A long discussions about the equal share of resources between WCG projects in another thread (http://www.worldcommunitygrid.org/forums/wcg/viewthread_thread,31751_offset,0) inspired me to bring this idea, which I think would help us to make the world to a better place even much faster! Please let me try to explain:

Now we have about 10 projects which are running in parallel. Some of them are intermittent, some have technical issues. But biggest part of them are running sharing resources pretty equally. Doing so every single project would need a couple of years to complete.

Why don't we take one project first, crunch it down with the whole power of WCG and continue with the next one? We would need couple of months instead of long years to complete one particular project. Most of sciences we are crunching for are needed ASAP and would safe lives - saving one year on our side we would save a lot of lives all over the world!

You would say - there are some "open-end" or very long running projects like FAAH or HPF - they would block WCG completely. But in those cases it's still possible to split such projects in parts - like experiments in FAAH are. Every part should have a target, which would bring a benefit for scientists and give them work for quite a while. I assume for such a project it's not a big difference, to have the work done slowly with constant speed or to have it done with high speed, but in chunks while having breaks between chunks.

What would be needed if we would go this way - is a group of independent experts who would decide, which project goes first. Ok, it could possibly lead to an ethic problem. In this case it can be done on a "First Come - First Served" base. Long-running project, which have just finished it's nexts chunk of work, would come in the end of the queue.

Another problem, which can be faced doing this approach - crunchers dedicated to only one particular project. For whichever reason they may be not happy with this approach even if their prefered science will benefit from it. They should have a possibility to crunch the project they want if this project has available WUs in the time between full-power runs. We don't want to loose those people.
At the moment we have badges as additional motivation for crunchers. A lot of people are crunching for a project till they get the badge they want and then go for a next project where they still don't have the wanted badge. But the main idea in badges is, that it's much more enjoyable (don't know if it's the right word) to go for a short-term targets than to crunch without seeing the light at the end of the tunnel. If we would have short-running projects, it would be additional motivation for people and would make more fun for everybody here. E.g. something like "Let's crunch the Leishmaniasis away in the next two months!"

What we else have - are hardware-intensive projects like CEP2 and projects bringing only a small amount of work time to time. Those projects I would still leave running in parallel to whatever is currently running.

Regarding software developing efforts for WCG - it's nothing to do here. Just give a highest feeder priority to the project which has it's "showtime" and low priority to all the other projects. As soon the work for this project (or project chunk) is done - give the priority to the next project in the queue. And so on...


This idea is based on the "agile" methodology (I'm pretty sure our techs now what it is). I'm working in software development sector and we are using this software development methodology for a couple of years now. We are using it now everywhere, not only in software development. And that brought the company a much better "time-to-market" value then it has ever had before. I think it would be a great improvement for WCG as well.
If you want some more information abot this methodology, you can start here:
http://en.wikipedia.org/wiki/Agile_software_development

It would be great to know opinions on this from WCG techs, community advisors and all you community friends!