A full week of work if not more directly to trash. I knew this was going to happen One day…
“We found a bug in the scheduling code”

It’s too easy too put it on bugs every time. These are not bugs but bad system set up And mismanagement. Why not say the truth! The scheduling code has not been Written correctly. Period. Let’s face the facts. This project is going very bad. If I was a key sponsor to this I would think twice before putting my logo at the bottom of a web page that has only bad news to announce. Day after day after day.

If the problems are known then take all the necessary measures and fix it once for all. If it is not known then deal with people who has the knowledge to set it up correctly but If you can’t just shut down the project and stop posting these useless and ridiculous news messages. I don’t understand how a system witch is not a big deal after all could work so badly. There are thousands of much more complex systems that works out there. Wake up guys!
A man walked on the moon 25 years ago!!!

I am not leaving. I am not saying goodbye SETI like others might say because I want to find ET but again let’s Face the facts, BOINC IS A JOKE. It is certainly a good idea but unfortunately It’s being manage by amateur who are learning on the back of thousands of users. If SETI doesn’t have the necessary resources to run both “classic” and BOINC then shut down one of Them. Pure and simple.

I am now crunching “classic” SETI but there is something wrong there also. The WU I Am currently crunching is a year and a half old !!! I guess they already know if this WU Contain something of interest in it or not by now. Why send those WU again and again.

Oh god !!! I went to the bank this morning and there was a message on the door;
“The bank is currently close. We found a bug in the scheduling code such that all deposits
were not being updated in the database. We are now running a fix script. All the money you
have deposit is in our safe but unfortunately we don’t know to whom it belong to”

I stick with the project and I have faith but it’s just amazing how bad things are going. Last
News from main SETI site is more then a month old where they still invite users to switch
To BOINC. It’s really amazing and it doesn’t make sence.

Sorry for the long post…. I feel better now.

> The scheduling code has not been Written correctly...

Uh... well that is what having a "bug" in the code MEANS.

I find your disrespect for the developers very rude and unprofessional. Have YOU ever tried developing a multi-server web application? I think not or you would have a little more respect for what they HAVE done. And if you HAVE, then don't sit there bitching about it! The source code is available for download. Download it and help them out instead of spamming the boards with a bunch of hot air.

FYI they didn't want to go live as early as they did. The URL for the 'active' site leaked out and spread around the world through message boards. If this had not happened, they would have done more testing and possibly some load-testing of the servers. They also could have prepared with better hardware. Right now they are waiting on a major disk upgrade to increase performance in their database. The supplier they ordered it from messed up so it probably won't get there until Friday

> I am now crunching “classic” SETI but there is something wrong there also.
> The WU I Am currently crunching is a year and a half old !!! I guess they
> already know if this WU Contain something of interest in it or not by now.
> Why send those WU again and again.

Because this is what they have been doing in classic to keep the crunchers happy. They just sent work units out over and over again to keep us busy because they have too much CPU power that they can't use effectively. This is why they developed BOINC. It will only give out work units 3 times. If they took the classic approach, we wouldn't be having nearly as many problems with work unit supply. Even when the project is fully up and running, they will run out of work units for us because there just won't be enough to go around. This is why they made it possible to get work from other projects as well. If seti@home is out of work, your client will just go on to ask predictor@home or einstein@home or climate prediction.net for work.

So really, you should be happy about this since your CPU time isn't being 'wasted' on useless re-calculations.



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

> FYI they didn't want to go live as early as they did. The URL for the
> 'active' site leaked out and spread around the world through message boards.
> If this had not happened, they would have done more testing and possibly some
> load-testing of the servers. They also could have prepared with better
> hardware. Right now they are waiting on a major disk upgrade to increase
> performance in their database. The supplier they ordered it from messed up

That's not quite right Toby. They chose to take it online of their own free will. When the address was leaked they could have blocked further users from joining, they didn't. I sat and watched the discussions on the Beta boards and refrained from joining (although I was sorely tempted to) until the official announcement. When I joined my machinies got host ID's in the high 2000 numbers, so not too many had joined while it was still unofficial. They could have stemmed the tide by not allowing new users to join but instead they chose to take it live officially. I'm not even going to make a call about whether or not it was a bad decision, but just point out that it was their decision, it wasn't forced on them by the users.

> This is why they made it possible to get work from other projects as well. If
> seti@home is out of work, your client will just go on to ask predictor@home or
> einstein@home or climate prediction.net for work.
>
> So really, you should be happy about this since your CPU time isn't being
> 'wasted' on useless re-calculations.

And a great idea it is too, much better than endlessly crunching the same WU's for no real gain.

OK I have been rude. I just don't understand how this project is managed.
On one side we cruch WU for nothing and on the other nothing works properly.
If it's like many says that they are short of hardware then what are they waiting for to close "classic" and get Boinc to work ?

It also looks like they solve one problem and create a few others in the process.
I feel like this will never end. You have all notice the new header
on the web pages...

Warning: session_start(): Cannot send session cache limiter - headers already sent (output started at /disks/koloth/raid5_b/users/boincadm/projects/sah/html/inc/db.inc:20) in /disks/koloth/raid5_b/users/boincadm/projects/sah/html/inc/util.inc on line 30

This is there since they are quote "trying to find the bottlenecks now."

I know they are working very hard and I respect that but we all have our
limit of competence. I guess we could all send one dollars to the project so they could deal with a professional ebusiness firm to clean up this mess and make things work once for all.

> OK I have been rude. I just don't understand how this project is managed.
> On one side we cruch WU for nothing and on the other nothing works properly.
> If it's like many says that they are short of hardware then what are they
> waiting for to close "classic" and get Boinc to work ?
>
> It also looks like they solve one problem and create a few others in the
> process.
> I feel like this will never end. You have all notice the new header
> on the web pages...
>
> Warning: session_start(): Cannot send session cache limiter - headers already
> sent (output started at
> /disks/koloth/raid5_b/users/boincadm/projects/sah/html/inc/db.inc:20) in
> /disks/koloth/raid5_b/users/boincadm/projects/sah/html/inc/util.inc on line
> 30
>
> This is there since they are quote "trying to find the bottlenecks now."
>
> I know they are working very hard and I respect that but we all have our
> limit of competence. I guess we could all send one dollars to the project so
> they could deal with a professional ebusiness firm to clean up this mess and
> make things work once for all.
>
What you are seeing is that testing was not completed during the Beta project, and the load is breaking things. They are fixing things as fast as they can. If a program has never been tested in a particular manner, it will break when run that way. I develop software for a living, and I recognize a phase that is normally hidden from most users.

@bfarrant:

They had already created 5 million accounts, so turning off account creation would not have helped much.