We need to get organized

[jasong]

I've said this before, but maybe I'm the only one anal retentive enough to care. If we got super-organized, we might be able to improve our accomplishments a little more. An example of a problem is that I'm sure there are people running Intels on FaD. If you're super hands-on about your hobby, that's great, but if you're looking to improve production for the project, the team, or yourself, you're wasting a TON of time running FaD on an Intel. (This is just an example)

I'm going to say something, and it's not a complaint, it's an advantage for the team if you like the idea. I've lost my job because of health problems related to my medication and the fact I work on a hot parking lot(Quitting solved the health problem, btw). I would like to set some some goals.

goals, in no particular order:

(1) Determine a way to calculate approximate points when a project is run in a selected configuration. Gaps are not a problem, educated guesses are permitted.

(2) Figure out what our goals our, and find a way to quantify them. For instance, assuming two projects take the same amount of time to install and we're advancing at the same rate in each project relative to the individual points rules, the thing that would matter would be the relative value of each computer configuration when it came to scoring. We could come up with a base configuration(1GHz AMD "something" running Linux) and come up with a point value relative to 100. If we REALLY wanted to be anal retentive we could have people rate projects from 1-10 so that the assignments will fit their personality or goals. Total the point values up for each project and multiplying them by the total value of all the configurations would tell us where the team needs to add or take away power. (Yes, I know I sound like a nerd)

(3)Alternatively, we could use the point values to trade time with others. I could run rc5-72 in exchange for someone else running LLR tests in Riesel Sieve, something which I've attempted before.(Actually I still want to do this, PM me if you're running an Intel on a project that likes Semprons better. If one does "too much" output, there can be a planned switch every once in awhile which could give one person "extra" time. Anything that likes Semprons is fine)

I know this sounds complicated, but taking advantage of some web page programming knowledge and a bit of algebra could simplify this. It's just difficult to express in regular language.

If anybody's interested, I could put forth some possible ideas in math form for review.

[IronBits]

There is also the tricky part, not everyone has full time net access, which many projects require, nor can work boxen run every project.
Good ideas, let's see more!

Send Xeons/P4s (HT to) to SOB :D

:cheers:

[jasong]

Originally posted by IronBits
There is also the tricky part, not everyone has full time net access, which many projects require, nor can work boxen run every project.
Good ideas, let's see more!
I've written a message on the Find-a-Drug boards and Riesel Sieve, I'm waiting for a response.

I've found this page, (http://www.mersenne.org/bench.htm) which could be useful for trading time to improve primality testing. The time it takes to test is also a definite concern. Not so much a concern with other projects.

[jasong]

I've lost my job because of health problems related to my medication and the fact I work on a hot parking lot(Quitting solved the health problem, btw).
Mental illness sucks. Other than that, I can't think of another reason why I would put this random sentence in there.

We need to figure some stuff out I think. I believe it would be helpful if we could start a review process to interview people and see what we can put where. A few ideas and a way to quantify them for strategic purposes.

First: An interview to determine:

(1) Run each individual computer for a week and, assuming your DC project is at lowest priority, note (a) the power of the machine(we can simply have people pick out their cpu from a table and use statistics to figure things out), and (b) out of total time, what percentage of total cpu power goes to project?(Unless computer is throttling, this should help us figure out where to ask people to go.)

(2) Rank projects listed from 1 to 10, 1-you hate the project, up to 10-an incredibly worthy project. Assuming someone assigns you a project, they would want to refer to this. Alternately, we could have a list of projects, and you number your favorite one,1, your next favorite 2, and so on, if they don't get a number, then they go to the end of the ranking and get equal "points."

(3) Notes: People should note any projects they have unusual prejudices about. For instance, I am strongly offended by United Devices projects censorship in their forums, so I try to avoid both the projects and their forums. Alternately, Find-a-Drug holds a special place in my heart as a high-quality project.

I can't think of anything else, but I hope I've breathed life into this thread.

Edit: I noticed we're about to experience multiple road kills in Rosetta. Maybe that's where we need to start.

edit 2: picking a leader might help.