SetiStash

[tiwazz]

hi guys...
i'm a newbie in seti@home n currently using the text
version...
i have come across this acticle saying that by using
"SetiStash", someone is able to finish a block in 10 mins
or so...
can anyone verify it...

thanx

 

[tiwazz]

sorry guys...here's the url :
http://www.jsihardware.com/guides/howtousesetistash/index.shtml

 

[artemedes]

DANGER DANGER WILL ROBINSON!!!


I noticed that the page you reference links to an seti client on their server not berkeley's. This makes it very suspicious in light of reports of a hacked client running around.

Also I think they exagerate how much faster the CLI is compared to the GUI. It is noticably faster but certainly not twice as fast.

If my suspicions are correct this could get that whole team (JSI) booted out of the SETI project.

If your a TA member, please don't use this client or Setistash until we can verify that it is indeed turning in valid results. This team has worked too hard to get where we are to allow a hacked client to disqualify us.

 

[Poof]

:|

What a pile of distortion all around...

mr_mann posted the same in a thread at Ars.

The WU that he had was obviously one with alot of RFI, eg., a large spike. The SETI client is designed so that if the file size of the result.sah file gets beyond a certain size (meaning a very "noisey" WU), then the analysis abruptly ends and the result is sent in. This can happen anywhere from a minute or so after the beginning of processing, up to close to the average for that particular platform... It all depends on the types of things found, power of spikes, etc. You would still get credit, but this is just so you know.

Another thing that would cause these "fast WUs" would be bad RAM. A number of people here and elsewhere have reported suddenly getting a series of fast WU completions in a row and shortly thereafter, discovered to their dismay, that their RAM stick crapped out on them.

It's sad that JSI was trying to suggest something other than the above... (aside from the use of that hacked client).. :|

 

[artemedes]

Yah Poof. I thought about that too. I have even had a few WU's that didn't finish completely. I just forgot to write those thoughts in my post. So thanks for filling in the blanks.

My major concern was the link to the client that is not on Berkeley's servers. If you use their client you have no idea what is different about it. I may be overly cautious and it may be an exact copy with a different name, but why would you risk it? I will only use a client I downloaded directly from the Berkeley team.

 

[Assimilator1]

Tiwazz

You've got hold of the wrong end of the stick ,that site doesn't say that SETI stash reduces WU times down to 10mins ,but just that it recorded a 1 off WU of 10 mins which does happen on the rare occasion (due to polluted WU's).
Also that site mentions that the CLi SETI 1/2's WU times ,this is only true if the GUI version had the graphics showing ,in which case he must be a newbie to SETI ,whilst the CLi version is faster it is much less than he claims IF the GUI version does not show the graphics whilst running.

Darn ,I see 2 others have posted whilst I was typing this! lol

Thanks for the heads up about SETIstash anyway ,it's a new 1 to me

 

[tiwazz]

thanx guys...
info noted...
just wanna verified it...
besides, i am part of team anand too...
just contributing in my own humble way...

 

[Assimilator1]

Glad your with us.
I see you only recently started from scratch ,enjoy the speed with which you'll shoot up the charts