For those of us not doing SETI (I'm not atm, partly because I don't have a very suitable computer at my hand since mine all have a bit little CPU cache)- what exactly happened there? People got totally imba credits?
Also it seems that not all platforms were updated to the latest code simultaneously so Linux users complained about being at a disadvantage compared to Windows users. Anyway it's frightening to see how emotional those RAC-discussions can turn if you change anything in the way credits are awarded.
BTW, there's a "Tag der offenen Tür" event at the Albert Einstein Institute (and other institutes at that site) in Golm near Potsdam today. Should be interesting, even tho there's no dedicated "Einstein@Home" event.
Yes, I've spent most of August on vacation in Norway, visiting some friends. It was really great :-D but now I'm really motivated for all the tech stuff again.
Quote:
Also it seems that not all platforms were updated to the latest code simultaneously so Linux users complained about being at a disadvantage compared to Windows users.
Somehow, this seems familiar, only that over here it was more or less the other way round. Things like this happen, I think much of the problem is caused by people getting all emotional about which Operating System is "the absolute ultimate uber-1337 choice for everything"... you know, the Windows Fanboys and Uber-Geeks... we get lots of those at Heise, too, so I hope you know what I mean and the others get an idea.
Quote:
Anyway it's frightening to see how emotional those RAC-discussions can turn if you change anything in the way credits are awarded.
FULL ACK to that. It's the same everywhere, as soon as you come up with a scoring system (however abstract it may be) there will be people who take it too seriously and get very emotional about it. Back when I was at school and learning about the Internet I took part in a community where you could discuss Star Trek and chat with lots of other users and well... that community had a system where you got "points" for voting in other peoples' polls or writing into their guest books and what should I say- it was complete mayhem. People where directing flames at each other on a daily basis (in a tone that would make Eminem sound nice, only hindered by the fact that even the word "idiot" was censored out and had to get little dots in the middle in order to go through the filters...), accusing others of cheating or of jealousy or whatever. Some people came up with scripts that got them lots of points and part of them were banned or got points taken off (but not all, of course) while some were suspected who were probably innocent. All this, of course, between the relative elite who already had DSL connections back in 2001, since the others could never dream of getting through enough profiles to get halfway competitive scores.
That was about the craziest thing I ever saw in terms of people going out of their way for something that was, ultimately, no good for them whatsoever. So all the credit-whores running around at BOINC really can't surprise me any more, because it's the same thing really...
Quote:
BTW, there's a "Tag der offenen Tür" event at the Albert Einstein Institute (and other institutes at that site) in Golm near Potsdam today. Should be interesting, even tho there's no dedicated "Einstein@Home" event.
Sounds interesting, but that's all through Germany for me and I don't really have the time. Maybe next year.
- The run-time of the workunits will be roughly the same we had at the beginning of S5R2 (12h at max on our reference machine - sorry, don't remember the credit values from the top of my head - was it 190?). We hope to reduce it even further when we put more work in optimizing the Apps.
- The amount of data necessary for a workunit will raise to 35-55 MB (depending on the base frequency). For the lower frequency bands we will use the same data files than we did for S5R2, so we intend to switch off deletion of the S5R2 data files on your machines at some point and you don't have to download a whole completely new set of data files again, just the ones you don't already have left over from S5R2.
- We also think about modifying the scheduler so that the workunits that require smaller amount of data are given to machines with slower connections. However given our previous experiences with scheduler hacks we might not be able to do this at all or on time, given that the scheduler is a mission-critical component that can easily bring down e.g. the database.
For those of us not doing SETI (I'm not atm, partly because I don't have a very suitable computer at my hand since mine all have a bit little CPU cache)- what exactly happened there? People got totally imba credits?
Also it seems that not all platforms were updated to the latest code simultaneously so Linux users complained about being at a disadvantage compared to Windows users. Anyway it's frightening to see how emotional those RAC-discussions can turn if you change anything in the way credits are awarded.
BTW, there's a "Tag der offenen Tür" event at the Albert Einstein Institute (and other institutes at that site) in Golm near Potsdam today. Should be interesting, even tho there's no dedicated "Einstein@Home" event.
CU
H-BE
Yeah, there have been a few rollout problems for MB, which now that Eric is back from vacation will all get addressed in due course.
Hmmm... I seem to remember something kind of similar to that somewhere...
Oh Yeah, right here with S5R2! :-)
Unfortunately, a lot of people over there seem to forget BOINC itself and everything that goes along with it is as much an experiment as the science we run on it for the projects. ;-)
The good thing about BOINC is that if one project is having "issues" you can just switch to another. I had to abandon SETI because my MACs just not are working over there right now. In the future, I'm sure that will change. I figure there is no sense in generating junk over at SETI. There are plenty of windows boxes to keep SETI happy.
As to S5R3, I guess participants who installed beta apps in the past will want to check now that they have either the most recent beta app installed or the latest official app.
If you have an outdated beta app with a app_info.xml file installed, the automatic app software update mechanism is suspended and you won't get a new app. But any outdated beta app will not have the extra feature Bernd mentioned that is used in S5R3 so you won't get S5R3 tasks once the new run begins, which will happen rather sooner than later as S5R2 is 95% complete now.
As to S5R3, I guess participants who installed beta apps in the past will want to check now that they have either the most recent beta app installed or the latest official app.
If you have an outdated beta app with a app_info.xml file installed, the automatic app software update mechanism is suspended and you won't get a new app.
CU
H-BE
Thank you, that is certainly good to know! I have installed the latest beta (v. 4.30) and based upon its apparent lack of problems, I was wondering if Bernd might soon make it the official application even before S5R2 is completed! :)
...once the new run begins, which will happen rather sooner than later as S5R2 is 95% complete now.
CU
H-BE
I wonder if the current end-of-run "monster units" will extend the S5R2 time-to-completion beyond the 5% apparent time remaining. If the percentage complete is measured in units completed, then these longer units will extend the conclusion of the current S5R2 run beyond what one might normally expect.
However, that in no way invalidates your great post, it just might delay the start of S5R3 for a bit longer. :)
...once the new run begins, which will happen rather sooner than later as S5R2 is 95% complete now.
CU
H-BE
I wonder if the current end-of-run "monster units" will extend the S5R2 time-to-completion beyond the 5% apparent time remaining. If the percentage complete is measured in units completed, then these longer units will extend the conclusion of the current S5R2 run beyond what one might normally expect.
However, that in no way invalidates your great post, it just might delay the start of S5R3 for a bit longer. :)
Hi!
There will be a period of time when all remaining S5R2 tasks have been sent out to crunchers. Then the crunchers have (for the monsters) 3 weeks deadline to finish, and it would not be cool to wait with distribution of S5R3 tasks until the last client has submitted it's S5R2 result, so my understanding is that S5R3 will begin before all monsters have been crunched (there will be a small overlap between S5R2 and S5R3). Bernd will be able to explain this much better.
The plan is that at some time at the end of this week we'll force the S5R2 workunits generator to write all remaining S5R2 workunits into the database (which will make it stop), and then start the workunit generator for S5R3.
The effect of this would be that for some days there will be both S5R2 and S5R3 workunits in the database of which no task has even been sent to a host yet. The transition will be rather smooth and long.
Hi Annika, long time no see.
)
Hi Annika, long time no see.
Also it seems that not all platforms were updated to the latest code simultaneously so Linux users complained about being at a disadvantage compared to Windows users. Anyway it's frightening to see how emotional those RAC-discussions can turn if you change anything in the way credits are awarded.
BTW, there's a "Tag der offenen Tür" event at the Albert Einstein Institute (and other institutes at that site) in Golm near Potsdam today. Should be interesting, even tho there's no dedicated "Einstein@Home" event.
CU
H-BE
RE: Hi Annika, long time no
)
Yes, I've spent most of August on vacation in Norway, visiting some friends. It was really great :-D but now I'm really motivated for all the tech stuff again.
Somehow, this seems familiar, only that over here it was more or less the other way round. Things like this happen, I think much of the problem is caused by people getting all emotional about which Operating System is "the absolute ultimate uber-1337 choice for everything"... you know, the Windows Fanboys and Uber-Geeks... we get lots of those at Heise, too, so I hope you know what I mean and the others get an idea.
FULL ACK to that. It's the same everywhere, as soon as you come up with a scoring system (however abstract it may be) there will be people who take it too seriously and get very emotional about it. Back when I was at school and learning about the Internet I took part in a community where you could discuss Star Trek and chat with lots of other users and well... that community had a system where you got "points" for voting in other peoples' polls or writing into their guest books and what should I say- it was complete mayhem. People where directing flames at each other on a daily basis (in a tone that would make Eminem sound nice, only hindered by the fact that even the word "idiot" was censored out and had to get little dots in the middle in order to go through the filters...), accusing others of cheating or of jealousy or whatever. Some people came up with scripts that got them lots of points and part of them were banned or got points taken off (but not all, of course) while some were suspected who were probably innocent. All this, of course, between the relative elite who already had DSL connections back in 2001, since the others could never dream of getting through enough profiles to get halfway competitive scores.
That was about the craziest thing I ever saw in terms of people going out of their way for something that was, ultimately, no good for them whatsoever. So all the credit-whores running around at BOINC really can't surprise me any more, because it's the same thing really...
Sounds interesting, but that's all through Germany for me and I don't really have the time. Maybe next year.
Greetz, Annika
Some update on S5R3: - The
)
Some update on S5R3:
- The run-time of the workunits will be roughly the same we had at the beginning of S5R2 (12h at max on our reference machine - sorry, don't remember the credit values from the top of my head - was it 190?). We hope to reduce it even further when we put more work in optimizing the Apps.
- The amount of data necessary for a workunit will raise to 35-55 MB (depending on the base frequency). For the lower frequency bands we will use the same data files than we did for S5R2, so we intend to switch off deletion of the S5R2 data files on your machines at some point and you don't have to download a whole completely new set of data files again, just the ones you don't already have left over from S5R2.
- We also think about modifying the scheduler so that the workunits that require smaller amount of data are given to machines with slower connections. However given our previous experiences with scheduler hacks we might not be able to do this at all or on time, given that the scheduler is a mission-critical component that can easily bring down e.g. the database.
BM
BM
RE: Hi Annika, long time no
)
Yeah, there have been a few rollout problems for MB, which now that Eric is back from vacation will all get addressed in due course.
Hmmm... I seem to remember something kind of similar to that somewhere...
Oh Yeah, right here with S5R2! :-)
Unfortunately, a lot of people over there seem to forget BOINC itself and everything that goes along with it is as much an experiment as the science we run on it for the projects. ;-)
Alinator
The good thing about BOINC is
)
The good thing about BOINC is that if one project is having "issues" you can just switch to another. I had to abandon SETI because my MACs just not are working over there right now. In the future, I'm sure that will change. I figure there is no sense in generating junk over at SETI. There are plenty of windows boxes to keep SETI happy.
As to S5R3, I guess
)
As to S5R3, I guess participants who installed beta apps in the past will want to check now that they have either the most recent beta app installed or the latest official app.
If you have an outdated beta app with a app_info.xml file installed, the automatic app software update mechanism is suspended and you won't get a new app. But any outdated beta app will not have the extra feature Bernd mentioned that is used in S5R3 so you won't get S5R3 tasks once the new run begins, which will happen rather sooner than later as S5R2 is 95% complete now.
CU
H-BE
RE: As to S5R3, I guess
)
Thank you, that is certainly good to know! I have installed the latest beta (v. 4.30) and based upon its apparent lack of problems, I was wondering if Bernd might soon make it the official application even before S5R2 is completed! :)
RE: ...once the new run
)
I wonder if the current end-of-run "monster units" will extend the S5R2 time-to-completion beyond the 5% apparent time remaining. If the percentage complete is measured in units completed, then these longer units will extend the conclusion of the current S5R2 run beyond what one might normally expect.
However, that in no way invalidates your great post, it just might delay the start of S5R3 for a bit longer. :)
RE: RE: ...once the new
)
Hi!
There will be a period of time when all remaining S5R2 tasks have been sent out to crunchers. Then the crunchers have (for the monsters) 3 weeks deadline to finish, and it would not be cool to wait with distribution of S5R3 tasks until the last client has submitted it's S5R2 result, so my understanding is that S5R3 will begin before all monsters have been crunched (there will be a small overlap between S5R2 and S5R3). Bernd will be able to explain this much better.
CU
H-BE
The plan is that at some time
)
The plan is that at some time at the end of this week we'll force the S5R2 workunits generator to write all remaining S5R2 workunits into the database (which will make it stop), and then start the workunit generator for S5R3.
The effect of this would be that for some days there will be both S5R2 and S5R3 workunits in the database of which no task has even been sent to a host yet. The transition will be rather smooth and long.
BM
BM