Foldit is undergoing a subtle modification to its energy function. Check out the details in our latest blog entry!
We are also introducing a new update group called 'newchapter' so you can try it out right away. See this post for instructions on how to safely switch update groups. Be sure to replace 'devprev' in those instructions with 'newchapter'.
Enjoy!
First - the client is painfully slow.
I opened the ED puzzle and did some shake / wiggle in .2, it seemed very slow. Then I tried a rebuild on 3 sections.
The first stop on the graph was fine. The second stop superimposed itself on the first, the third on the second… which meant that there was only only stop on the graph, not three. It only registered the last pose done.
So I did a Blue Fuse v1.1 to go to 1.0 CI, and got out the stopwatch. 25 minutes and 18 seconds later, the script finally finished. That's not just slow, that's wading through cement.
My computer is a Mac a year and a half old (top of the line). I don't have the patience for this or the desire to go buy a new computer.
When you have removed the cement, I'll test it again.
I had the same experience as AD doing manual rebuilds; it would help to make a comparison between this new version and the old version on the same puzzles. Without being able to do that I believe NewChapter is extremely slow wiggling. Scripts as simple as Blue Fuse and Snapper are taking forever.
What am I doing wrong? I change "main" to "newchapter" but each time I run the software "main" reappears in the text.
Do you have an open client? If you close a client, the last client closed will override the options.txt file.
When I pull a section in the symmetry puzzle, it crashes (repeatably):
***** STARTING THREAD ActionLocalMinimizePull
core.scoring.CartesianBondedEnergy: Creating new peptide-bonded energy container (41)
ERROR: !core::pose::symmetry::is_symmetric( pose )
ERROR:: Exit from: ......\source\src\core\scoring\ScoreFunction.cc line: 629
UNHANDLED EXCEPTION
1: RaiseException +72 bytes (no line)
2: library_main +33223925 bytes (no line)
3: library_main +31113281 bytes (no line)
4: library_main +31113488 bytes (no line)
5: library_main +22865679 bytes (no line)
6: library_main +21124968 bytes (no line)
7: library_main +21094688 bytes (no line)
8: library_main +6713472 bytes (no line)
9: library_main +6610933 bytes (no line)
10: library_main +6710473 bytes (no line)
11: library_main +4403950 bytes (no line)
12: library_main +4407094 bytes (no line)
13: BaseThreadInitThunk +14 bytes (no line)
14: RtlInitializeExceptionChain +132 bytes (no line)
15: RtlInitializeExceptionChain +90 bytes (no line)
We've made some adjustments to wiggle to make it not run as long for a given number of iterations, and also make the length of each iteration a bit more consistent. Blue Fuse 1.1 on the ED should take about as long to run in newchapter as it does with less precise wiggle in main. An update to newchapter will be posted today with this update.
We'll be posting an update to newchapter shortly that should address the slowness of wiggle, the pull crash mentioned above, and bands not pulling (http://fold.it/portal/node/996591).
But you likely didn't see notice this serious bug from my comment…
"Then I tried a rebuild on 3 sections.
The first stop on the graph was fine. The second stop superimposed itself on the first, the third on the second… which meant that there was only only stop on the graph, not three. It only registered the last pose done."
Mike tried this also with the same results (see his comment below mine above).
Essentially, you can't do manual rebuild in the newchapter client - the graph is not working.
Making comparisons possible. And if local saves from main or devprev can be used it would even make more evaluation possible.