SETI's response to Open Letter and petition

[Poof]

From here:

At about 5:10 p.m. local time (3:10 a.m. EST) I received the following mail from Dr Anderson, Seti project director, as the reply to our open letter with 822 signatures I mailed him on Monday morning. BTW, there are now 854 signatures.


quote:
--------------------------------------------------------------------------------

----- Original Message -----
From: "David Anderson" <davea@ssl.berkeley.edu>
Sent: Wednesday, November 06, 2002 5:06 PM
Subject: Re: Open letter/petition to stop cheating in Seti

Mr. "Kosmaj":

Thanks very much for the petition and the list of suspicious accounts.
We'll examine all of them.
Please thank the members of your team for helping us in this matter,
and for contributing so much to SETI@home.

--------------------------------------------------------------------------------


So, finally we got the response and the recognition from Berkeley that IMHO we deserve.

[I signed my mail with my real name and that's how he addressed me but here I prefer to stay behind my nick.]

I think that we have all seen some very good response to purging cheat accounts, even despite the recent DoS attacks of yesterday where someone or "several someones" were generating tens of thousands of new bogus SETI accounts, effectively bogging down the data servers.

Appreciation goes out to the TeAm for those who signed the petition and who continue to be vigilant in making this a fair race to find ET!

 

[Robor]

Isn't it sad that the little punk cheaters who got busted are still trying to screw with SETI? I hope the next project is more resistant to these losers.

Thanks for the info, Poof!

 

[JHutch]

DoS attacks can be hard to prevent. Usually, you have to wait until one starts and then block IPs. So, I'd imagine SETI2 will still have problems with script kiddies (just like everything else of the 'Net).

Good to see them addressing the cheating accounts. Hopefully SETI2 will close the obvious gaps in security. I remember way back when, at the beginning of the project, I duped about 30-40 units by restoring a backup of SetiQueue. I never thought they would credit more than one instance of the WU.

JHutch

 

[Smoke]

Thanks for the update, Poof.

There is still an outstanding issue that I hope gets dealt with very soon. Information has been passed on by me and others concerning accounts that have been allowed to amass hugh amounts of WUs with computers that return corrupted WUs in only minutes and sometimes seconds either "by design" or "by accident but allowed to continue operating".

Let me attempt to be a little more clear on this:

We all get these occassional WUs that are internally corrupted and aborted in a short amount of time. These are sent on to Berkeley and we receive credit for them just as we do for a normally analyzed WU. I have noticed that they seem to appear around 0.3% of the time based my observations of TeAm Smokeball. TS does approximately 1,000 WUs @ week and it is normal to see 2 - 4 WUs sent in during that time that only took a few seconds to a few minutes to complete.

We have seen on our own TeAm an occassional computer go awry and start sending in WUs at startling rates. We have been quick to respond to these situations and have had the offending computer stop processing SETI or fixed. BTW, I may be bringing up this particular topic very shortly in another thread for a TeAm Discussion.

Anyway, there are some accounts (not ours) that have had computers start running amok and finishing WUs faster than they could even be transmitted. Statistical records (read evidence) show that some people have allowed these computers to continue to run and even set up separate Qs to provide these computers with enough WUs. This was allowed to go on for a long period of time and it was NO ACCIDENT.

More on this later.

 

[Poof]

Smokeball - all you have said is true and has apparently been identified as one of the "cheats".

From what I've seen that they're doing is first addressing the accounts that have been and continue to be sent to them by team leaders/founders. Then there are groups of folks who are going through the Top 1000 to find overtly excessive WU production (ie., obscene production done by fairly "new" registrants) and also going through the Registration classes themselves to identify obvious standouts there... After that, it will be tricky to find suspicious accounts if the cheater lays low and appears "reasonable" in their WU production. Usually that attitude doesn't remain long because the cheater cheats in order to either 1.) rise in the ranks quickly to gain glory so eventually there will come a time for the need to ramp up... or 2.) cheat because they found that they can "break" the system and gain glory from doing so.

Things continue to get curiouser and curiouser...

The last I heard is that more than 9,000,000 bogus WUs have been purged so far! :Q

 

[Confused]

Originally posted by: Smokeball
We have seen on our own TeAm an occassional computer go awry and start sending in WUs at startling rates. We have been quick to respond to these situations and have had the offending computer stop processing SETI or fixed. BTW, I may be bringing up this particular topic very shortly in another thread for a TeAm Discussion.

I will be the first to say and be let known that there is a user under mine and Alex's name on the TeAm with about 190 WUs, and a very low average time. When we were trying to assimilate our college, we set up a joint account so there were no arguements as to who got how many machines (this was before the invention of the British Bulldog Racers) and there was one PC there that we think had a bad stick of RAM, and was putting out WUs in a couple of minutes each. This was isolated to this one account. We left the account tho as there were a few WUs that were completed properly and validly for the account and the TeAm, so that is why we left it on the Team. We were using OK's queue, and we contacted him and got the user blocked on his queue as we couldn't stop the client at the time (it was a weekend or evening when we noticed it).

So, that's why that user account has a very low average, and you will see that the last time a WU was returned on that account was a long time ago!

Just to clear that up again, as that has come up for discussion a couple of times in the past


Confused

 

[Assimilator1]

I don't think you have anything to worry about Garry ,you stopped it in good time
Over 2.5 yrs ago I had a PC whose RAM went defective & spewed out about 20-30 odd screwed WU's in a few days, I dont think Berkley will be worried about that though

 

[Smoke]

Hmmm