PDA

View Full Version : Need help with explanations please



MAD-ness
01-18-2003, 05:11 AM
There are a few questions which members of my team have asked me which I have had trouble answering.

I have searched the forums here, visited the prothsearch page for Sierpinski numbers (http://www.prothsearch.net/sierp.html), visited Chris Caldwell's Sierpinski Number page (http://primes.utm.edu/glossary/page.php?sort=SierpinskiNumber), and gone over the info at the bottom of the front page at www.seventeenorbust.com and have found no acceptable answers to them.

So, here I am.

1) what exactly are cEMs? If I remember correctly from somewhere (perhaps the SoB page or the FAQ before it was pulled) cEMs stands for "Corrected Exponent Mod.' I always sort of wondered what mod was an abbreviation for. Having just dug through the Sierpinski math pages my guess would be that it stands for "modulo." Would it be possible for someone to explain in some detail (and hopefully simultaneously and/or seperately) exactly what an exponent mod is?

2) How were the k-values between 78557 and the 17 k-values that remained at the start of the SB project eliminated? The obvious assumption is that a prime was found for those k values, but there is a lot of additional data that those of us who aren't math geniuses are missing. For example, short-cuts, proofs that eliminate larges amounts of numbers, etc. I know that people like Payam and others have been working for quite a long time to prove that the famous 17 k values were not Sierpinski numbers, but how were the k numbers that were eliminated before computers were used primarily proven to not be Sierpinski numbers?

3) I will just quote the question as it was asked -

"Now this is probably really dumb, but I can't find the answer anywhere and I haven't seen it asked before.

So, as part of the Sierpinski Problem it seems to be accepted that "in 1962, John Selfridge discovered the Sierpinski number k = 78557, which is now believed to be in fact the smallest such number."

Which by itself is cool, what I'm after is how did Selfridge prove that this number is indeed a Sierpinski number. All the Prime Pages just state this as if it were fact, without linking it to anything. It interests me that this could
be proven in 1962 whereas today we are throwing heaps of computing power at the remaining, smaller numbers and it could still take more than a decade to finish."

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

Here are some of my own (likely ignorant or confused) questions that are somewhat (or not at all) related:

"k = 78557, now conjectured to be the smallest Sierpinski number, was found by John Selfridge in 1962."

http://primes.utm.edu/glossary/page.php?sort=SierpinskiNumber

How did Selfridge prove k=78557 is a Sierpinski number (and how is any k value proven to be a Sierpinski number)?

What evidence makes k=78557 as the smallest Sierpinski number be treated as a conjecture and not as an open question or a wild ass guess,?

"A Sierpinski number is a positive, odd integer k for which the integers k.2n+1 are all composite (that is, for every positive integer n)."

http://primes.utm.edu/glossary/page.php?sort=SierpinskiNumber

"n 1960 Sierpinski showed that there were infinitely many such numbers k (all solutions to a family of congruences), but he did not explicitly give a numerical example. The congruences provided a sufficient, but not necessary, condition for an integer to be a Sierpinski number"

http://primes.utm.edu/glossary/page.php?sort=SierpinskiNumber

Upon following the link provided on the Prime Pages for "congruences" I found that "one of the most important tools in elementary number theory is modular arithmetic (or congruences). "

So, apparently you can do modulo math of some sort (this where the 'mod' in cEM comes into play?). I followed a bit of another thread here regarding Paul Jobling's NewPGen and some of the participants were discussing doing something with modulo math to eliminate n values (related to sieving).

What are these "solutions to a family of congruences" that Sierpinski discovered?

I think that this might be the answer (or part of it) as to the congruences:

To get an impression of the rate at which the 39278 multipliers k < 78557 are successively eliminated, let us define fm to be the number of these k giving their first prime k.2n + 1 for an exponent n in the interval 2m < n < 2m+1. Then f0 = 7238 is the number of those k for which k.2 + 1 is a prime, the first one obviously being k = 1. More generally, the following frequencies have been observed:

m = 0 1 2 3 4 5 6 7 8 9 10 11 12
fm = 7238 10194 9582 6272 3045 1445 685 331 195 114 47 34 26

From the above tables of primes and currently known search limits we infer that f13 = 11, f14 = 18, f15 = 12, f16 = 5, f17 = 5 and f18 > 2.

http://www.prothsearch.net/sierp.html

I got to this table and I got completely lost. Anyone capable of paraphrasing or extracting the basic facts of it for me?

Also, as to how the other k values got eliminated, at http://www.prothsearch.net/sierp.html there is a table of the 49 k values which were eliminated between 1981 (k = 2897) and August 1997. Four more k values were eliminated between that time and when SeventeenorBust was began (with the only 17 k values left to eliminate).

The table shows the n value for each eliminated k value (the k, n pair that was prime, thus eliminating the k value as a potential Sierpinski number). Of the eliminated k values most contain a prime at an n value MUCH lower than the n values at which the most recent SeventeenorBust prime was found (k=54767, n=1337287). The largest of the k values included in the aforementioned chart was found to be prime at n=462058 (Lew Baxter, k=34999, n=462058) and most of the k values in that chart were eliminated at n values half or a quarter the size.

The point is, they were much easier to compute than any of the primes that Seventeen or Bust has found.

When the following was published in 1983, there were only 69 remaining potential Sierpinski numbers below k=78557. How were 78488 eliminated so quickly?



1337287


It is implied that there are patterns (or atleast theoretical probabilities) for prime numbers in the Sierpinski search. Where can one find about more about these and how they were theorized, calculated, etc.?

I am hoping for both links to detailed info that might help answer these questions as well as some general answers that laymen might not get hurt trying to understand.

If any of this has been asked before or the answers are already gathered somewhere feel free to link to them. I may not understand all I read but I am quite capable of going to a site and reading it and following links and trying to understand it.

I edited this a few dozen times and still probably have a lot of errors but at 3 AM this is about as good as it is gonna get.

Thanks in advance,

MAD-ness

edit: the code tags owned the post but without them the quoted table of info probably won't line up. See http://www.prothsearch.net/sierp.html towards the bottom of the page to find the table (and more).

MAD-ness
01-18-2003, 05:18 AM
k = 78557 was found by John Selfridge in 1962.

It isn't made clear in the materials I have looked over whether or not Selfridge was responsible for the conjecture that k = 78557 is the smallest Sierpinski number. Since it doesn't say this explicitly on the Prime Pages entry I am under the impression that the conjecture was proposed by someone else and Selfridge was simply the person who discovered that k = 78557 is a Sierpinski number.

Any clarification possible?

nuutti
01-18-2003, 05:39 AM
Smaller primes are easier to find. Most of the numbers were removed by finding a prime less that 100 digit long.
Finding primes that small was possible even using computers
in the 1970s.

And you can go to library (university library)and read one of the articles mentioned
in webpages. Like :
[BCW] R. Baillie, G. Cormack and H. C. Williams, The problem of Sierpinski concerning k.2n + 1, Math. Comp. 37 (1981), 229-231.

Yours,

Nuutti

smh
01-18-2003, 08:17 AM
Originally posted by MAD-ness
2) How were the k-values between 78557 and the 17 k-values that remained at the start of the SB project eliminated? The obvious assumption is that a prime was found for those k values, but there is a lot of additional data that those of us who aren't math geniuses are missing. For example, short-cuts, proofs that eliminate larges amounts of numbers, etc. I know that people like Payam and others have been working for quite a long time to prove that the famous 17 k values were not Sierpinski numbers, but how were the k numbers that were eliminated before computers were used primarily proven to not be Sierpinski numbers?


By looking for small primes. 2/3 of the numbers below 78557 produces a prime for N < 8, see below.



Originally posted by MAD-ness
To get an impression of the rate at which the 39278 multipliers k < 78557 are successively eliminated, let us define fm to be the number of these k giving their first prime k.2n + 1 for an exponent n in the interval 2m < n < 2m+1. Then f0 = 7238 is the number of those k for which k.2 + 1 is a prime, the first one obviously being k = 1. More generally, the following frequencies have been observed:

m = 0 1 2 3 4 5 6 7 8 9 10 11 12
fm = 7238 10194 9582 6272 3045 1445 685 331 195 114 47 34 26

From the above tables of primes and currently known search limits we infer that f13 = 11, f14 = 18, f15 = 12, f16 = 5, f17 = 5 and f18 > 2.

http://www.prothsearch.net/sierp.html

I got to this table and I got completely lost. Anyone capable of paraphrasing or extracting the basic facts of it for me?


What that table says is that there are 7238 numbers (K) elliminated for m=0 when using 2^m <= n < 2^(m+1).

For m=0 N is between (and including) 2^0 and 2^1
for m=1 N is between (and including) 2^1 and 2^2
etc

Two months ago i posted this (http://www.free-dc.org/forum/showthread.php?threadid=1910) thread. Think this makes it a bit more clear.

Of course, the 5 primes found by SoB are not included, since they were not known at that time. Maybe someone cares to update that post.


Originally posted by MAD-ness
Of the eliminated k values most contain a prime at an n value MUCH lower than the n values at which the most recent SeventeenorBust prime was found (k=54767, n=1337287). The largest of the k values included in the aforementioned chart was found to be prime at n=462058 (Lew Baxter, k=34999, n=462058) and most of the k values in that chart were eliminated at n values half or a quarter the size.

The point is, they were much easier to compute than any of the primes that Seventeen or Bust has found.


Yes, the easiest are always picked out first. The 5 SoB found were also much easier than any of the following to be found. It only gets harder and harder to find the next one.

jjjjL
01-18-2003, 01:39 PM
Originally posted by MAD-ness


1) what exactly are cEMs? If I remember correctly from somewhere (perhaps the SoB page or the FAQ before it was pulled) cEMs stands for "Corrected Exponent Mod.' I always sort of wondered what mod was an abbreviation for. Having just dug through the Sierpinski math pages my guess would be that it stands for "modulo." Would it be possible for someone to explain in some detail (and hopefully simultaneously and/or seperately) exactly what an exponent mod is?



Well, I'm sure no one else is going to step up and field this one... so here goes.

"cEMs are the pronouncable acronym for corrected Exponentiation Modulus's. To understand the origin of this unit, you have to understand a little about the origin of SB. Historically, the first versions of SB (v0.1 - v0.90) used GMP (http://www.swox.com/gmp/) as an underlying math lib. Anyone involved with SB back in April 2002 might remember the first month or so where we simply used EMs (http://www-personal.engin.umich.edu/~lhelm/primes/oldskool/). An EM was equal to one pass through the squaring (exponentiation) and modulus loop in the original GMP versions of SB. When we had vital systems of SB working, we went back and decided to redefine the work measurement. This is because the speed of the EM is not constant or linear with respect to the input size. The larger the number under test becomes, the longer a single EM takes. For GMP, the steps supposedly grow as O(n ln n) the input length of the numbers. However, we found that the FFT multiplication routines in GMP were dominated by many sub-routines which ran in O(n^2) time. Therefore, we added a correction factor and redefined EMs as cEMs. Specifically, the forumula to calculate cEMs for an entire test is:

cEMs = n^2 / 1 billion.

cEMs were great. They worked really well.

Now, fast forward half a year later. George Woltman appoaches the SB team. He offers his math routines from PRP which are highly superior to GMP's more general routines. We integrate PRPs routines but don't immediately change units. However, the unit should probably change for several reasons. For one, PRP techincally doesn't even do a modulus operation anymore... making it a little inappropriate for the unit itself to refer to modulus'. [For those interested, the modulus is done "for free" as a side-effect of the discrete weighted transforms used to square numbers.] Also, the routine is not O(n^2)... it's closer to O(n ln ln ln n). So to make things even more confusing, it turns out the PRP routines are not only faster, they also "slow down slower" as the size of their inputs increases. This means that cEMs artifically inflate slightly as test sizes increase.

Currently, Dave has isolated specific test points in the database which we are using to create a new unit the doesn't skew over time. Unless something changes, cEMs will eventually be replaced by Flops (Floating-Point Operations). Once an appropriate correction factor is decided on, Flops will replace cEMs. EMs made sense for a month. cEMs made sense for a couple more months. And now, Flops make sense going forward so we'll change as soon as possible... making this entire question on the FAQ only a depreciated curiosity. ;)"

-Louie

MAD-ness
01-18-2003, 06:31 PM
Thanks for the info everyone.

Louie: thanks for the detailed history and explanation of the cEM unit. While it might seem baffling there are people who want to know just out of sheer curiosity. Of course when dealing with a group of people who are dedicated to proving obscure number theory perhaps this curiosity isn't all that hard to relate to. ;)

The information on how the calculations are done is great too (both in PRP and GMP).

dudlio
01-19-2003, 10:08 AM
...how did Selfridge prove that 78557 is indeed a Sierpinski number?

The general idea is that Sierpinski numbers have a finite, repeating set of "factors" known as a covering set. The covering set for 78557 is { 3, 5, 7, 13, 19, 37, 73 }

Conversely, if k had a non-repeating and unbounded set of factors, you wouldn't be able to prove that k2^n+1 is always composite.

So, I'm assuming that Selfridge's proof was demonstrating the covering set for 78557.

I still don't understand the process of finding a covering set, however, in order for there to be a Sierpinski number smaller than 78557, it would have to be a highly unusual one.

http://www.glasgowg43.freeserve.co.uk/siercvr.htm

http://www.glasgowg43.freeserve.co.uk/siernash.htm

dudlio
01-19-2003, 11:05 AM
Also try this link with proofs. Scroll down to "Zero Nash Weights" for a decent plaintext explanation.

http://www.glasgowg43.freeserve.co.uk/nashdef.htm

jjjjL
01-19-2003, 02:07 PM
although the covering set appears complex, it's not. Back when we were trying our hand at creating our own custom sieves, Dave derived the proof that k=78557 is a Seirpinski Number without aid. His proof is equivilent to the original Selfridge proof... perhaps simpler to follow though.

in fact, i think with a simple explaination of modular math, it's easy to explain the entire proof in the faq if you so desired. i'm saying that prefaced by a few questions about modular math like "What is some of the math involved in this project?" someone reading though the FAQ would have a decent chance of understanding why the covering set applys the way it does.

actually, thinking about it as an outside participant, that would thrill me to have a description like that for a project I was working on.

MAD-ness - in case I haven't made it clear before... I totally appreciate the fact that you're compiling a FAQ for the project. we need one... badly. heck, we don't even have a readme file with our executable! all mine and Dave's technical communications classes going to waste! ;) well... in any case, I just want you to know that the work you do to compile a FAQ is every bit as important as the work I'm doing to maintain the project and I'm really happy to have you as a participant and a partner. thanks a bunch man.

-Louie

MAD-ness
01-20-2003, 02:15 AM
I thought about it for a while before posting but I am still not sure how to respond to your post Louie.

I do not know if I am interpreting your comments correctly and if I am not then all of the following becomes pretty much null and void:

I was not/am not attempting to compile an official FAQ for the project. To be honest, the thought hadn't even occured to me, though I know that you had mentioned that you would not mind if someone compiled one.

My goal with compiling a FAQ was two-fold:

1) create a much needed source of information for and about the project (as I percieved a lack of said resource) that can be referenced by users

2) create a useful resource for my teammates to utilize (a place I can easily point them to for explanations) and, hopefully a resource that will raise awareness of Team Prime Rib by being useful to other project participants.

----

I fear that this will come across in the wrong tone or be mis-interpreted. :(

I would have no problem with you taking portions of the FAQ I am compiling (or the whole thing) and using them on the SeventeenorBust page. Nor would I have problems if other teams were to do the same (though I would definitely appreciate credit and an email beforehand would be classy). The content of these FAQ entries is hardly original content created by me. I merely collected what information I could find and made educated guesses as to what was correct and relevant. :)

Your post makes me a bit nervous. I do not consider myself the most qualified (or even well qualified) person to write any sort of FAQ for this project. However, when faced with multiple questions (most of which I have seen answered one or more times on this forum or elsewhere) about the project and the client and seeing no one else really step up with a comprehensive source of information, I figured that my ignorant, half-arsed hack of a job of doing it would be better than nothing.

Anyways, back to discussing something useful -

I have finished the initial batch of client related FAQ entries and am in the process of getting them put into proper HTML and hosted on the TPR web site. I posted the contents in rough draft form on the TPR thread at the Ars forums but the response was not as critical or active as I had hoped.

When we get them up on our page I will definitely link them here and hopefully peer review can help to ensure that they are properly worded, properly credited and as clear as possible.

In fact, I think I will attach the .txt file here.

Just please be gentle. :)

I will fix anything that people can show to be incorrect, lacking in quality, etc. as soon as possible. Most of it is stuff that I collected from quotes and questions and discussions on forums.

MAD-ness
01-20-2003, 02:24 AM
Just a note, I downloaded the .txt file I attached and it saved it as "tprsobclientfaq.txt.php" for some reason.

It still opens properly with notepad (and a browser might well open it as a text file as well, I haven't tested it).

MAD-ness
01-20-2003, 03:22 AM
It is odd how things fall together. Right after I posted the .txt file here I got confirmation that the FAQ was up on our web site. :)

http://teamprimerib.com/sob/faq.php

Also, a simple "getting started guide" written by Rob Reid a while back (I think it is modified enough to apply to the newer client versions though) and some instructions on getting a dual cpu system running that prokaryote posted on this forum a long time ago are included on this page:

http://teamprimerib.com/sob/newb.php

(note, the getting started guide is written for members of TPR so it has TPR/Ars specific elements which non-members could obviously ignore)

jjjjL
01-20-2003, 04:17 AM
MAD-ness,

woops... my bad. i assumed you were compiling a project FAQ. i'm sorry. i don't mean to be putting pressure on you and all.

i just remembered you emailing me suggesting that we put together a thread and make a FAQ. when I saw you building one, i sorta assumed you had done just that.

actually, i wasn't sure. i figured you were making one more for your team but in any case, i'm sure it's still generally helpful. i haven't read through it yet but i will soon.

of all the people who could write a FAQ, you are near the top my list in terms of who i'd like to have do it. you're on our forum almost as much as I am and I see you answering questions that normally only I was answering before... and I have yet to see you make any really blatently wrong assuptions... even when you're just guessing about how the project is run. ;)

in a better system, Dave, myself or maybe Mike would write a FAQ. I asked Mike to do it awhile ago... but he never got around to it. I'd imagine he's pretty busy with his consulting business. Dave is equally busy with his job. That leaves me... and I spend all my time on school or on other project aspects... most recently, the sieving sub-project.

also, it's probably best that I don't write the FAQ. Dave and I honestly know way too much about SB to come up with the right questions anymore. It was a little easier when I wrote the original FAQ when we were just getting starting. I was more objective and had a better idea of what the average user would want to know. At this point, everything about the system seems so obvious to me that I have a lot of trouble even thinking of what people need to know. Madness knows so much about the project, I imagine he suffers from a reduced form of my ailment.

in general, would much rather "consult" on the creation of a FAQ. i have all the answers, but I don't have the questions. and even if someone gave me a good list of questions, it would still be a chore for me to fill them all in myself. i'd prefer something like what's going on right now... someone (madness in this case) fleshes out the answers that he can, and I just fill in the details.

beyond just the FAQ, there are several things that i could use help on if someone were interested in being a project assistant of sorts. actually, Dave and I have often fantisized about having an assistant (http://www-personal.engin.umich.edu/~lhelm/primes/sbassistant-xxx.gif) Dave will kill me for posting that but it just had to come out. ;)

-Louie

MAD-ness
01-20-2003, 05:17 AM
Either/or.

Hopefully with a good chunk of it done that will give people something to build on that is a bit less intimidating.

Anyways, I don't wanna think about math or FAQs or HTML for a bit as it is (past) time for bed. :)

I will be more than willing to work with whoever, on whatever to get something put together.

Cowering
01-20-2003, 04:27 PM
MAD-ness,

Could I use parts of your FAQ to help my teammates? It's now impossible to explain all the possible ways to get SB working to newbies i'm trying to recruit.

My own question about seiving vs sb is.. i see the 'highest n' in the client climbing rapidly towards 'max n tested' on the project stats page (observed when I return a block).. is there any page that lists the highest N yet sieved? I'm guessing only n=2m-3m is complete, and we could run out pretty quickly if we get a large surge in CPU power (assuming that 99.8% of the numbers are eliminated via the sieve already)

It is fun chasing after you in stats though, however, it's a pain to explain the whole SB setup to each new person recruited.

Cowering
TeamRetro Leader

Nuri
01-20-2003, 06:23 PM
Great job Mad-ness.

Louie, I tried to compile a couple of questions for the FAQ for beginners. May be you might find some of them useful (and some really silly, but that's the logic). Sorry, they are not in any order (in fact, some may be asking the same thing in different ways). I am writing as they come to my mind.

PS: Just for brainstorming and to give you some ideas. Not that I am really asking those. :D

PS2: I think you should definitely have a question submission button at the top of the FAQ page when it's on. I'm sure you'll receive some questions that none of us can even dream of.

Regards,

Nuri

-----

What does seventeenorbust mean?
How long will the project last? (or When will the project finish?)
What are you doing in this project?
Please explain briefly. What is going on here?
Why should I participate?
What do I gain by participating seventeenorbust project?
Can I give a link to the project in my web site?
How can I cite the project?
What platforms are supported?
I belong to the press and want to write about seventeenorbust. What material can I use?
There are tons of distributed computing projects out there. Why is SoB better?
What is the science behind the project?
What is the difference between GIMPS and SoB?
Is there any chance that I will find a 10 million digit prime if I run SoB?
How can I understand how many digits will it be if the number I am searching now turns out to be a prime?
Can I participate if my computer does not have an internet connection?
Will my computer have to be online all the time?
How frequently do I have to connect to the internet for the project?
Why are you searching for primes anyway?
What will be the benefit of this project to science?
How long will a test take?
Will the work I have done be lost if I shut down my computer in the middile of a test?
What are the chances that I will find a prime?
Who is Sierpinski?
What is a proth test?
What is sieving? Should I sieve too?
I stopped and exited the client, but I can't figure out how to start is again. What should I do?
What are the hardware requirements to run the program?
What are the system requirements to run the program?
Will I have to interfere a lot?
What about security?
Why don't you have fancy graphs like Seti@home?
Does this program have a screenserver version? (hmm, in fact, why not? if it will attract 20,000 more users and give a real boost to the project)
Can I SoB while running a scrrensaver?
Can I run SoB and Seti@home at the same time?
How will I know when to upgrade the client?
How do I create a team?
How do I participate in a team?
Will my stats be affected if I change my user info?
What will happen to my stats if I change teams?
I chose my county as Burkina Faso but it doesn't show up in the production by county page?
What does IANA IP address mean?
What will happen if I don't return a test for a looooong time?
Will I get credit if a prime is found using my computer?
How does the computing power of SoB compare with existing supercomputers?
It says result 3 in the log file. What does it mean?
It says "residue:", then some letters and numbers afterwards in the log file. What does it mean?
What does a block mean?
Should I select "transmit intermediate blocks" in the config menu?
It says "temporarily unable to connect -- block added to submit queue" in the log file. what does it mean?
It says "couldn't report to server [can't connect], retry in 1200 secs [error: 0]" in the log file. What does it mean? What should I do?
It says "couldn't report to server [report denied], retry in 3600 secs [error: -3]" in the log file. What does it mean? What should I do?
Can I process the same test on two different computers? (Will it finish earlier?)
Can I process two tests at the same time?
I have a dial-up modem. How will I understand when to connect to the internet?
What can I do to run SoB faster?
Will I double my output if I run two clients at the same time on the same computer?
Why is it not possible to choose the test I want?
How does the server distribute tests?
What is the deadline to return a test?
What proportion of the tests assigned do not return?
Will I be penalized if I don't return a test?
How do I run SoB as a service?
How are we doing in terms of largest primes?
What about privacy?
Are there any safeguards against people sending in fake primes?
Are there any safeguards against people sending in fake composites?
Are there any safeguards against people that cheat to climb up the stats?
How will I understand if my client found a prime?
Will I be credited for running the project?
What are the protected features of the website you are mentioning in the signup page?
How safe is my personal information that I gave against attacks from hackers?
My computer is slow and I want to participate. Why don't you reserve some smaller numbers for slower computers?
Why don't you publish newsletters?
How can I subscribe to your newsletter?
How much will it cost me to participate in SoB?
What will I gain by participating in SoB?
Is this software open source? What is your policy on that?
How can I report a bug?
There are trillions of primes out there. What is so special about "Sierpinsky primes"? (that's a tricky FAQ)
I want to run SoB on my PC at work. Should I get the concent of my boss?
I can not find the answers to my questions on this page. What should I do?
How can I follow my progress w.r.t. other participants?
How can I follow the progress of the project?
I want to participate in different distributed computed projects a well. Which projects do you suggest?


-----

Wow. I started for a couple, and it came out to be a hundred. ;)
I guess it will be the largest FAQ in DC history if you answer all.
That's all for now.

MAD-ness
01-21-2003, 03:13 AM
Nuri:

I intentionally set aside the more project/goal/history related questions for now. It was simply too much at once and I was still trying to put together in my mind a better picture of the math and history, etc. So, I went with what I felt more comfortable with and worked on the client questions. Many of your questions I had already thought of, seen, and some I have already written up (but not included). Some I hadn't thought of.

Thanks.

(I think, see the title of this post for further information). ;)

That list gives me something to get cracking on. At this point I think I will just go through and address what I can, add them to the TPR FAQ and then take the time to re-evaluate where I am at. If I look at the whole thing at once I get a bit glassy-eyed. :)

Cowering: The short answer is "yes."

The long answer is: If I quoted someone you should obviously quote them as well. Most of these are based upon information that I gleaned from others, so I don't make any claim to intellectual property rights of any sort. ;) Credit to "MAD-ness of Team Prime Rib" somewhere if you use whole entries would be appreciated.

That said, if you just rip them off I am not likely to notice or to come and holler at you even if I do. ;)

BTW, does Team Retro have a web-site? We at TPR are curious about who is out-producing us. :)

Suggestions that I have so far for the FAQ is to have it indexed and to have it categorized. Both of these I will add as soon as I can get around to it. Another suggestion was to add in links when sites are mentioned (i.e. the log-in page) and I will add that as well (probably sooner than the previous two suggestions as it is easier).

Cowering
01-21-2003, 12:43 PM
Mad-ness, I'll try to stay a little more on topic this time.

TeamRetro is almost exclusively made up of users of my old game cataloging tools ( found at www.retrogames.com/cowering.html ). We just find that old games have much more replay value than the 3-d crap they call games nowadays. I've made a lot of contacts in all the years i've been doing the databases, and now i get to "enjoy" some ROI.

BTW, Does TeamPrimeRib reallocate CPUS to stay #1 on your various projects? I'd hate to think i gotta race a team with an army of reserves!

We have so many high powered machines because it takes a p4-2.0GHZ+ to emulate some of these 20 year old game consoles, and there is no reason to waste all that CPU time when you are not playing Pac-Man or Space Invaders :)


Back on topic: I FINALLY figured out why i have so many 0 cEM users. They are making an account on the website, but not getting the password email for upto an hour... In the meantime, they are trying to make another account with the client in config->signup. Maybe you can explain in the FAQ that using 'existing user' in config is useless!! It might have been useful before, but why would you HAVE to click existing user when all that is needed is to fill in a username on the main config menu??? I might have missed something from the previous FAQ.

Also, how does one get a user into the 'Ex-Members' list of a team? I have two on that list on TeamRetro who can't remember how they did it :(

horus2
01-21-2003, 01:37 PM
Originally posted by Cowering
BTW, Does TeamPrimeRib reallocate CPUS to stay #1 on your various projects? I'd hate to think i gotta race a team with an army of reserves!


Yes. TPR and its sister teams have 5-6THz all told. But most of that will stay where it is - only a small amount is generally mobile .

MAD-ness
01-21-2003, 01:38 PM
I believe that the only way to get a person on your "ex-member" or "former member" list is for them to leave the team.

They probably did this unintentionally.

In regards to the existing user sign-up in the client, I am not 100% sure that I understand you, but I think that this is a relic in the client from the time when you could sign up on the web page OR via the client. At first you had to sign up on the page, then they had it so you could sign up via the client. Finally, they got the log-in and advanced stats set up on the SoB site so now you have to sign up via the web page for good (if you do a new sign up). That is just a guess though.

As to Team Prime Rib...the Seventeen or Bust team is a sort of off-shoot of the Team Prime Rib GIMPS team which is pretty large (probably 500-1000 ghz at a rough guess) and we are all part of the Ars Technica DC teams. The Ars teams as a whole have a HUGE amount of processing power. That said, we are a pretty diverse and scattered group, there is no supreme overload dude directing people to one place or another we are pretty anarchic. ;)

It is possible that people might come to help from elsewhere but most people are busy with other projects. I guess time will tell.

Cowering
01-21-2003, 05:09 PM
Originally posted by MAD-ness
I believe that the only way to get a person on your "ex-member" or "former member" list is for them to leave the team.

They probably did this unintentionally.


No, we were trying to make all the users who will never actually 'login' go away.. The ones with multiple failed attempts to create a user account.

The best I can do is to remove them from the team since it won't matter if they never got/finished a test. But there are a couple of users who want to combine their two or three usernames into one account, but when they leave the team, it shows them with no affiliation, but they never make it to the ex-member list, so we lose the blocks they finished. I'm baffled at how 'MrJeff2' got onto the ex-member list, as I can find no where on the website to click to make it happen.

MAD-ness
01-22-2003, 01:37 AM
I don't know what to tell you. Maybe Louie will see the post and have an answer. I haven't fiddled with the team management options much. I guess I will add that to the "The List." ;)

jjjjL
01-22-2003, 02:21 AM
people who quit your team become ex-members.

each block is associated with a user and team. when the stats script builds the roster, it uses a seperate table of users and sees which are currently associated with your team. then to associate blocks with them, it finds all the blocks done by members of your team. if there are blocks completed by your team but that person isn't on your user roster, it assumes the person must not be on your team anymore and is therefore an ex-member.

people who you remove with the admin options should become ex-members too. all it does is null out their teamid. looks like it worked when i kicked SunOS off my team (Michigan) as a test.

if you give me usernames, and tell me more details i could look into it a bit for you.

-Louie