Maybe they're debugging on their own machines or on the main Rosetta?

Which is the sense to test an app on one or two machines and distribute it to tens of thousands of different pcs?
Which is the sense to test an app to production's project if you have a test project?
Which is the sense to have a bug since August 2020 and repeat it until today?

They'd use Ralph if they needed to.

And this is clear

---

Maybe they're debugging on their own machines or on the main Rosetta?

Which is the sense to test an app on one or two machines and distribute it to tens of thousands of different pcs?
Which is the sense to test an app to production's project if you have a test project?

Because tens of thousands can test better than a few hundred on Ralph. If it's probably not going to crash or cause alarm to the user, best to test on many more machines.

Which is the sense to have a bug since August 2020 and repeat it until today?

What bug?

---

Because tens of thousands can test better than a few hundred on Ralph. If it's probably not going to crash or cause alarm to the user, best to test on many more machines.

I continue to prefer stable apps in Rosetta and test apps in Ralph. But it's only my opinion...

Which is the sense to have a bug since August 2020 and repeat it until today?

What bug?

File: ......srcprotocolsmotif_graftingmoversMotifGraftMover.cc:537
For this scaffold there are not suitable scaffold grafts within your constrains

---

Because tens of thousands can test better than a few hundred on Ralph. If it's probably not going to crash or cause alarm to the user, best to test on many more machines.

I continue to prefer stable apps in Rosetta and test apps in Ralph. But it's only my opinion...

It doesn't really matter where they're run, but I agree it's better to put the unstable ones in the beta where the users don't care so much. But if they only seldom crash, does it matter? I can't remember the last broken Rosetta task.

What bug?

File: ......srcprotocolsmotif_graftingmoversMotifGraftMover.cc:537
For this scaffold there are not suitable scaffold grafts within your constrains

Did that produce a computation error or did you just happen to notice it in your results? Because all mine appear to work ok.

---

Did that produce a computation error or did you just happen to notice it in your results? Because all mine appear to work ok.

There are, periodically, problems like these

---

Did that produce a computation error or did you just happen to notice it in your results? Because all mine appear to work ok.

There are, periodically, problems like these

Not sure why that would be a problem, especially if they crash very quickly. You've only wasted a few seconds of CPU time. And you've helped them find a bug.

---

And you've helped them find a bug.

That's what Ralph is meant to be for, not the main project.

---

Grant
Darwin NT

And you've helped them find a bug.

That's what Ralph is meant to be for, not the main project.

Does it really matter? Somebody somewhere needs to run the faulty task. I would agree with you only if the faulty task could crash the machine or something, so only the diehard users get them.

And why don't they do what other projects do, have a tickbox for "run beta tasks". Then you don't need two projects.

---

Did that produce a computation error or did you just happen to notice it in your results? Because all mine appear to work ok.

There are, periodically, problems like these

Not sure why that would be a problem, especially if they crash very quickly. You've only wasted a few seconds of CPU time. And you've helped them find a bug.

The "ea00010f_nav1-7" tasks that Grant found, and I also get, just run a few seconds. Yes, not a problem.

The XW_JG tasks run for several hours before crashing, but do award credits for time run.

The others, with an Error reported, crash after several hours and award credit too - hopefully the error message is found and helps.

I don't know if these are errors peculiar to my PC or if many people get them.
If me, it's my problem.
If many get the same, hopefully something gets re-coded (they all seem to be bcov's jobs here at Rosetta).

But in general, I agree. They ought to be run first at Ralph (which I don't run), but even if they get through to us here the errored tasks can be useful too.

It used to be that credits weren't awarded at all for crashed tasks, but now that they are, there's no loss to me that concerns me.
I list them when I get the time just to highlight it to others - and so others might advise me if it's likely they're a problem specific to my end.

Edit: I note no errors on my other (i3) PC (running offsite), nor my laptop, nor phone, so there's a distinct possibility it may be more related to my PC than a task/coding error.

---

And why don't they do what other projects do, have a tickbox for "run beta tasks". Then you don't need two projects.

Beta/test project is not only for new apps, but also to test infrastucture update/new server/etc

---

I would agree with you only if the faulty task could crash the machine or something, so only the diehard users get them.

This has also happened in the past...

---

And why don't they do what other projects do, have a tickbox for "run beta tasks". Then you don't need two projects.

Beta/test project is not only for new apps, but also to test infrastucture update/new server/etc

The server status page says Rosetta's servers run on Ubuntu 16.04, which will see updates end in April.
Unless they pay for the extended support, I'm guessing we might see a bit of downtime somewhat soon as they update to a more recent Ubuntu version.

---

And why don't they do what other projects do, have a tickbox for "run beta tasks". Then you don't need two projects.

Beta/test project is not only for new apps, but also to test infrastucture update/new server/etc

The server status page says Rosetta's servers run on Ubuntu 16.04, which will see updates end in April.
Unless they pay for the extended support, I'm guessing we might see a bit of downtime somewhat soon as they update to a more recent Ubuntu version.

Surely it can run without updates? It's like those people who panic when Microsoft ends "support" for an older OS. What that actually means is you can now use your PC in peace and quiet without forced updates twice a week!

---

Sure it could, but it's a server that will get no more security updates while running out of a University.
I don't think they would risk it.


I myself like to run Windows XP every now and then for nostalgia purposes lol

---

Sure it could, but it's a server that will get no more security updates while running out of a University.
I don't think they would risk it.

The latest version of Ubuntu Server LTS is 20.04.2, released 2 days ago.

---

My point is that Rosetta@home is running on 16.04 which will no longer receive security updates from April onwards, as I stated on my original post.

I don't think they'll want to remain running on 16.04 but that's up to them, I guess.

---

Sure it could, but it's a server that will get no more security updates while running out of a University.
I don't think they would risk it.

The University I worked at had no security rules whatsoever. You got as many fixed or dynamic IPs you wanted to do whatever you liked with, and there was no firewall etc, you were on the raw internet. I er.... may have had a server doing things they didn't know about. It ran for a few years before they grumbled about high bandwidth!

I myself like to run Windows XP every now and then for nostalgia purposes lol

Trying to use XP makes me want to punch the screen. How did we ever put up with that?

---

My point is that Rosetta@home is running on 16.04 which will no longer receive security updates from April onwards, as I stated on my original post.

I don't think they'll want to remain running on 16.04 but that's up to them, I guess.

Depends if it screws up the program and if it's easy to fix. The most important thing is that the science works.

But wasn't there some fussy company who insisted on https before they'd join? And that caused a lot of bother!!

---

i still using Windows XP but with a non SSE cpu ( XP2800+) it's hard to stay to chrome 33 or firefox 48 ^^ and i don't talk about SHA-2 algorithm ...

But wasn't there some fussy company who insisted on https before they'd join? And that caused a lot of bother!!

If it's your data at risk you have the right to make a few requests especially if what you are doing is trying to be first to the market with the results.