Opinions are like noses, everybody has one, heres mine.

I was involved with the beta testing of BOINC, that seemed to go rather smoothly. Since BOINC has gone public, they have been down, longer than they have been up. I have had great difficulty recruiting members for my team because the schedulers are never responding. If you cant receive work, or send results, how do I get people to try it? I have work waiting to upload for over 2 weeks now. My BOINC computers are idle.

My suggestion would be to go back into beta, fix the problems and TEST the system before going back public again. Wait until you can consistantly send work, and collect results before going public again. Keep classic online through new years, thats working.

Hi there.

I know nothing of what is going on in the back ground, It is best to wait for a wile as the project is very new and is running on 1/3 for the hardware that
seti classic. IMHO I think they sould pull the plug on seti classic and use the hardware for Boinc/Seti as we are not gaining any more science by processing the same work units 10,000 times on set classic.

I also think that SETI classic should be shut down so the good hardware can be used for SETI BOINC.

It would be nice if they would provide information on the website if they have no new work units, so we don't waste CPU cycles and can run other projects until new SETI WUs are available again.

My own nose... erm... oppinion, I mean =)

Going back to beta probably wouldn't do much to help the project. It would slow down the number of people joining, and probably have a few people leave (especially if they reset the credits), but there would still be a LOT more users than even the original beta had to deal with. It'd be nearly as bad as it is right now.

Besides, as somebody else pointed out, what better way to stress test BOINC then to leave it public? Sure it'll make some people quite frustrated/angry, but there's nothing like an actual load to pound the servers into the ground. It's a rough ride (far bumpier than even the beta as I'm sure you'll agree =D) but these people are going to come eventually, so you might as well let them (and keep classic SETI a very viable option in the meantime as well).


_______________________________

yeh thats the reason i came to this thread. im idle more than im runnin. plus for some reason the old SETI doesnt recognise my account, yet it transfered over here. go figure.

> yeh thats the reason i came to this thread. im idle more than im runnin. plus
> for some reason the old SETI doesnt recognise my account, yet it transfered
> over here. go figure.
Opposite with me. Dropped my 28,000 + WU's in the new account. I for one am fed up. I'm going back to the old 'Seti' until they get their act cleaned up. My bet is many will never return. I should have over 10,000 on the new one but it never gets updated. As soon as my chache runs out I'm out.
>
>

> yeh thats the reason i came to this thread. im idle more than im runnin. plus
> for some reason the old SETI doesnt recognise my account, yet it transfered
> over here. go figure.
>
>

My Seti I account also did not merge with my Seti-BOINC account, although my team transferred over. Go figure.

Also . . .I have uploaded close to 75 WU in the past month with no credit for them showing, and have about 25 WU on four of my boxes that cannot upload. Thank goodness there is Predictor to move in, although both of these bOINC programs have issues!

Jammy






Skepticism Is A Virtue

Why won't my sigs post?

Skepticism Is A Virtue

Hmmm . . .I guess these forums take only HTML and not UBB code.

Skepticism Is A Virtue

We think you are right. It seems it has been put out before problems have been ironed out.
It is not a user-friendly system, so we are still running classic SETI.
New users (eg not very tech folk like us)will not be impressed, and will find something else to do.

Please please please guys, take it off-line, sort the bugs out, and then re-release it.

I understand your disappointment, and in many ways I share it (idle computers, many completed WU waiting to be uploaded, no credits).

However, please understand the scale of what they are trying to accomplish. We forget that we are not dealing with a large corporation with deep pockets, datacenters full of enterprise equipment, and 24/7 IT staff.

I can remember the initial growing pains when the classic SETI project was just starting out... there were very similar issues.

Going back to beta at this point seems pointless; they simply need time to adjust their configurations to handle the massive load that all the additional users are putting on their servers.

I strongly doubt that these issues will be around in a few weeks... Patience...

I agree, but some stats wouldn't go a miss!

My NOT so humble opinion -- They need to fix the problems, NOW and let us upload the work units we're done with, not upgrade. Besides, everyone KNOWS you never upgrade on a Friday!

Fix the problems you're having now and come back to us in 30 days when you've got the bugs squashed in the current version. THEN talk to us about upgrading to 4.x. Make sure you have work units available and schedulars that aren't down more than they're up and then we can upgrade!

They dont care if you can upload or not. If you cant then the WU's will just get resent eventually and they still get the result.

I dont know why the scrabble to upgrade is happening although there may be some embaresment felt in the Seti camp that other projects are using a more recent revision of the software they themselves are developing.

Also I suspect users are leaving in relativly large numbers to crunch for Predictor and Cpdn. By upgrading they will at least allow users the option of crunching for all projects (rumour is predictor is accelerating their plans to upgrade).

I ahve to say that looking at the record to date for the Seti team the idea of them attempting to upgrade the system fills me with dread.

But we'll see. I wonder if they will remember to make the new client available to download?

I regret switching to BOINC at this point. I rarely had trouble with the old client software, and I was nearing 1000 WUs. Now that I've switched, the database only shows the 400 I had completed when they transferred the data, and I can't seem to download any new work or upload what gets completed.

I'm of the opinion that you don't go live with software until you're sure everything is properly configured to handle it. Now they're talking about an upgrade? What good is upgrading if the schedulers and splitters aren't servicing the users' client applications?