PDA

View Full Version : Brainstorming/ Making this project better



hhh
07-03-2005, 07:39 AM
Dear friends,

I think we are at a point of major changes in this project.

-We found 2 primes in half a year,
-had a client release,
-we have a new wiki,
-have intensified secondfass PRP
-and are starting secongpass sieve,
-and are going to have a reservation page for sieving.

Furthermore, we are considering
- a remake of the largest prime que, and
-at a moment, there will be Version3 and
- a new scoring system.

What's the problem, so?

I know that everybody is very busy, (me too), and I know how much time it takes to change things, and to do everything. But I would anyway ask to take the time to think about what is good, what not, and what can be made better. And then to take the courage to do it.

What I like:
-We have only 3 forums, just the right amount for the number of posts.
-The wiki will be a great help one day.
-The projectwas easyly understandable when I started.

Things I regret a bit:
-Only a few people are in the forum (but this is perhaps normal, and much more people are reading then posting. That, I cannot know). Is it too difficult to find here?
EDIT: I agree I was wrong. While I'm writing this edit, 43 people have already read this post. That's fine. Sorry to have underestimated you.
-The wiki is not useful at all, yet. If everybody who is able to do it heads over and writes some 20 lines, it would be already good.
-There is always a great enthousiasm when a new idea comes up, but then after 20 or 30 posts, the thread starts to sink down and dissapears at the end of the page. This is the case right now with the Reservation page and the largest prime que. The final conclusion lacks.

I know that the people behind the project are doing a lot of stuff the normal reader of the forum does not even remark. They are coordinating the secondpass sieving, factoring small k, keeping things running, and I have a large respect for them. Perhaps is the fact that one does not see this work the reason why I dare to write this post. Because I would like to see advance the things that I see and don't see the things that advance.

Please tell your opinion, and what you think is important, for this project.

If I may ask that, I would prefer this thread to remain general and to discuss specific things in the specific threads. It would keep things simpler.

Feel free to snub me, if you think it necessary, I can take it. (Though, I know that everybody is nice here and will find a polite way to tell me if i'm wrong).
Sincerely Yours, H.

royanee
07-04-2005, 01:18 AM
I've always had difficulties with wikis except for wikipedia itself. I think things are going well, and would only get rid of the largest prime queue as I don't see it as doing much other than cluttering and confusing. I'm not against people searching for a 100k digit prime, I just don't think that attempting to beat out mersenne.org's search will just deviate from our focus on our set of k. They don't have k (other than 1) and will check all n whether they find a prime or not. I'm sorry if this is blunt, but I wanted to explain truthfully.

Greenbank
07-05-2005, 01:33 PM
Wiki:

I've promised to do some pages for the Wiki. I've been playing with my own programs (and the maths involved) and think I could write a few bits on sieving, factoring, prp-testing and Proth's theorem.

Sieving:

I'm playing around with a wrapper for the sieving client.

It started off as a util for myself to help me keep track of various machines I have sieving, but it could easily be extended if people like the idea.

Hmm, I'll give it a few more days work and then I'll create a seperate thread about it, probably in the Sieving forum.

vjs
07-05-2005, 01:56 PM
Have you seen sobistrator???

Matt
07-05-2005, 02:13 PM
I made a VERY simple PHP shell script to read the SoBStatus.dat and fact.txt files and give me output like this:

Proth Sieve Status Monitor
--------------------------
Current Range: 815000 -> 816400
Factors Found: 49
Progress: 816168 (83.4%)
Average Speed: 467kps
Estimated Completion: 11 Jul 2005 01:14:19
[#################################--------]

I can give people the source if they want it.

Greenbank
07-06-2005, 04:24 AM
I haven't seen Sobistrator as I don't use Windows (for sieving).

The dual 3GHz P4 Xeon on my desk is Win2K. I run one SoB client on there for prp-testing (secondpass queue) but I need the other CPU for work.

My other machines are all Linux. (P4 3GHz - SoB, Athlon 2GHz sieve and P4 1.7GHz sieve). The last machine is sieving because it's an ancient version of Linux and the libc libraries are too old to run sbclient. That could be solved if you released a statically linked sbclient for Linux.

I'll post a description of what I'm thinking of doing in a new thread in the sieving forum.