Back in the "good ol' days" of classic SETI@home, there were server problems, but they were simple and repaired relatively quickly. Nowdays (June 2005), all we get are server problems, database problems, power problems, communications problems, and on and on and on.... As I write this, I've been waiting for three days for SAH to bless me with a work unit, and all I get is the maddening "no work from project". And forgive my bad manners, but this BOINC thing sucks canal water! Apparently, I'm not alone in feeling this way. At the demise of classic SAH, it had (I believe) some 5,000,000 users. SAH/BOINC has fewer than 125,000. Apparently, some 97% of the old users didn't bother with the switch. Frankly, had I not been doing this since 1999, I doubt I would have switched either. I'm not sure who was supposed to benefit from BOINC, but the SAH users sure as hell didn't.

---

Well there is only about 350,000 active Classic users and I was one until about a year ago when I changed to Boinc and have not looked back since. This is new and exiting and the way of the future.. Sorry you do not agree but that's the way the cookie crumbles.

I too have had problems getting Seti WUs, but I am attached to other projects, which by the way is the reason for BOINC's existence, and have never run out of work. If you are one of those that refuses to attach to more than one project you will never be happy with Boinc.

---

Semper Eadem
So long Paul, it has been a hell of a ride.

Park your ego's, fire up the computers, Science YES, Credits No.

If you take a look at this article about the golden past, where everything was fine and smooth, nothing ever went wrong and no problems occured, you would perhaps have a different view of it;)

Power and network problems affected both, Classic and Boinc, and were clearly outside the responsibilizy of the project team.
Database and server probs are IMHO unavoidable, if you have to maintain two projects with the staff of only one, on the hardware of one and a half.

I had no real problems thanks to Boinc, as some other project simply took over, and my CPU stayed warm and cozy.

BTW: The sending of up to 30 time the same WU (afaik about 6 times on average), just to please the crunchers, without any sense, is imho worse than some minor outages.
And active classic crunchers are only about 350.000, the others decided befor that it's not to their liking.

---

Gruesse vom Saenger

For questions about Boinc look in the BOINC-Wiki

Maybe you can change your "Connect to network about every" day setting to an amount that will allow you to weather these outages.

I've been getting work for the last three days but it has been tough to get through at times.

---

Bob,

It's easy to get discouraged but maybe we can help out a bit.

You have 3 computers showing. Are they all the same machine or do you really have 3 boxes? The number one machine returned 2 work units yesterday at allmost this same exact time. What messages are you seeing in the log? Have you tried a manual update?

---

Boinc....Boinc....Boinc....Boinc....

Back in the "good ol' days" of classic SETI@home, there were server problems, but they were simple and repaired relatively quickly. Nowdays (June 2005), all we get are server problems, database problems, power problems, communications problems, and on and on and on.... As I write this, I've been waiting for three days for SAH to bless me with a work unit, and all I get is the maddening "no work from project". And forgive my bad manners, but this BOINC thing sucks canal water! Apparently, I'm not alone in feeling this way. At the demise of classic SAH, it had (I believe) some 5,000,000 users. SAH/BOINC has fewer than 125,000. Apparently, some 97% of the old users didn't bother with the switch. Frankly, had I not been doing this since 1999, I doubt I would have switched either. I'm not sure who was supposed to benefit from BOINC, but the SAH users sure as hell didn't.

Bob, sorry you see it that way.
Sure there are plenty of things I don't like about BOINC and certainly there have been numerous operational issues.
That said, if you've been doing SETI since 1999, you should remember that things wern't much different back then - yet you stuck with it.
It's a miconception to say that there are 5,000,000 Classic users, there are not and never were. That figure is the total number of people who came and went over the whole 6 years. At its peak, there were I think around 350,000 users active in the same 28 day period. BOINC/SETI has I believe around 50,000 _active_ users from a total of 125,000 that have signed-up at one time or another. I don't know what the recent figures for active classic users is, but your 97% figure is just way off the mark. At an absolute minimum, 1 seventh of the classic users have switched - thre reality is more likely to be over 30%.
Also, the 50,000 active participants do not include BOINC users who do not run SETI.
As for who was supposed to benefit from BOINC, it's the Distributed Computing Community and the Science, not me or you.

I'm not attacking your frustration, I feel the same some days. I get annoyed if there looks like they'll be no SETI work, as I don't really want to run other projects (at present). That said, running single project with a 7 cache, I simply do not run out of work.

My advice is chill out a bit, take it easy and have patience. Progress is being made, sometimes (it seems to us) painfully slow, yet we don't know what else is going on behind the scenes.

---

Bob,

Geek@Play pointed out that you have three machines showing. If you don't have three actual machines you may need to merge the duplicates before they get any credit added to them.

You remind me of me when I first started BOINC. I just about had SETI classic all figured out and then BOINC came along. After a while it really starts to grow on you. It's sort of like a new pair of shoes. The possibilities for BOINC seem boundless.

---

Let's not gang up on Bob. I think he is right in large measure.
* Why is it we should be accepting of these terrible issues on almost a weekly basis? Yes, I believe classic had its startup pain, but why shouldn't its replacement actually have fewer? Maybe there is a rational answer, but on the face of it organizations need to learn from their mistakes and not accept repeating them, yada yada. (Isn't this version of seti funded by some large government grants where the initial one was not? Is that the real issue??)
* Second, this confusion about what version boinc to use seems endless. Some sort of change control should be in place so that nothing gets released to the beta population until some sort of alpha population does a good job getting the obvious issues in the open and resolved. I can't speak for the details, but there seems to be a boinc version 4.xy on a daily basis. That's ok for alpha, not ok for beta, and I hope never an issue for the larger population who just wants to find ET.
* Third, there is a fermenting anger about credits on the message board. The system is either broke, not well thought out, or something, but there are a lot of users who want some sort of merit badge but want it to be a fair one.
* Fourth, if the only issue is bugs and not some sort of ineptitude, then why are people being forced off classic right now? What's the rush? Boinc/seti is certainly not yet ready for prime time as judged by these message boards and actual experience.

That's all. I'm sure I'll get my share of hate mail, but I think somebody needs to support Bob's point of view.

---

May this Farce be with You

Let's not gang up on Bob. I think he is right in large measure.

Nobody is ganging-up - everyone is to some degree agreeing with Bob and trying to heklp him.

* Why is it we should be accepting of these terrible issues on almost a weekly basis?

Why not? How much did you pay for your copy of BOINC? I bet I got mine cheaper because it doesn't seem too bad a deal to me.
Put simply if you can't accept BOINC for what it is, uninstall it and go on your way, do something else.
BOINC is not only free, it is totally and completely unnecessary. There is no aspect of your life that is affected in any way by wether BOINC works or not. This is unlike say using a well known word processor at work, which you don't pay for (work does) that doesn't do its job - it affects your work.

Yes, I believe classic had its startup pain, but why shouldn't its replacement actually have fewer? Maybe there is a rational answer, but on the face of it organizations need to learn from their mistakes and not accept repeating them, yada yada. (Isn't this version of seti funded by some large government grants where the initial one was not? Is that the real issue??)

BOINC and Classic SETI are completely different things. Original SETI@Home was a science application, with internet communications bolted-on. BOINC is a Distributed Computing Framework, with bolt-on science apps.
If you want to compare like-with-like, compare the BOINC/SETI Science application with the Classic Science application. The BOINC/SETI science application performs flawlwssly.

* Second, this confusion about what version boinc to use seems endless. Some sort of change control should be in place so that nothing gets released to the beta population until some sort of alpha population does a good job getting the obvious issues in the open and resolved. I can't speak for the details, but there seems to be a boinc version 4.xy on a daily basis. That's ok for alpha, not ok for beta, and I hope never an issue for the larger population who just wants to find ET.

Vearsions are clearly labeled 'Recomended Version' and 'Development Version'. Those that get 'Recomended' status are those that have had alpha-test input.

* Third, there is a fermenting anger about credits on the message board. The system is either broke, not well thought out, or something, but there are a lot of users who want some sort of merit badge but want it to be a fair one.
* Fourth, if the only issue is bugs and not some sort of ineptitude, then why are people being forced off classic right now? What's the rush? Boinc/seti is certainly not yet ready for prime time as judged by these message boards and actual experience.

Can't disagree that the credit system is a problem. I havn't come up with a better alternative yet, have you?

That's all. I'm sure I'll get my share of hate mail, but I think somebody needs to support Bob's point of view.

[/quote]

Not intended to be 'hate-mail', just putting a few things in perspective, I hope.

---

That's all. I'm sure I'll get my share of hate mail, but I think somebody needs to support Bob's point of view.

Not from me -- he has valid concerns and frustrations -- I'm still current over in Seti Classic -- which has been running quite well even during the latest bout of issues here. Further, when there were connectivity issues (which affected BOINC SETI and Classic SETI) once they are resolved, Classic recovers far more rapidly than BOINC -- we're talking 3 hours versus 24 or more hours.

Also, I've heard the litany of Classic was troublesome early on -- understand that -- I've been a SETI classic member for over 4 years. But the thing here with SETI BOINC is that what we see with that litany is the thinking that BOINC started at the same 'zero' base point as Classic -- which doesn't make sense, at least some of the learning from Classic should have accrued to BOINC. I'd note also that the problems with SETI BOINC seem these days to happen quite a bit more often than over at Einstein BOINC -- so there is another variable to be considered as well.

What is frustrating to me is not only the seemingly high frequency of outages, but also the very long recovery periods. An outage of 4 hours or more seems to require a recovery period of 12 to 24 to 48 hours. Heck, for the past week or more when we've not been in 'outage' we've been in 'recovery' -- like now.

I agree that changing cache settings (best done apparently by AVOIDING newer clients and sticking with 4.19 which I've done), one can weather most of the outages. For me, the other solution has been to increasingly use EINSTEIN as my second project -- and periodically tweaking shares so Einstein gets a larger share of CPU (as a 'reward' for stability). I also, as needed, will restart classic on workstations running BOINC.

But I agree, the frustration level is quite high at the moment -- and I suspect that frustration is not only with the 120K or so SETI BOINC users, but also in the project support group.

---

I'd accept that part of the litany if we weren't seeing a more stable implementation of BOINC over in the Einstein project -- which has run surprisingly solid for the past month or more. They have a fairly comparable user base (close to 40% of SETI BOINC with the same user counting approach. (I accept that folks comparing 5 million SETI classic users to 120K SETI BOINC users are very much doing apples and oranges.). That being said, Seti classic 'costs' the same as BOINC and an argument that 'it is free, therefore we should *expect* it to be unreliable' strikes me as a tad specious. It is also interesting that even with the change in home web page (BOINC SETI replacing Classic SETI) Classic is STILL acquiring *new* users at a rate higher than BOINC SETI.

Don't get me wrong, as you can see from my stats and Seti BOINC farm, the concept of BOINC strikes me as an excellent one, and, well implemented should be preferred to SETI Classic -- especially for multiple project support. But for some reason, over the past month or so, BOINC - SETI has been quite problematic.


[/quote]

Why not? How much did you pay for your copy of BOINC? I bet I got mine cheaper because it doesn't seem too bad a deal to me.
Put simply if you can't accept BOINC for what it is, uninstall it and go on your way, do something else.
BOINC is not only free, it is totally and completely unnecessary. There is no aspect of your life that is affected in any way by wether BOINC works or not. This is unlike say using a well known word processor at work, which you don't pay for (work does) that doesn't do its job - it affects your work.

[quote]

---

I suspect that Bob is afraid to jump back in here. Doesn't matter anyway. Right now my boxes are uploading the finished work units but staying at "ready to report" status. The scheduler may be down again.

Sigh................................

Not to worry, they will get the bugs out eventually!

---

Boinc....Boinc....Boinc....Boinc....

Points to ponder
Win NT was rewritten from scratch started with 3.51 and went to nt 4.0 with 6 plus major service packs written by how many programmers?
Win 2000 was base on nt with service pack up to 4, winxp continue up to sp 2.

How much money went into the research and dev of windows NT to XP?

I think the DEV have done a great job with what they have to work with, donated servers, etc.

Just my two cents

---

Fred
BOINC Alpha, BOINC Beta, LHC Alpha, Einstein Alpha

Any one of us could take over and carry Bob's banner on this one.; we can all argue it both ways. So just to balance this out a bit more and support Bob and BarryAZ. The recovery periods sucks badly. On Friday I predicted 27 hours before it was back...I got it wrong....give it another 24 at least. I need to be careful what I say as I walked into a minefield yesterday on this. I do feel this project is available before its ready. There is much develpoment work to do yet and systems to get right. It won't happen overnight but if we want to keep the classic migrants UCB have gotta work hard and fast to bring some stability. Also we gotta see better alpha and beta activity in testing of ALL elemnts of the system; crucial!

We are where we are as they say. Stick with it Bob and give em hell like some of us do to get it right.

---

Points to ponder
Win NT was rewritten from scratch started with 3.51 and went to nt 4.0 with 6 plus major service packs written by how many programmers?
Win 2000 was base on nt with service pack up to 4, winxp continue up to sp 2.

How much money went into the research and dev of windows NT to XP?

I think the DEV have done a great job with what they have to work with, donated servers, etc.

Just my two cents

With respect...there aint no comparison between this system's complexity and an operating system - NT or whatever. Pointless comparison in my opinion. This is simple stuff compared to Windows 98 let alone NT or 2003 or whatever.

---

[/quote]

With respect...there aint no comparison between this system's complexity and an operating system - NT or whatever. Pointless comparison in my opinion. This is simple stuff compared to Windows 98 let alone NT or 2003 or whatever.[/quote]

I agree -- and, along with the position of 'it is free - have no expectations' - tends to give 'permission' for not resolving problems.

I am still curious as to what has gone on over the past month over in SETI BOINC that we are not seeing over in Einstein BOINC.

These excessive recovery periods are just that. I wonder if that is a SETI/BOINC design specific problem that needs some looking into.

---

I also wonder if sometime in the past hour or so that the schedulers went offline again....

---

Points to ponder
Win NT was rewritten from scratch started with 3.51 and went to nt 4.0 with 6 plus major service packs written by how many programmers?
Win 2000 was base on nt with service pack up to 4, winxp continue up to sp 2.

How much money went into the research and dev of windows NT to XP?

I think the DEV have done a great job with what they have to work with, donated servers, etc.

Just my two cents

Actually, WinNT was based in part on code taken directly from the Athena operating system owned by DEC. DEC was given free use of NT in perpetuity as part of the settlement. I think part of the reason v5.0 was released as Win2k and not NT5 was to avoid the obligation, but it is also possible that they may have finally purged the portions of the os they were sued for.

If Boinc/Seti could do ONE thing to alleviate the complaints- then IMO that would be to Educate the user. There have been many complaints based upon inaccurate information. I don't think blaming the user is right since Seti/Boinc hasn't put much "Official" information out there for users to easily access. They are offering access to the WIKI which is a good start.

Pauls documentation is excellent, BUT it's 700+ pages and no user really wants to have to read that much just to crunch seti. It's an excellent place to go LOOK for answers. Seti needs a short version, that won't turn off the users that are in a hurry (most every new user).

Perhaps something with the following sections:

What is Boinc-
define what boinc stands for, the fact that there are two parts(boinc and App), etc

How do I sign up-
describe how to make an account, how to attach,

Why is Boinc replacing Classic-
describe the max number of WU available, and the fact that the current user base(classic and boinc) is capable of doing all the work that arecibo can provide. Cite the fact that Classic is resending WU's up to 30 times just to make users happy, and that isn't helping science since only 2 or 3 returned WUs are needed by the scientists. Mention that Boinc was created to provide other projects access to the Super computer that we are, and thereby actually do more science with the same data base of users. etc. etc.

How is Boinc different from Classic
describe the ability to do other projects. List them with links. Describe the new scheduler. Describe briefly how the new scheduler works, then Link to a more detailed explanation of the formulas/program of how it works (LTD is 2 x connect to, etc)

anyway, you get the picture.

For example, User ksnash came here hating everything. he didn't understand WHY he hated it, so his posts were filled with inaccurate info. He stuck to it, he learned, and now his posts are atleast "Informed" complaints. He still hates it, but now he cites Valid (to his mind) points. At least he's stuck with it and learned. Most users won't be that patient, they'll just leave.

So, the users need to be "informed" in an easy to access fashion.

(2cents inserted)

tony

---

my kid always has to be reminded of chores or tasks, and I think users are not that different. Something needs to be said in many places. Place a link to the info in:

the account verification email
the account creation page
the front page
the message board page
the help/questions page