I can tell you that when I originally signed up I intended to support primarily the cancer projects. I used the default settings which enabled all the projects except for one computationally intense project which I forget what it was. Somehow even though I really didn't intend to support the Aids projects they ended up with my longest runtimes and most points before I became aware of how unfair this was about 2 years ago.

Now I am aware and know how to limit my contribution to the Aids projects if I choose.

The bottom line for me is that probably 95% of volunteers never come into the forums and don't know much about how to control the WCG app. They install the app with default settings and a 1-day deadline project grabs a disproportionate share of the time they donate.

I feel that the Aids projects have been stealing time from the cancer projects with the 1-day deadline. The big thing that I don't understand is why WCG allows this,

Shorter deadlines mean that the project jumps the queue ahead of longer deadlines and therefore an enhanced share of units processed.

Lengthening queues mean that there are fewer units returned for the first few days and then that project get its fair share thereafter, so a lower proportion.

It is always possible for us to manipulate the processing selection by using a combination of Device Profile selection and app_config processing, but how many participants actually know how to do both, or even either.

Only those active on fora are likely to know.

Mike

The same thing has happened to me yesterday. One of machines got enough tasks for all day, and other got nothing. There was some talk about randomizing WUs. As I was talking earlier, FAH2 is more suitable for people running their machines 24/7 and more dedicated to project, regardless of computing performance (if possible to finish tasks in 24h), because of project nature, but one cruncher here misunderstanding that and accused me for being arrogant and ignorant! If we must find some compromise for all crunchers to be happy, and not to slow down research too much, maybe Tech should consider some 36-48h deadline in future. 7d deadline experiment should finish soon anyway.

Lol, yes, and they're rubbing it in

Statistics Date		Total Run Time
(y:d:h:m:s)		Points Generated		Results Returned
01/14/2020		0:001:05:54:49		7,949		7
01/13/2020		1:178:23:37:41		2,769,525		2,377
01/12/2020		1:353:09:37:22		3,714,642		3,191
01/11/2020		2:211:10:54:45		4,791,364		4,169
01/10/2020		1:349:09:42:58		3,635,072		3,164
01/09/2020		1:143:05:32:25		2,429,763		2,175
01/08/2020		2:329:12:30:10		5,149,338		4,614
01/07/2020		6:245:08:58:04		12,204,578		10,821

Not even a quarter of what was before the JDI experiment.
Anyone interested to experience day seven, oh wait, that's when the light comes on.

This was my exact reason for limiting the project on my hosts. Every time I turned around work units from this project were eating up half the threads on my CPUs and other projects I was trying to run were suffering because of it.