What exactly is the difference? Do we get a better calculation or 'nicer' wu if we set it to 24 hours rather than 1 hour??

Please let me know

---

You get less network traffic.

---

dag
--Finding aliens is cool, but understanding the structure of proteins is useful.

If I handed you a pile of plastic building blocks and told you to assemble them into a fairly straight line... starting with THIS blue square... and it took you an hour to assemble all of the other blocks... and then I said "now that's great, do it again... same blocks, but this time start with this green rectangle"... it would take you another hour to assemble all of the other pieces.

If you selected a 2hr preference, you might produce two models with all of the pieces. If you selected a 24hr preference, you might produce 24 models. All of the models are different. And all are meaningful scientific results.

If you selected a 2hr preference and sent back the results, you'd then be handed another set of blocks and asked to assemble more models. ...and it's possible they would be identical to the blocks you just had... there are still billions of other ways to assemble them that need to be examined.

If your computer is only on for 24hrs a week, then it might be cutting it pretty close to have a 24hr work unit with a one week deadline. If you compute 24hrs a day, it's a different story. So the preferences are there just to give you greater flexibility to match your preference to how you use your computer.

---

Add this signature to your EMail:
Running Microsoft's "System Idle Process" will never help cure cancer, AIDS nor Alzheimer's. But running Rosetta@home just might!
https://boinc.bakerlab.org/rosetta/

The project wants around 10,000 models/decoys while we're testing new algorithms, and around 100,000 models/decoys while we're working on CASP for each protein. For the 10,000 setup - the science doesn't care if they get 1 model from 10,000 different contributors, 10 models from 1000 contributors, 100 from 100 contributors, or 1000 from 10 contributors or any combination that adds up to 10,000. For the science, it doesn't matter how they get the 10,000 models.

For each computer that starts work on a WU, we have to download several megs worth of data, and when we're done with that WU, we upload a few hundred kilobytes of data back. If your system isn't having problems, then the longer WU times allow contributors to reduce their bandwidth usage AND the bandwidth usage of the Rosetta servers.

---