PDA

View Full Version : DNETC@HOME


Rusty
11th September 2010, 11:49 AM
http://sesef.pl/dnetc/img/dnetc.png

About DNETC@HOME
DNETC@Home is a wrapper between BOINC and distributed.net.
You can read more about it here.... (http://www.distributed.net/Main_Page)


http://dnetc.net/

DigiK-oz
13th September 2010, 06:04 PM
Well, I myself do not like wrapper projects....RC5-72 is already in the Vault, so do you really want to double the impact a single project has on scores?

Rusty
24th September 2010, 11:53 PM
Could this a step towards Dnet running BOINC

Beyond
25th September 2010, 02:17 PM
Could this a step towards Dnet running BOINC
I may be incorrect, but from the rumblings I've heard, distributed.net is not happy about the DNETC project wrapping their client. This is probably why they won't make simple changes that would make DNETC more useable, such as allowing 1 WU/GPU, multi GPU/WU works fine with the stream RC5-72 client but has problems when parceled into BOINC WUs. The credit system is also suspect. While both Collatz & MilkyWay have posted the reasoning behind their credit systems, DNETC AFAIK has only stated that they wanted to give credit somewhere between Collatz & MW. Then they proceeded to raise their credit to ~ 1.5x - 1.6x that of Collatz (for ATI 5xxx GPUs) and also much higher than MW. The project IMO is a haven for credit at any cost crunchers and as DigiK-oz states above is just a wrap of the RC5-72 client. Personally I don't think we need 2 iterations of RC5-72 in the Vault.

Xaverius
25th September 2010, 02:27 PM
We from the Dutch Power Cows are also in contact with distributed.net because of the points dnetc@home is given in de distributed.net stats. The points should be awarded to the teams who participate in dnetc@home and not to dnetc@home.

Second, we stated it would be more fair if distributed.net would make their own BOINC-port and the statistics would be one. We asked them to remove dnetc@home from the stats like google was removed from the folding@home stats when they used their roolbar for folding@home back in 2002.

runesk
26th September 2010, 06:55 PM
We from the Dutch Power Cows are also in contact with distributed.net because of the points dnetc@home is given in de distributed.net stats. The points should be awarded to the teams who participate in dnetc@home and not to dnetc@home.

Second, we stated it would be more fair if distributed.net would make their own BOINC-port and the statistics would be one. We asked them to remove dnetc@home from the stats like google was removed from the folding@home stats when they used their roolbar for folding@home back in 2002.

These wrapper projects tend to have these start up issues regarding credits.

Back in the days when yoyo@home made a wrapper for Muon1 DPAD, one could join yoyo using a username using the Muon1 team indicator, and one would get credits both at yoyo and a Muon1. This were changed and all credits granted were removed for Muon1.

My opinion is that no project that will give this multi-credits should be added to the vault.

.R