SB has come a long way

[b2uc]

I was just looking over some old Sierpinski sites and over at the riesel number sites and then went on to look at numerous so called distributed computing efforts. The SoB project sure has come a long long way. My hat off to you ('You' being a collective you of everyonet that has contributed) on such a great job on doing...but not limited to... the following:

1. Making an easy to use, almost idiot proof client.
2. Making said client faster and faster and always striving for optimizations.
3. Making an incredible sieve that is not only fast but able to sieve 12 K's at once while giving the project the versitility to check in and out p Ranges.
4. System of checking in and out K and n values automatically thru the client.
5. System for both team and individual stats that drive us stat heads to try even harder.
6. Making the site beautiful. (Some projects don't get the simple fact that a very nice, professional, easy and logical layout filled with tons of information is what will drive membership up.)
7. DC Forums that are both highly active, informative, and used to make the sieving effort flow smoothly.
8. Always striving for new and improved ways to speed up the finding of Primes. The P-1 Factoring is a great concept whose time is almost here to shine in help reducing the work load for PRP testing. P-1 Factoring built into the SoB client would be super helpful.
9. Great stats...or did I mention that already:)
10. Workload spread so that each K progesses almost evenly thru the n values.
11. Maximum checkout times for individual work units so progression of the project can keep moving forward and you don't have to worry about if some person you haven't heard from in 4 months is working on a unit or not.
12. Friendly atmosphere that actually incourages the understanding of the project, number theory, and exactly why we are doing this, instead of an elitist, threatening environment held only by a few.


I'm sure you other guys and gals can list more. It was just a reawakening of an appreciation for the hard work that has culminated from several people to make SoB, in my opinion, the best distributed project on the net at the moment.

[larsivi]

Don't forget the cool accounts you can use if you don't care too much about stats:

1. secret - doublechecks PRP-tests, especially those predating SOB (some numbers at http://www.seventeenorbust.com/secret/ )
2. supersecret - doublechecks the doublechecking (the same as secret at the moment)
3. el33t (or was it l33t?) - gives PRP-tests that will give the biggest prime ever if succesful. The Max n Tested at 13.4M is the result of such a test.
4. garbage - assigns the lowest pending but unfinished test (I don't know if the same test as assigned to all garbage clients)

[eatmadustch]

Originally posted by larsivi
3. el33t (or was it l33t?) - gives PRP-tests that will give the biggest prime ever if succesful. The Max n Tested at 13.4M is the result of such a test.
4. garbage - assigns the lowest pending but unfinished test (I don't know if the same test as assigned to all garbage clients)

it's 31337 actually, not el33t, not l33t ;)
Louie has also clearly stated that he wants noone to run garbage, the whole point of it is that on their machine and not some "quitter's". Also, there's no point in you running garbage, you just get normal prp tests (just slightly lower) assigned!

What I really like about sob
the math is not too dificult to understand (not like some dc projects trying to falsify the zeta function!)
the client -- eventually ;) -- doesn't need to be baby-sat
nice stats (or have they been mentioned already ;))
sob has already found a prime, and will probably find another
sob is actually doing something important (not like say projects trying to break encryption [distributed.net, the neo project ...])

what I don't like about sob limits itsself to 3 points:
the forum is only accessable through the "forum is under construction" page, which has been there for quite a while now. Why don't you just take that page out louie, and link it directly (in a new browser window, if possible) to this forum?
the cEM/s unit should be replaced!
the IP to country translations isn't solved very elegantly. I have to go through a like 10 page list and search for the few empty fields to put in a ch. I think there should be one button where I can click "I am from Switzerland (ch)" and then forget about it!

[larsivi]

Also, there's no point in you running garbage, you just get normal prp tests (just slightly lower) assigned!

I don't, but I DID test it ouf of curiosity. I was then assigned the k=22699/n=3667222 test, which is the lowest unfinished test on the stats page.

[Keroberts1]

if someone was to run garbage then would it also assign the same lowest test to them or would they get a different test than one of louie's computers running it?

[MikeH]

if someone was to run garbage then would it also assign the same lowest test to them or would they get a different test than one of louie's computers running it?Simple answer - yes.

You can check for yourself by changing user to "garbage", changing to something else, then back to "garbage", each time clearing the cache. You'll get the same unit which is always n=lowest 'n upper bounds' +1.

This is exactly why there should be only one relatively slow PC running garbage - multiple garbages will just result in the fastest PC doing the work, and the other being wasted.:spank:

Mike.