Why has SETI management not officially told us that all of the work our computers are doing is completely worthless? Why do I have to dig through the message board posts to just happen to find one from Matt Lebofsky that explains that all work downloaded after 8/11 will be rejected by the server? I still have work units downloaded on 8/16 and 8/17 that are due 8/30 and 8/31! They don't have the common decency to inform their volunteers that the work our computers are doing is absolutely worthless! If you are reading this and don't know what I'm talking about, go find Matt Lebofsky's post of 8/19, then read this again. There are many more fine distributed computing projects that deserve your computer time that is now being totally wasted on SETI/BOINC! SETI management has no integrity. They have lost all TRUST. Trust is crucial, you cannot run a volunteer project without it. I personally applaud Matt's honesty, but where is the official notification? It should be plastered in bold on the main page: abandon your current work, it is worthless! When they don't have the common decency to tell volunteers that information, then they don't deserve volunteers. I encourage everyone to quit SETI/BOINC immediately and permanently. And I welcome, in fact I demand, a response to this message from SETI management. They have killed their own project. Trust is gone, all gone. You cannot recover trust when you mistreat your volunteers this badly. And that is a very sad state of affairs that they have brought upon themselves.

Good post Elmer! You have stated what I've been trying to write for quite a while and never expressed it as well as you did.

I agree 100% with you about communication problems. I remember that
thread very well being the one who've asked Matt about those units.

OK for thechnical problems it just happen although I am convince
that the techs are currently "learning" and are not good enough for
the job. Not telling us what's really going on so we don't waste our
time is another thing and very important one.

They're just doing everything wrong and nothing is up to the standard
we would expect from the brightest distributed computing project.

Actualy if I was the master of this project and I would try get rid of
the participants I would do exactly what they're doing now.

Regards
Marc

-.-. --.- -.. -..- . - .-.-. -.- --... ...--

I agree with most of what is written in Elmer's message. I pointed this out several days ago as well and also said – I had it, and will take my hands off from this SETI-project. To be able to have volunteers you must show a trustworthy project and I can’t really say that’s the case today.
I still stick to what I said earlier – SETI management should face the facts that they’re not up to it at this very moment and take down this project and return for new volunteers when they have something trustworthy to offer. It would be the most honest thing and should reinstate some kind of respect towards the project it self and all thousands of people around the world.
I do not agree on “trust can’t be recovered”! Trust can be recovered if you’re able to show some guts and admit that there is a problem and that we’re working on it, in a professional manor! I would happily close down my BIONC and restart in a couple of weeks when I know that the work my boxes producing is useful and not dumped in cyberspace…

"August 18, 2004
We are now recovering the database from a snapshot made a few days ago. This process will continue through tomorrow."


Maybe I'm not understanding English enough, but coupled with the posts of not really up before 16. and then they started copying the database, my guess is the database was reverted back to 11. August, and highly likely any wu downloaded or result uploaded/reported after 11. August is unknown to the database and therefore lost...

But if this interpretation is wrong, it wouldn't be the first time has mis-interpreted English... :)

> my guess is the database was reverted back to 11. August, and highly likely any wu downloaded or result uploaded/reported after 11. August is unknown to the
database and therefore lost...

"GUESSING" couldn't we avoid doing that by reading CLEAR and RESPONSABLE
message from the project ?

According to the site (which IMHO has an official statement that they've had problems, and that data has been lost), they have not yet restarted the project.

Under the circumstances, I find the work units encouraging -- it suggests that for once things are going better than hoped.

What we're seeing is a project where the folks involved thought that everything was ready, and that things would scale nicely, and it didn't go well

But the only real way to find out is to try.

This is not a big budget science project -- in fact, it seems to run on a shoestring, and live largely by donations of our computing resources, servers given by Sun and SNAP and others.

Someone else complained that this isn't up to "commercial quality" and I agree, because it's not a commercial project. It's research. It's academic.

Part of the experiment is to see how much can be done on virtually no money.

So, the validators aren't running, they've found problems with the sheer number of files in a directory, and the researchers took most of the weekend off because they have lives.

In the meantime, one of my machines is waiting for work units, and the other is crunching away happily.

They have also had what seems to be a fair amount of bad luck as far as hardware goes. First there was the bad stick of RAM that prevented them from bringing an additional transitioner (?) online. Then of course the RAID failure recently which has precipitated the current situation. There have been other such things that we know of (a couple of network failures and such) which undoubtedly pulled some of the devs from their bug fixing for a while. And of course considering the level of communication, I'm sure there have been others that we are unaware of. It could be the fact that they are running on a very tight budget and are therefore buying their supplies from slightly less dependable sources. Or maybe just "bad luck" whatever that is :)

Also keep in mind the difference between BOINC and seti@home. Most (not all) of the problems have been with the seti@home part of the project, not the BOINC platform. If you are losing trust/confidence, please do so in seti@home and not BOINC. Keep in mind that we have TOO MUCH CPU power for them to use. They *WANT* us to move to other BOINC-based projects. That was the entire goal of developing this thing in the first place. Right now of course there is only predictor@home to choose from and they may or may not be accepting new users. Keep in mind that ClimatePrediction.net will be starting up its public project in a few days. Others will undoubtedly follow. I really don't care if you quit seti. Check out some other BOINC things. That is why I'm developing my cross-project stats website.

And this is what happens when I'm bored at work... I go posting long messages all over the place :)



-------------------------------------------
- A member of The Knights Who Say NI!
Possibly the best stats site in the universe:
http://boinc-kwsn.no-ip.info

>
> "GUESSING" couldn't we avoid doing that by reading CLEAR and RESPONSABLE
> message from the project ?
>
>

Well, my interpretation of "snapshot" is based on the use in the transition-plan, there they took a "snapshot" of "classic" 14. May, and any later changes in "classic" is unknown to the database...

"recovering the database from a snapshot" is therefore for me interpreted the same way, anything that has happened after the "snapshot" was taken and before the system was up again is unknown for the recovered database, and is therefore lost.
"a few days" isn't so exact and can be mis-interpreted, but the later messages gives more exact time-line.

My "guess" was, since the system was down from 11. to 16, that they reverted back to 11. August. They mentions 13-18, but since the system wasn't up the 12th there really isn't a big difference here.


But going by this thread, clearly some either stopped reading after the 1st news-post, or has a completely different interpretation of "recovering the database from a snapshot".
For the 1st group, there isn't any help, for the 2nd group, it's not certain a re-wording wouldn't confuse someone else so still not all would understand the message...

Lastly, it can be me getting confused by the English so it's really something else they're meaning with the message. :)

Now! this is what I call a well thought out and spot on post, pay special attention to the second paragraph, and then go read these;

Woody wanted more.
Documentation Menu
Release Notes
BOINC FAQ
BOINC Owner's Manual
SETI@Home Web Site Owner's Manual
SETI FAQ
BOINC Stats Page



> They have also had what seems to be a fair amount of bad luck as far as
> hardware goes. First there was the bad stick of RAM that prevented them from
> bringing an additional transitioner (?) online. Then of course the RAID
> failure recently which has precipitated the current situation. There have
> been other such things that we know of (a couple of network failures and such)
> which undoubtedly pulled some of the devs from their bug fixing for a while.
> And of course considering the level of communication, I'm sure there have been
> others that we are unaware of. It could be the fact that they are running on
> a very tight budget and are therefore buying their supplies from slightly less
> dependable sources. Or maybe just "bad luck" whatever that is :)
>
> Also keep in mind the difference between BOINC and seti@home. Most (not all)
> of the problems have been with the seti@home part of the project, not the
> BOINC platform. If you are losing trust/confidence, please do so in seti@home
> and not BOINC. Keep in mind that we have TOO MUCH CPU power for them to use.
> They *WANT* us to move to other BOINC-based projects. That was the entire
> goal of developing this thing in the first place. Right now of course there
> is only predictor@home to choose from and they may or may not be accepting new
> users. Keep in mind that ClimatePrediction.net will be starting up its public
> project in a few days. Others will undoubtedly follow. I really don't care
> if you quit seti. Check out some other BOINC things. That is why I'm
> developing my cross-project stats website.
>
> And this is what happens when I'm bored at work... I go posting long messages
> all over the place :)
>


> -------------------------------------------
> - A member of The Knights Who Say NI!
> Possibly the best stats site in the universe:
> http://boinc-kwsn.no-ip.info
>

This 'SPACE' for Rent.

M7

> Also keep in mind the difference between BOINC and seti@home. Most (not all)
> of the problems have been with the seti@home part of the project, not the
> BOINC platform.

I don't think we know quite enough to say that -- what actually belongs to BOINC and what actually belongs to SETI@HOME.

If the stats are right, 92% of BOINC users are SETI users.

Predictor has 5,500 users and CP has 516.

What we do know is what the project has said. They'll start their work-week in about 12 hours from now.

To prevent any further confusion who really is giving out "offical" information on these boards, as opposed to some self appointed insider or outsider; I have the following proposal. Dr. Anderson should post a message on the front page under the news column who at the University of California, Berkeley, Space Sciences Laboratory (SSL) has the authority to speak for him and the above named organization. (This way some hack can't create a false account and make an unauthorized posting on the boards under his name)

Please give all the participants of this project a list of names who can post "officially" on these boards and therefore speak with authority in your name, that of SSL, and the Seti/Boinc project.

Your response would be greatly appreciated and thank you in advance for your time.

---

Account frozen...

> I don't think we know quite enough to say that -- what actually belongs to
> BOINC and what actually belongs to SETI@HOME.
>
> If the stats are right, 92% of BOINC users are SETI users.
>
> Predictor has 5,500 users and CP has 516.

The number of users has nothing to do with what is BOINC vs. seti. I don't think you were saying this but I'm not sure why you stuck in that particular piece of information. Much of the problems have been hardware related which is specific to the seti@home project. Also, anything that deals specifically with seti@home (such as the splitters) has nothing to do with the BOINC platform. Lastly, I point to predictor and climate prediction. While seti has been down for the past 2+ weeks, they have been up and running for much of that time. This clearly indicates that BOINC itself is working but seti@home is having problems with its implementation on their equipment.

It really is too bad that the premature kickoff of BOINC is going to leave a bad taste in many people's mouths. I'm just trying to slip in a spoon full of sugar to make the medicine go down :) Unfortunately there seem to be many around here who LIKE the taste of the medicine and shunn the sugar. Oh well... I will keep crunching for SOME project no matter what goes down here.



-------------------------------------------
- A member of The Knights Who Say NI!
Possibly the best stats site in the universe:
http://boinc-kwsn.no-ip.info

> Why has SETI management not officially told us that all of the work our
> computers are doing is completely worthless? Why do I have to dig through the
> message board posts to just happen to find one from Matt Lebofsky that
> explains that all work downloaded after 8/11 will be rejected by the server?
> I still have work units downloaded on 8/16 and 8/17 that are due 8/30 and
> 8/31! They don't have the common decency to inform their volunteers that the
> work our computers are doing is absolutely worthless! If you are reading this
> and don't know what I'm talking about, go find Matt Lebofsky's post of 8/19,
> then read this again. There are many more fine distributed computing projects
> that deserve your computer time that is now being totally wasted on
> SETI/BOINC! SETI management has no integrity. They have lost all TRUST.
> Trust is crucial, you cannot run a volunteer project without it. I personally
> applaud Matt's honesty, but where is the official notification? It should be
> plastered in bold on the main page: abandon your current work, it is
> worthless! When they don't have the common decency to tell volunteers that
> information, then they don't deserve volunteers. I encourage everyone to quit
> SETI/BOINC immediately and permanently. And I welcome, in fact I demand, a
> response to this message from SETI management. They have killed their own
> project. Trust is gone, all gone. You cannot recover trust when you mistreat
> your volunteers this badly. And that is a very sad state of affairs that they
> have brought upon themselves.
>

I agree 100% with this post! When the data in a system is corrupt . . . users tend not to trust the system any longer and that usually is the beginning of the end. You can mess about with software, but you can not mess about with data!

>
Hydnum Repandum </br>
Open SETI Initiative

I needed to point out some stuff.

A snapshot as used by Matt in one of his posts actually refers to a backup technique whereby a virtual copy of the filesystem is taken and added on too for each update since the beginning of the snapshot until the backup to tape or disk is completed.

Basically at scheduled intervals the snapshot process is started, during this process the project is shutdown the process is started against the database, upload, and download directories, and then the project is started back up. Without this technology the project would have to be shutdown for 8-12 hours instead of the 30 minute shutdown that happens in the morning.

----- Rom
BOINC Development Team, U.C. Berkeley
[url=http://www.boinc.dk/index.php?page=user_statistics&project=sah&userid=85465]<img border="0" height="70">

Any idea when the validator will be back on? I think a lot of the feeling here is that it's like driving a car with a broken odometer.

Actually I have a mixed feeling. I'm still confident it will evolve in a good direction and in a good way. When I read some posts I'm quite sure this project is still running better then the starting of S@H Classic although I wasn't there at that time.

Running the software itself has not been a problem since the official launch using broadband without proxy on a PC running Win XP.
I'm quite sure the DEV team is still working to improve the software and to remove some minor bugs that are remaining.

The real problems are what I believe linked to missing hardware, what does for me means lack of financial funds, something that probably is quite common for University projects, at less here in Belgium.

Problems of database structure are maybe actually a week point that should be improved for sure. Lack of WU's is another problem that will hopefully be past history in the near future.

The only remark I have is the lack of some real information about what is going on and the lack of fresh news on the front page.

To the team of Berkeley: keep on the good work!
To the DEV team: keep on the good hunting for improving the software!
To the crunchers: keep confident in this project. It is the future of science!

Greetings from Belgium.

Any idea when the validator will be back on?
===========

When I decide to Turn it On ... ;)

I do have pretty much the same idea of the whole situation :

Probably the ultimate bottleneck is for 80% or more, the underlying hardware or the lack thereof due to the immense costs and the relative success (wealth of crunchers).
We've seen this happen with a lot of promising scientific projects at universities... Most of the troubles we did have the last weeks are most likely due to the hardware infrastructure, complicated by the transition of going from classic seti to BOINC.

Nevertheless don't worry, be happy...

Greetings from Belgium ;-))

> Probably the ultimate bottleneck is for 80% or more, the underlying hardware

Don't forget about the SNAP story... Hardware was there. They just blew it !