stephenjbrooksMastodon: strudel charactersjbstrudel charactermstdn.io RSS feed

Site has had 25161067 accesses.