[djohnson]

Does anyone know what the difference is between 3.0 and 3.0.3? Maybe it will help those of us with 1 meg caches.

[DocWest]

Yeah. maybe.

But it's gonna produce slower results for everyone anyway.
It does more processing.

The seti folks are complaining because their servers can't keep up with the load, so they figure they'll slow everyone down to decrease the load.

damn. looks like my times are just gonna keep dropping

According to the SETI 3.03 page, Berkeley is limiting their bandwith, not their servers can't keep up. Maybe its just an excuse by I don't see why it would be.

http://setiathome.ssl.berkeley.edu/version303.html

------------------
THE Master has spoken

[This message has been edited by MasterZeus (edited 12-14-2000).]

[druber]

so is v2 actually obsolete, then? has anyone using 2.04 been rejected when trying to return/dl a new wu? reading that seti link in the last past made it sound like it's coming soon, anyway, but as always the info is vague vague vague.

[reader50]

2.04 still going strong here. I get the "upgrade soon" message every time, but it keeps offerng new work units to process.

[slboett]

Yep, 3.x is miserable if you are only out for units! If you are trrying to really "help", then I guess it doesn't matter - but really, who's in this for such noble reasons???

Scott

PS - 2.0.4 seems to work fine, although the upgrade message will still pop up, as was stated.

[druber]

is the upgrade message in the form of a dialogue box? or does it just pop as text within the seti data screen? i have a computer on the network and don't want to be there to hit enter to keep it running..

[slboett]

Dialog box - and it doesn't appear right away...very tricky of them. I once saw a little freeware app that would dismiss dialog boxes after a preset amount of time...might be good to find that or make an AppleScript that does it???

Scott

[druber]

mm, very tricky indeed. oh well. 3.0 isn't too horrendous. or at least, it could get worse if i upgraded.

[reader50]

The "upgrade soon" dialog comes up when the new work unit has been downloaded. Apparently right when the crunching starts. If you have a slow modem, or if Berkeley's servers are running behind, you will not see the dialog until the new unit has come through or until the client gives up waiting.

The dialog does not have to be dismissed before SETI will start crunching again. I have often checked in the morning. Dismiss the dialog, discover that the new work unit is 80% done. Never had a full unit finish and "lap" the dialog box, I don't know what would happen then. Maybe dismiss the dialog and have the next one pop up to be dismissed.

[DocWest]

3.0.3 is a real dog.

using 3.0 I can usually do a unit in about 11 to 11.5 hours.

i'm doing my first 3.0.3 unit and its taken 10.5 hours to do 50%.

I might just stick with 3.0

[crazycylon]

If you are in it to get lots of units done use the older version until you are forced to upgrade. I was doing a unit every 8-11 hours with 3.0. With 3.0.3 I am only 20% complete after 3.5 hours. Normally I would be at 30% complete by this time.

------------------
As a man comes, so he departs.

[djohnson]

I agree with you on the slowness. I wonder how much more science stuff they have added to effectively double the time on my 466 G3... Oh well, I wonder how all of those 4hour/wu intel/amd chips are doing now...

[Fred CHOTTIN]

I confirm: SETI 3.03 is a snail. My first WU under 3.03 wiil request more than 30 hours against 15-25 hours on my iMac rev.B 233 MHz. I hope its discriminating features is better than prior version for stopping analysis of uninteresting WU. SETI said in documentation related to 3.03 that some WU could ended very quickly if no interesting data were detected and that more interesting ones were studied thoroughly thus the differences in times for treatment. On the other hand, on version 2.04, times were practically equal whatever the WU.

------------------
Fight against the Dark Force (Microsoft)

[DocWest]

Well, my first unit with 3.0.3 finished in 18:17, compared to about 11:30 for 3.0
Pretty damn slow.
I think I'll just use 3.0 from now on.

[poocat]

this is just silly. 3.03 is taking absurdly long. 8 and a half hours for 50%?!?
with 2.04 i was getting seven hour units.
with 3.0 i was getting 6-10 hour units.
now? uggg.

does anyone have 3.0 up somewhere i can get it from?
i don't have it any more, and that makes me sad.

thanks,
poocat.

The SETI page states that the 3.0.3 does its analyze within a chirp rate of +- 20 Hz/second instead of +- 10 Hz/second as the 3.0 client did. I suppose this will about double the time it will take to process a unit. For all of us who are in this for the fun of "racing" against others, it would be nice if the units returned with the 3.0.3 client would count twice.
Anyone got a good friend at SETI who could possibly bring that idea to the SETI folks?
/Daniel

[reader50]

poocat, you have mail. Probably spam too, but definitely mail.

You're all missing the point, folks. The goal of this experiment is science, not accumulating the most units crunched. And besides, soon version 3.03 will be mandatory for everyone and the speed difference between versions will be moot.

I'm glad to see they are doing more processing on each unit.

[poocat]

oh, definitely spam. spam, spam, and spam some more...
...but also mail.
thanks a lot!

back in the ranks of the speedy.


poocat.

[elzinat]

is there ANY chance in Hell that SETI finds anything? Of course not!
the science is only "the point" in theory, not in practice.

The Real point is that it's fun! (and wastes lots of power).
and when everybody gets bored of SETI (which may never happen), the door will be opened for more useful distributed computing projects like this.

So crunch away with the fastest thing you've got!

[NB: This post doesn't exist.]
------------------
be happy!
-mac freak

[This message has been edited by elzinat (edited 12-19-2000).]

[druber]

uh oh, someone's making a good point. seek shelter if you can. at least duck behind your sturdy minitower. and by all means, ignore the posts that don't exist!