![]() |
|
|
#45 | |
|
If I May
"Chris Halsall"
Sep 2002
Barbados
2·5·7·139 Posts |
Quote:
When the "Let GPU decide" option is set, the system's SQL query sorts by rand() if the UA is MISFIT. The thinking is to spread out automatic system's assignments across the whole range we own, to include not just the less profitable (as defined by Factors Found / GHz Days Saved). (I can't use the rand() sort for human interaction because (as you've demonstrated) it can cause different assignments than what the preview shows.) The reason for this is I wanted to give a fair share of assignments to those who are using a fully automated environment the opportunity to work on lower TF levels as well as the "last step". Also, I wanted to bring the "Estimated Completion" time below 30 days for the 63M and 64M ranges (no real reason, I just wanted to). Sorry... I should have stated this publicly so everyone understood the different behavior. |
|
|
|
|
|
|
#46 |
|
Sep 2008
Bromley, England
43 Posts |
|
|
|
|
|
|
#47 |
|
"Bill Staffen"
Jan 2013
Pittsburgh, PA, USA
23×53 Posts |
That first paragraph makes so much more sense to me now than it did then. I don't know if you're still working on that, Challsal, but I do know you're busy. If you want, and have a spec of the api you can send, I can probably modify the work fetcher to use it rather than the HTTP POST. I have some free time at home and am a programmer.
|
|
|
|
|
|
#48 | |
|
If I May
"Chris Halsall"
Sep 2002
Barbados
230028 Posts |
Quote:
The API is actually already online (and has been for months). The issue is I have to spend some "quality time" ("calm time") with GPU72 to provide a web-based GUI interface to create and make available CIDs. Soon.... |
|
|
|
|
|
|
#49 |
|
"Bill Staffen"
Jan 2013
Pittsburgh, PA, USA
23·53 Posts |
Fair enough. Since it's already spitting out CIDs for the p-1s I didn't think it would be a thing to spit them out for the TFs, too. It's a standing offer in case you change your mind.
|
|
|
|
|
|
#50 |
|
If I May
"Chris Halsall"
Sep 2002
Barbados
2×5×7×139 Posts |
|
|
|
|
|
|
#51 |
|
Aug 2012
New Hampshire
23×101 Posts |
F.Y.I. MISFIT does its uploads via HTTP get to the public gimps "upload results" page. If you wanted to author a class in .net that can upload the results via primenet API I would put that into MISFIT.
If you created the class you could then go all the way an create a .NET console uploader and you would not need the perl uploader anymore. (nothing against perl, just completing a thought) Later you could do the same for GPU72 api and I'd put that in MISFIT too. (incase you wanted to do something....) Last fiddled with by swl551 on 2013-03-21 at 21:10 |
|
|
|
|
|
#52 |
|
"Bill Staffen"
Jan 2013
Pittsburgh, PA, USA
1101010002 Posts |
Is the spec online somewhere?
|
|
|
|
|
|
#53 |
|
"Mr. Meeseeks"
Jan 2012
California, USA
216810 Posts |
|
|
|
|
|
|
#54 |
|
Basketry That Evening!
"Bunslow the Bold"
Jun 2011
40<A<43 -89<O<-88
3·29·83 Posts |
|
|
|
|
|
|
#55 |
|
"Bill Staffen"
Jan 2013
Pittsburgh, PA, USA
23×53 Posts |
Oh, is misfit windows only?
|
|
|
|
![]() |
Similar Threads
|
||||
| Thread | Thread Starter | Forum | Replies | Last Post |
| a lyric finding tool | MattcAnderson | Soap Box | 1 | 2017-01-01 15:18 |
| Auto-XYYXF tool | fivemack | NFS@Home | 3 | 2016-07-06 04:01 |
| TF fetching/reporting toolkit for Linux | swl551 | PrimeNet | 20 | 2014-06-19 15:00 |
| GPU72's fetching/reporting API | Aramis Wyler | GPU to 72 | 44 | 2013-04-10 13:41 |
| tool to help with Top 5000 list | Mini-Geek | Lounge | 6 | 2008-05-04 03:19 |