A great example...


log in

Advanced search

Message boards : Cafe SETI : A great example...

Author Message
Seutoneus
Send message
Joined: 3 Apr 99
Posts: 3
Credit: 223
RAC: 0
United States
Message 12358 - Posted: 25 Jul 2004, 17:43:22 UTC

...of the old phrase "If it ain't broke, don't fix it".


<p>Seti@Home worked fine on my poor, old computer when I first discovered the project and downloaded the software on July 22, 2000.


<p>On July 18, 2004, Seti@Home begain tossing up fatal errors on the three machines I had since added to the project. And so, I went to the SETI web site and downloaded and installed BOINC; running Seti@Home BOINC almost four years to the day I first started crunching work units.


<p>This time, however, on three different PCs, none with a processor running below a PIII, I immediately, I ran into difficulties. "No work from project", a newly-created account which wiped out my old statistics and started over from scratch, and a great deal of frustration as the messages returned indicated that the app would attempt to reconnect in an hour to try and get more work (with one message being returned that there was no response from the scheduler, and it'd try again in four days!).


Seti@Home ran fine for four years, crunching work units 24 hours a day, 365 days a year, and it was fine.

Seti@Home BOINC has been running for a week. There's been a few work units receieved, crunched, and returned between all three machines. More often than not, there's been no work to do, sever problems, and more frustration.

<p>It's disappointing. Having my PCs work on the Seti project was satisfying. It let me be a part not only of a world community, but also part of what I felt to be an important project. Seti@Home wasn't "broke", but they fixed it anyway, and now it doesn't work.

<p>I realise that there's always going to be transition problems when things go from one version to another; but the problems faced with BOINC are so frustrating that I've decided to end my participation in the Seti@Home project for now, and let my computers crunch data for the grid.org project.

<p>I truly hope the problems with BOINC get worked out, and I'll keep checking to see if they have been. I still believe it to be an important project, but I just don't see the sense in being part of a project that simply isn't working.

<p>If the problems finally get worked out, I'll be back, with more machines working on the project, but for now, it's just too frustrating.

Profile Misfit
Volunteer tester
Avatar
Send message
Joined: 21 Jun 01
Posts: 21790
Credit: 2,510,901
RAC: 0
United States
Message 12384 - Posted: 25 Jul 2004, 18:37:22 UTC - in response to Message 12358.
Last modified: 25 Jul 2004, 18:53:29 UTC

> ...of the old phrase "If it ain't broke, don't fix it".

>
> Seti@Home ran fine for four years, crunching work units 24 hours a day, 365
> days a year, and it was fine.
>
> Seti@Home BOINC has been running for a week. There's been a few work units
> receieved, crunched, and returned between all three machines. More often than
> not, there's been no work to do, sever problems, and more frustration.
>
> It's disappointing. Having my PCs work on the Seti project was satisfying.

What needs to be kept in mind is that Seti Classic was putting out OLD WUs that had already been crunched several times. While that may have given us a cool looking screen saver it wasn't helping the science, but only boosting our numbers. Whether your CPU is just looping thru its system idle process or aimlessly working on OLD WU's the net result is the same; its just wasting cycles. Seriously, I'd rather get a "no work" message thereby allowing me to switch to another project than crunch old seti for nothing. What we really need is for Berkeley to release all the new unprocessed WU's (see CALENDAR). Once they get around to sending out the 'yellow' data HOPEFULLY these no-work problems will go away.

EclipseHA
Send message
Joined: 28 Jul 99
Posts: 1018
Credit: 238,883
RAC: 149
United States
Message 12517 - Posted: 27 Jul 2004, 0:36:24 UTC - in response to Message 12384.

> What we really
> need is for Berkeley to release all the new unprocessed WU's (see <a> href="http://setiathome2.ssl.berkeley.edu/calendar/calendar.html">CALENDAR[/url]).
> Once they get around to sending out the 'yellow' data HOPEFULLY these no-work
> problems will go away.
>

I just wonder why the "yellow" days tapes were not split for Classic months back, as Seti/Boinc is doing "about the same" science. If classic is recrunching old data, based on your link, all I can ask is "why?"! Seems there's a boatload of data that needs to be crunched based on this link, and plenty to split for seti/Boinc too!

Seem's the link is fairly up to date, as it shows the WU's we now get are those in blue! (currently being sent for crunching)

Profile Slartibartfast
Avatar
Send message
Joined: 19 Mar 03
Posts: 21
Credit: 10,840
RAC: 0
United Kingdom
Message 12814 - Posted: 27 Jul 2004, 17:06:43 UTC - in response to Message 12517.

>
> I just wonder why the "yellow" days tapes were not split for Classic months
> back, as Seti/Boinc is doing "about the same" science.

I'm prepared to be shot down wrong on this... but as far as I can remember reading, the new data was not usable on classic. Rather than make a "classic 2" which would again have been specific to Seti, the choice was made to move to (and help develop) a common platform
[url=http://setiweb.ssl.berkeley.edu/team_display.php?teamid=112265]
<a>

Message boards : Cafe SETI : A great example...

Copyright © 2014 University of California