Guys,

What is it with your group and the lack of any form of communication?

You have twitter and facebook accounts you don't update. You have this forum to communicate, you have the recent news section on the home page to communicate, but you do NOT use any of them.

What is the problem? Twitter, a few seconds to send out an alert that there is a problem. Facebook a few minutes, same with the forum. I have brought this point up before, it appears it is time to bring it up again.

I'm glad you started this thread because I've been giving a lot of thought to this topic and wondered whether and where I should write anything.

It appears the project has no plan in place for communicating with its volunteers. The communication, when it comes, seems haphazard, as if it is done by individuals who just happen to notice a possible need to respond. Occasionally a post has the tone of the chastened, as if the grad student, in making an error that led to the trashing of a whole batch of workunits, has been reminded by a supervisor that his carelessness led to the wasting of resources far greater than his own time. Likewise, the posts in the science section seem to come in batches, as if, perhaps in a meeting, Dr. Baker or some other supervisor had suggested the junior scientists should let us volunteers know what they were working on. It suggests to me that while Dr. Baker and others are very much aware of and value the contributions of volunteers they have delegated most of the communication work to others with vague instructions that inevitably lead to infrequent, often misplaced communications. (I note that Dr. Baker does post about upcoming publications).

For example, Rocco posted a clear, indeed excellent, explanation of the recent workunit shortage. Unfortunately it was dropped deep in the "Problems" thread after several days of anxious posts and complaints (scattered throughout the fora) by volunteers. Rocco mentions receiving an email notice about the empty queue. That notice, or whatever triggered that notice, should have triggered another notice to be placed on the front page of the website. Someone, someone specific, should have been responsible for letting the volunteers know about an event (or nonevent as it were) that effected them. This requires a plan for what sorts of information needs to be shared*, in what form and where the information will be conveyed**, and who is responsible for getting the information out***.

Another example: I would suggest every grad student and post-doc be required to make a post in the science section describing their work. Those posts are always well received and in addition to a description of the science they often provide important clues to the behaviour we volunteers observe regarding specific types of workunits.****


Best,
Snags

* I have quite a bit more to say on this topic so I'll expand on this bit in a subsequent post.
** a few thoughts to come
*** Just that it should be specific people for specific information rather than "Somebody needs to..." or "One of you guys should..."
**** Yep, more thoughts, saved for later.

I believe someone named Keith Langston or something like that is or was one of the cheif IT guys for the project. However I have not seen his name come up in a long time.

I notice that Modsense is not around any more. He used to report these things to the project.

There used to be updates by the grad students about their tasks, now there is none.

I understand that Dr. Baker is very busy as head of the department and project and can not really deal directly with things such as server outages and lack of tasks and the such. But the IT team should deal with this. There was some talk about sending out updates via twitter, but that died long ago.

I once suggested they borrow a student from the communications department to come and post to the web the news they have about new tasks and outages and the such. But that just got backburnered.

This project seems more about turning and burning the tasks rather than sharing information about the tasks or letting people know about problems with them, etc.

There is more information coming from POEM and Einstein and Milkyway than there is from here.

So come on team, start talking to us, or maybe we should just go away since you just ignore us anyway.

Ready to send 1

No communication.

Here we go again.

I am around, but have no facts to offer as to what is going on. I've been "heads down" at work the past week and not had time to keep up with the boards.

Personal observation of last week's tasks is that they use excessive amounts of memory. I saw a few over 600MB. Enough that the default BOINC settings were only running 3 on my 4 CPU machine due to memory. ...and so I'm guessing they've decided that more coding is needed on that type of work before they send out more of them. That's what I would probably do in the same position.

---

Rosetta Moderator: Mod.Sense

I am around, but have no facts to offer as to what is going on. I've been "heads down" at work the past week and not had time to keep up with the boards.

Personal observation of last week's tasks is that they use excessive amounts of memory. I saw a few over 600MB. Enough that the default BOINC settings were only running 3 on my 4 CPU machine due to memory. ...and so I'm guessing they've decided that more coding is needed on that type of work before they send out more of them. That's what I would probably do in the same position.

Thanks for the educated guess

And the silence is deafening!!
No communication, no explanation as to what went wrong and what is being done to fix it.

Getting annoyed with teams lack of respect for the volunteers.

I'd prefer more communications too.

It seems that Rosetta@Home is trying to outdo the average German BOINC project in lack of communications (at least in English).

I'm curious about the experiences some of you have had with other projects, do they communicate better than Rosetta? I've only had experience with a couple other projects, and frankly Rosetta's communication is pretty good, comparatively speaking.

Speaking for myself, I don't particularly care how much they communicate with me. I enjoy it when they post info to us, particularly on the science end of things. But I don't feel like I'm "A full partner" in anything, or that I'm owed anything. When they send WU's to me, I'm happy to run them. When they don't, I'll run WU's from my backup project, or just not run anything for a while. The way I look at it, my computer is available to them when they need it. That's it.

I'm curious about the experiences some of you have had with other projects, do they communicate better than Rosetta? I've only had experience with a couple other projects, and frankly Rosetta's communication is pretty good, comparatively speaking

After R@H had it's server meltdown over the change of the year, I added World Community Grid projects, first to cover only the downtime of R@H. Have since added it to all my hosts and run it on a few exclusively.
When they know that there is a batch of WUs ending or a whole project is coming to an end, they post it on the projects forum days if not a couple of weeks in advance, so there is no doubt about what is going on.
Didn't really have a server outage that I am aware of, but then their sysadmins might be a bit 'closer to the ball' than for example those of R@H...

Ralf

I'm curious about the experiences some of you have had with other projects, do they communicate better than Rosetta? I've only had experience with a couple other projects, and frankly Rosetta's communication is pretty good, comparatively speaking.

Speaking for myself, I don't particularly care how much they communicate with me. I enjoy it when they post info to us, particularly on the science end of things. But I don't feel like I'm "A full partner" in anything, or that I'm owed anything. When they send WU's to me, I'm happy to run them. When they don't, I'll run WU's from my backup project, or just not run anything for a while. The way I look at it, my computer is available to them when they need it. That's it.

From Poem@home: Aug 02 , 2011 - Workunit Shortage
From Einstein@home: Power outage in Hannover brought BRP servers down 14 Jul 2011 5:56:39 UTC Followed by: AEI Hannover servers up again 14 Jul 2011 12:56:26 UTC
From Milkyway@home:
Recent Outage
Milkyway@home was down for a few hours yesterday. Also news about their AC problems and workunits having a maximum time limit bug in them.

This blows away Rosetta feeble communications. So if the above 3 projects can pop out one liners or longer about their problems what is the issue with the Rosetta team and their deafening silence?

Forgot the a team member posted something in another thread which got buried:
https://boinc.bakerlab.org/rosetta/forum_thread.php?id=5701&nowrap=true#70834

Guys,
What is it with your group and the lack of any form of communication?

Same problem on Ralph@home...
We give you our cpu time AND our time to feedback problems on forum.
Why you cannot communicate with us??

---

Overall, the information Rosetta@home provides about their project far surpasses that of many other projects. In the early years the project and volunteers clearly spent a good deal of effort compiling information useful to and of interest to the crunchers and putting it on webpages and creating several long message board threads of FAQs. More recently a suggestion that publications be announced on the home page finally caught the eye of Dr. Baker and this is now being done. (The publications page was also finally updated and hopefully there is a regular procedure in place to prevent it from falling out of date again).

Likewise other events such as running out of workunits should be placed in the news section in a timely manner. Rocco's post was excellent but, as Greg notes, even a quick one liner would have alleviated the anxiety of many crunchers. For news, the when and the where it's conveyed makes a difference.



There is another category of information that we are not getting enough of lately and that's to do with how different types of workunits behave. A couple of years ago if someone posted on the boards questioning whether the behaviour they were seeing was normal or not I was fairly confident in my judgement on the matter. I knew how rosetta workunits behaved and could even explain (well, to myself at least) how and why that behavoiur was different from that of workunits from other projects. Over the years rosetta has added new strategies which results in new normal behaviours: the 100 model limit, the "jumpiness" of protein interface workunits. There are currently questions about the flxdsgn units (beginning here), their apparent 1 model limit, the odd 1201 cpu time for all (seemingly frequent) invalid workunits. There is speculation of a bug and determination to abort similar workunits but no response from the project. As I stated in that thread you ask us to report problems and we are clearly willing to do so but we need you to respond and provide us with up-to-date information on what's normal behaviour.


Ideally, there is so much more to this (identifying normal vs. abnormal behaviour) than simply typing a few responses on the message boards. The information needs to be gathered together in one easily referenced location. I believe if I am going to complain about this I should also offer some suggestions for improvement but what immediately comes to mind seems so time consuming and labor intensive that for the moment I'll leave with this: Mod.Sense once mentioned updating the FAQs. I am volunteering my services, whatever small value they may have. I could barely be considered computer literate, take far too long using far too many words to say far too little, can both be narrowly obsessed and easily lead astray. That said, should you decide I might be of some assistance, I hereby offer several hours of my time each week to update the FAQs or otherwise improve the information made available to the volunteers.



Best,
Snags

BOINC manager has a place for notices. Why is it not used by Rosetta?

If you run out of WU, say so. Nothing wrong with that. But keeping the crunchers in the dark leads to thrashing and ultimately you lose your team out of frustration.

People spend time trying to figure out what is wrong with their machine when there is nothing wrong. The project simply has no work today. Would it be so hard to say that?

I have learned, over the years, when working wtih volunteers that if you don't keep them busy then you better tell them what is going on or they leave feeling that they are not needed.

And, perhaps that is the case. Rosetta really does not need us. It seems they often have no work for us. So better to go elsewhere.

Personally I have two projects running, Seti which always seems to have work, and Rosetta which seems to have work on and off. When Rosetta is running it gets one of my two cores. When it has no work, Seti gets both cores. That's OK with me but I don't spend any time trying to figure out why I don't have a rosetta WU unless it goes on for days.

Net Net, it only takes a very small amount of communication to make things run smoothly. But it seems the Rosetta team feels none is better. I don't agree.

BOINC manager has a place for notices. Why is it not used by Rosetta?

Fair comment - no idea.

It seems they often have no work for us. So better to go elsewhere.

I don't agree with that though - I've been running Rosetta for around 6 years and I can probably count the periods where the work has run dry for more than a weekend on one hand.

---

BOINC manager has a place for notices. Why is it not used by Rosetta?

Fair comment - no idea.

The notices tab is a very recent addition to the BOINC manager. The messages tab, now named the event log and put in the "advanced" menu, used to be the place where a project could put a notice. Few projects used this mechanism. Maybe it will be used more now.

yeah thats a good idea, but then they would have to create a program for putting an alert out to boinc manager that there is no more work. that would require effort which they could put into leaving a note on the home page as well.

Thing is while it is rare for R@H to not have work ready to go, there are times this happens. When it is just a day, fine. When it is 3 days going on a week and no one says a blanking thing anywhere, you have to wonder who's hiding in fear.

You will notice no one has replied to the two threads going on about the lack of communication. What chance do you think they would use a automated program or a live human to say they are out of work due to whatever reason?? Slim to none following the latest example.

I see work is starting to come out again, but it would have been nice for a note of some sorts about what happened.

Greg - you can already tell when the project is out of work by looking at the server status page. I think that people are looking a "one liner" now and then about what is being done to resolve the issue and what the expectations for a return to normalcy are.

Last year at this time the project was cranking out 150+ TeraFLOPS - granted that was when SETI was having its issues but I think one of the big reasons R@H is now only producing 100 TeraFLOPS on a good day is the abhorrent lack of communication.

Greg - you can already tell when the project is out of work by looking at the server status page. I think that people are looking a "one liner" now and then about what is being done to resolve the issue and what the expectations for a return to normalcy are.

Last year at this time the project was cranking out 150+ TeraFLOPS - granted that was when SETI was having its issues but I think one of the big reasons R@H is now only producing 100 TeraFLOPS on a good day is the abhorrent lack of communication.

Should we reduce the computer resources they can use even more to help them find more time for communications?

Greg - you can already tell when the project is out of work by looking at the server status page. I think that people are looking a "one liner" now and then about what is being done to resolve the issue and what the expectations for a return to normalcy are.

Last year at this time the project was cranking out 150+ TeraFLOPS - granted that was when SETI was having its issues but I think one of the big reasons R@H is now only producing 100 TeraFLOPS on a good day is the abhorrent lack of communication.

Should we reduce the computer resources they can use even more to help them find more time for communications?

Would they even notice?