Rosetta 3.11
Interesting:

Implemented a Rosetta thread manager to facilitate multithreaded protocol development and to avoid thread explosions when nested requests for multithreaded code execution are made.

Maybe in future R@H app.....

---

Cosmology @ home (for example), implements the same principle, but really 'multithreaded', with VirtualBox and boinc2docker (1 application).
I think it would be more effective to change the code for SIMD.

I think it would be more effective to change the code for SIMD.

I agree.
But, with multi-thread, we may simulate bigger proteins..

---

This new version has not only multithread support, but a lot of new features, improvements and bugfix.
Some:

- MultistageRosettaScripts
- Support for peptoid macrocycle structure prediction in the simple_cycpep_predict application.
- GALigandDock: density scoring
- New ResidueSelectors
- New SimpleMetrics
- General improvements to centralize disk use and remove repeat access, especially w/r/t scoring. This makes Rosetta more usable on ultra-high-processor-count supercomputers without disk hammering when all threads try to grab scorefunction data at once.
- Added support for linking Rosetta against Tensorflow (extras=tensorflow option during compilation) to facilitate development of machine learning-based protocols.

I hope, one day, we will have these features on R@H

---

Very interesting, but honestly I got tired of throwing electricity because of WUs, which give error after a few hours of work.
If R @ H has improved applications, I gladly give my contribution, but they must be able to be selected independently, without having to necessarily download other bad ones.

P.S. for information purposes only, AMD provides a good compiler for C/C++ : https://developer.amd.com/amd-aocc/

Very interesting, but honestly I got tired of throwing electricity because of WUs, which give error after a few hours of work.

I'm also reducing my cpus on this project, because the lack of communications...

---

I have had only one error and 122 valids. I am adding a Ryzen 3700x.

---

I have had only one error and 122 valids.

It's not only a problem of errors, it's a question of attitude.
For example: no one admin follow the thread of errors, saiing "ok, guys we are working on debug".
Or, other example, they released a new version of android app and no one wus in queue.
The latest news in home page is of 5 months ago...

---

I didn't have errors either. They are not active and doesn't share much info about the updates but according to the news and their twitter account research goes well. so as long as it helps the science I will keep crunching. I think rosetta is probably the most valuable boinc project. there are many good projects but rosetta seems to be the most effective to change our lifes in the short term.

---

I think rosetta is probably the most valuable boinc project. there are many good projects but rosetta seems to be the most effective to change our lifes in the short term.

Do not misunderstand me.
Rosetta is a great project, from scientific side.
But it seems there is little consideration of volunteers.

---

But it seems there is little consideration of volunteers.

There isn't much hand-holding for us, that is true. And their code is a mystery that only rjs5 can have any hope of understanding.
More frequent updates would be helpful; there have been a few unexplained outages too.

But I agree with kaancanbaz that it has the greatest potential to change our lives in the short term (probably medium and long too).
While they do mainly biomedical, if they could do some green energy work, that could be more transformative than the other projects specifically addressing the problem, for example.

So I will put up with the bad PR, but I agree that they could help themselves by doing more.

---

I am interested, as I am blood group 0 +. I derive many benefits from proteins.
If it had multithreading and / or SIMD oriented applications, it would be better for me.