Posts by MikeMarsUK

1) Message boards : Current tests : Forum code update. (Message 2180)
Posted 19 Aug 2006 by MikeMarsUK
Post:
Bugs you may find are:

* Moving posts as a moderator might no longer work
* Editing posts as a moderator no long available (not actually a bug, a feature change)
* Deletion of posts may alway give the poster the reason \'obscene\'

They may have been fixed in the version you have.
2) Message boards : Current tests : New crediting system (Message 2178)
Posted 19 Aug 2006 by MikeMarsUK
Post:
...

I agree, if we add encryption to rosetta, we could use the internal benchmark. It\'s something to consider.


You\'d also need to encrypt the in-memory benchmark results as well as the upload and download, also obfuscate the actual benchmark code so that people can\'t follow what it\'s doing in a debugger and modify the values on the fly. (Remember if people can use a debugger, an out-of-process routine can do the same once the memory locations are known).

The recent debugging code (map files etc) would prove extremely useful to someone trying to do this, but aren\'t necessary.

The new system does the job, and should do it well after the bugs are ironed out, so why mess around with insecure ways of doing the same thing?
3) Message boards : Current tests : New crediting system (Message 2074)
Posted 16 Aug 2006 by MikeMarsUK
Post:
...

I\'m not sure I understand your point. I think you are saying \"But it should be okay to change the rules going forward, right?\" If so, I agree 100%. In fact, I am a *huge* fan of the way SETI is doing it. And optomized clients mean exactly that...they crunch the SAME job in less time. More work = more credit = more science! Who can argue with that?


I think your argument is based on a misunderstanding - An optimised client (i.e., a drop-in replacement manager) does no more science work than the original manager it replaced.

You\'d need to optimise the science application (which is what Rosetta has been doing since it started) in order to increase the science output. The main effect optimising a manager has is to increase the credit score (a few have other functions as well, such as managing farms of computers, or giving you additional statistics).

What is being discussed is firstly whether \'science = credit\' is a good idea going forward, and secondly whether it\'s a good idea going back, as well. My vote is \'yes\' for both.
4) Message boards : Current tests : New crediting system (Message 2050)
Posted 15 Aug 2006 by MikeMarsUK
Post:
Could you do the credit calc from Feb, and then estimate the older credit based on the recent boinc credit vs. recent work credit ratio?

i.e., if the Feb-August boinc credit is 10% higher than the Feb-August work credit, then apply the same ratio to the pre-Feb boinc credit to get an estimated pre-Feb work credit?

The reason that I am trying to think of ways to get the old credit into the equation is so that the exported XML can reflect the new system ASAP. I also would like to move to the new system exclusively.
5) Message boards : Current tests : New crediting system (Message 2007)
Posted 12 Aug 2006 by MikeMarsUK
Post:
...
I think it is a smart idea to introduce the new crediting system to Rosetta first as an alternative. In the long run however you need to decide on one system, which can only be the credit/model system, since it will be much fairer. The exact transition can be determined later and indeed the credit/model scheme allows quick and smooth adjusting in any case.
...


I would mostly agree with this, but the problem with having two sets of statistics is that it doesn\'t stop the \'cheating\' flame wars, and could even possibly make it worse. If it wasn\'t for that, I\'d quite like to have the two to compare.

Is there any chance of calculating the historical \'work credit\' based on average claimed credit per work unit? (i.e., for previous models crunched over the years).

If Rosetta publishes the user.xml stats etc, I think it should be the work credits stats which are published, but if this is to be done, it would be best to grant work credit for all decoys that the participant has processed. My guess is that Rosetta does have the data to do this (although perhaps not the time).






©2019 University of Washington
http://www.bakerlab.org