Make the code open source

Case number:699969-986241
Topic:General
Opened by:ChristianK
Status:Closed
Type:Suggestion
Opened on:Friday, June 5, 2009 - 08:42
Last modified:Friday, April 22, 2011 - 21:24

As the results of the folding go into the public domain, why keep the code of fold.it closed?

At the moment there are a lot of open issues on the bug tracker and the developer team doesn't seem to have enough manpower for all of them.
Making fold.it open source might help with the manpower problem.

(Fri, 06/05/2009 - 08:42  |  15 comments)


zoran's picture
User offline. Last seen 6 years 37 weeks ago. Offline
Joined: 11/10/2007
Groups: Window Group

foldit code will eventually be shared for all non-profit purposes. we expect the game to evolve out of the beta first. also, we have to figure out how to share the code, without allowing people to use their own clients, and cheat by reporting modified scores.

mcurtis's picture
User offline. Last seen 9 years 19 weeks ago. Offline
Joined: 05/09/2008
Groups: None

Foldit relies substantially on open source packages such as, Boost, boinc, glut, libircclient, etc.

And yet, another year has passed, and the Foldit source remains closed.

:-(

zoran's picture
User offline. Last seen 6 years 37 weeks ago. Offline
Joined: 11/10/2007
Groups: Window Group

I have good news on this front. As was out intention since the inception of this project, UW will issue a free license for all non-commercial uses of Foldit. It will be available both as a standalone interface that can be linked with other biochemistry software, and as an extra feature for the Rosetta protein structure code available through the long established Rosetta Commons. Commercial licenses Foldit interface will also be available for an annual fee.

Some of the difficulties of pushing this through stemmed from the fact that Rosetta biochemistry code included with Foldit already has a distribution protocol that we had to adhere to.

Joined: 05/19/2009
Groups: Go Science

That sounds great.

If I understand the system right at the moment client have to upload the final version of the protein anyway to the server to report their score.
The server could recalculate the score to be sure. People will probably also be able to hack enough into fold.it to change the score that gets reported without having the source code.

Having a replay like function where everyone can look at all action that were taken to get to the final score would also help to catch cheaters.

Joined: 10/30/2009
Groups: None

Thinking closed source is a protection against hacking or cheating is a big mistake. You can just delay them (and not very much if your software is really interesting to hack) by closing the source.

The most important is to improve the quality of the code of Foldit, and quality could be greatly improved if more programmers could contribute to the debugging or development.
That's why open source could be very interesting.

Don't only open the source, give the right to reuse it too (with AGPL for example)

Joined: 09/18/2009
Groups: SETI.Germany

Open-source? Yes, yes and YES!
I very want to have it this way.

There are many users here, wo have the technical know-how to fix bugs and make improvements.
Often, minor bugs are reported here, who could be removed quickly, but are not, because of the lack of manpower (and time?).
So, by open-source you could easily get more people helping you (people living at other time-zones), and I trust everyone here, that he has in his mind, that the way to solutions are the most important point of the game, and not the score.

Okay, safety and cheating issues is a point, but as the users ChristianK and serval2412 said before, there IS a way to prevent cheating.

regards with best wishes,
Alex

Joined: 09/18/2009
Groups: SETI.Germany

ChristianK:
Your idea to submit not only the solution but also the moves towards it is very good!
The game should not only upload the solution, but the UNDO-info, where the states are already stored.

This brings me to the suggestion that the game in general should not only store the current state, but also the undo-content, because so you can save many states in one file, which would increase file size, but reduce file quantity, which would also make group-sharing solutions more lucid.

beta_helix's picture
User offline. Last seen 6 days 14 hours ago. Offline
Joined: 05/09/2008
Groups: None
Status: Open » Closed

Marking the status as DONE.

Here is the link to the Rosetta Commons for those interested: http://www.rosettacommons.org/

Joined: 09/18/2009
Groups: SETI.Germany

That's some nice news :)

Tlaloc's picture
User offline. Last seen 10 weeks 6 days ago. Offline
Joined: 08/04/2008
Groups: Mojo Risin'

Question: If I understand the licensing right, it is impossible for me to get the source code without paying at least $35,000 per year. I am not an educational institution, and I am not a company. I just want to hack in some changes into the sources. David Baker's seemed to indicate that I should be able to get the sources, but my reading of the license indicates that I can't.

ipatrol's picture
User offline. Last seen 6 years 8 weeks ago. Offline
Joined: 12/04/2010
Groups: None
Status: Closed » Open

Still not done. Look, you have to realize that no amount of DRM will prevent people from trying to hack this, and why should you honestly care about that? The only thing to keep secure is the points system, so you should process that on your server instead of client-side. Your rosetta@home system can augment that by securely evaluating submitted proteins for scoring. BOINC already does this internally and there's a great essay on this very issue. If people can create new clients that implement innovative ways to fold protiens, then so be it. Mabye we can combine r@h credits with foldit points to allow people to submit partially-finished protens to the rosetta system for further analysis. In a project like this, with lives that could be saved, openness can only make us stronger. If people can be trusted to fix protiens, then why not the code itself?

Joined: 05/17/2008
Groups: None

Open Source is a great idea but within the scope of a project like this, it presents formidable challenges. If you think the current development staff is "short-handed", think what is going to happen when more bodies are required for defect management (bug reporting), documentation, workload management (let's face it, someone has to keep track of what is being worked on), and quality control. I think that it would become more "hassles" than what it is worth to Baker Labs to consider this too seriously without taking all the other "burdens" into account. The project would lose cohesiveness too quickly if the work wasn't properly directed and many of the Open Source advocates would quickly become disinterested when they found that they weren't allowed to be "artists" and were relegated to the role of "contract programmers". I believe in Open Source but without the proper "command and control" schema in place, the "negatives" would far outweigh the "positives". Just my two cents for what it is worth.

B_2's picture
User offline. Last seen 4 years 25 weeks ago. Offline
Joined: 11/29/2008
Groups: None
Status: Open » Closed

Another old thread

Joined: 09/16/2010
Groups: None
Topic: General » Crash/Hang
Status: Closed » Open

Resolved? No it is not resolved.
Foldit is crashing on my ubuntu box 3 times an hour and I don't have source code to use to find fixes. I am starting to wonder why I'm still here.

B_2's picture
User offline. Last seen 4 years 25 weeks ago. Offline
Joined: 11/29/2008
Groups: None
Topic: Crash/Hang » General
Status: Open » Closed

I think your bug report more properly belongs in one of the similar Linux crash feedback items such as this one http://fold.it/portal/node/989138

This was a general discussion on open sourcing the fold.it game, and I think that has been answered, and anyone still campaigning for open source should read the post by tallguy-13088 near the end of this thread.

Sitemap

Developed by: UW Center for Game Science, UW Institute for Protein Design, Northeastern University, Vanderbilt University Meiler Lab, UC Davis
Supported by: DARPA, NSF, NIH, HHMI, Amazon, Microsoft, Adobe, RosettaCommons