Bases 251-500 reservations/statuses/primes

[rogue]

Originally Posted by KEP

It sounds nice with such a script. I'm not sure you can make it in time for my posting next weekend (or sooner), but for future updates, I'm sure Gary at least would like it. I know not much about using vbscript in a spreadsheet, but if it can update the tables so there is no flaws, and very little time is used, then it deffinently is worth looking into.

I think that it will be a good idea to send the list to Gary, such that no other start these bases, since I'll (even though I have to do it manually) add them in the Sierpinski code.

To be honest, I found out, that I could save tons of typing, by using a spreadsheet to create the main code regarding the 499 Riesel codes, and since most of the Riesel bases can be copy pasted to the Sierpinski side, only very little effort is needed compared to what I initially had to do

If you can make a VB script, that can create the sub-pages (i.e. like R3, S3 and other conjectures with >25 k's remaining), then I'll not make those subpages, since it will require quite a lot of work/processes (even though I only have to change the base and type "save as"), that it would be need to avoid

But how fast Rogue could you potentially have such a vbscript working?
Can you make a vbscript that can create a new HTML file for different bases like those subpages that we know from R3? (and how fast, you think)?

I am uncertain of the effort, thus make no promises. Gary told me in the past that it would be too hard to do and he might be right. Only by trying will I know. There are a number of intricacies with the data that will make such a script difficult to write. I'll try to grab some time this weekend to work on it.

I've sent the list to Gary.

[unconnected]

Riesel base 433, k=92.
Primes attached.

Remaining k's:
6*433^n-1

Trivially factors: 15 k's
Base completed to 25K and released.

[gd_barnes]

Originally Posted by KEP

If you would like Gary, I wouldn't mind helping you create the HTML code, for the unstarted bases. If you would like that, you can e-mail me the details of what part of HTML you would like for me to create, and what information you would like to have embedded in the cells of the tables, in order for you to accept it. Maybe this can help you, and too be honest I wouldn't mind using some time on creating the HTML code, such that you only need to COPY+PASTE whenever a new base is started, and then add the nescessary information for top10 primes and k's remaining.

I suggest that for bases with k>1M, it should be safe to create the HTML sites like we know them for Riesel base 3, where you only will have to add the k's remaining and the starting aswell as update date

But anyway, if you're interested, in my offer, let's discuss it further via e-mail.

Regards

Kenneth

What you're offering takes very little time off of what I have to do. It's easy to create the boxes and list the base using Microsoft Front Page. That's nothing at all. Let me give you an idea of everything I do for new bases:

1. Do the easy HTML part of copying the boxes and showing the base on both the conjectures and reservations pages.
2. Factor b and b+1 to properly check and possibly analyze algebraic factors.
3. If Sierp, check for GFN primes.
4. Pull up Robert's sheet to get the conjecture and covering set. (I don't take people's word for that.)
5. Factor b-1 to get the correct trivial factors.
6. Save off the primes.
7. Sort the primes descending by n, format them for the pages, and list them. (I now have an automated spreadsheet that does this somewhat quickly.)
8. If applicable save off the results.
9. If applicable, add the base to the 1k thread. Also sieve the k to get the correct weight.
10. Remove the base from the untested bases thread.
11. Follow up with someone if there is an error or not all of the information is listed. (Quite common)


Can you guys see why even for easy proven bases without algebraic factors, this can average 10 mins. per base? Throw in a new base that has 1 k remaining, has algebraic factors, and has results to save off and you can be talking nearly 30 mins. Check out the recent R243. Batalov only reserved it. He hadn't searched it yet. It took me 30-35 mins. to get it listed correctly on the pages. Check it out and you'll see why.

It's not just the HTML. Getting the "HTML construct" for each base is easy. Listing and verifying that all of the information is correct is not easy and adding to that updating various threads makes it downright time-consuming.

I think you've probably already seen my PM response already where I allude to the fact that what is needed is far more effort than what you are anticipating. I've had people do the HTML for me before. Ultimately it took me longer. Why? Because sometimes it wasn't right and I still needed to do a lot more than just update the web pages.

So, if you or Ian (as he has offered) is willing to do all of the above for bases 513-1024, then let me know. Otherwise, I would appreciate it if people would stop searching a ton of small-conjectured bases and wanting to dump them on me. Your CPU time would be much better spent on smaller bases with larger conjectures. Also, I think it's only fair to leave some of the small conjectures for newer people as they come in.

BTW, on your k=8 proven Sierp conjectures that you dumped on me from way back, that was what I was using to post them 2 a day for a while. I just searched the ones with one k remaining to n=25K; proving a few more. So your effort did not go to waste there. I only did it for 2 reasons: (1) So your effort wasn't wasted. (2) To "catch it up" somewhat with the Riesel side. But...I have no intention of doing such an effort for CK=10 or 12 or 14 or whatever on the Sierp side. I made that exception once.

For now, if people want to post 1 new base < 512 per day, I'm fine with that but like I said in the other thread, I/we still need to think about how bases > 512 should be handled in the future.


Gary

[gd_barnes]

I just now had time to read most of the posts in this thread.

So how are you going to do this Kenneth? The pages are dynamic and being updated constantly. Are you only going to send a page of proven bases that you and Mark have worked on or are you going to incorporate them in the existing pages? You can't do the latter.

You said that the k's with trivial factors are the same for Riesel and Sierp. That is incorrect so you need to make sure you understand what the difference is. Will you be listing GFns that don't have primes on the Sierp bases? How will you make that determination? Will you only be listing the proven bases? If not, it's possible for the Sierp side to have algebraic factors also. Will you know how to check for them? The covering set can be different for the 2 sides even if the conjecture/base is the same. Are you checking Robert's sheet for that? If the primes have the same n-value, are you sorting them descending by k within the n-value so that the largest one is listed first?

Regardless, this does not change anything regarding bases > 512. The project as it is currently designed is way too big. We need to work our way slowly upwards.


Gary

[MyDogBuster]

1. Do the easy HTML part of copying the boxes and showing the base on both the conjectures and reservations pages.
2. Factor b and b+1 to properly check and possibly analyze algebraic factors.
3. If Sierp, check for GFN primes.
4. Pull up Robert's sheet to get the conjecture and covering set. (I don't take people's word for that.)
5. Factor b-1 to get the correct trivial factors.
6. Save off the primes.
7. Sort the primes descending by n, format them for the pages, and list them. (I now have an automated spreadsheet that does this somewhat quickly.)
8. If applicable save off the results.
9. If applicable, add the base to the 1k thread. Also sieve the k to get the correct weight.
10. Remove the base from the untested bases thread.
11. Follow up with someone if there is an error or not all of the information is listed. (Quite common)


So, if you or Ian (as he has offered) is willing to do all of the above for bases 513-1024, then let me know.

I still volunteer. I do most of that on my bases anyway now. I, like you, ALSO double check things like covering sets and CK's. Just so that everyone knows before hand, I will not open up 513-1024 to an uncontrolled assault on the new bases. Even though I probably have more time that Gary does to do this stuff, I also have a life and I also have over 70 cores that I manage. Don't know what limitations I can handle but it won't be much higher that what is in place now. Getting this stuff done quickly is not as important to me as getting it done correctly.

[KEP]

Originally Posted by gd_barnes

I just now had time to read most of the posts in this thread.

So how are you going to do this Kenneth? The pages are dynamic and being updated constantly. Are you only going to send a page of proven bases that you and Mark have worked on or are you going to incorporate them in the existing pages? You can't do the latter.

You said that the k's with trivial factors are the same for Riesel and Sierp. That is incorrect so you need to make sure you understand what the difference is. Will you be listing GFns that don't have primes on the Sierp bases? How will you make that determination? Will you only be listing the proven bases? If not, it's possible for the Sierp side to have algebraic factors also. Will you know how to check for them? The covering set can be different for the 2 sides even if the conjecture/base is the same. Are you checking Robert's sheet for that? If the primes have the same n-value, are you sorting them descending by k within the n-value so that the largest one is listed first?

Regardless, this does not change anything regarding bases > 512. The project as it is currently designed is way too big. We need to work our way slowly upwards.


Gary

Answering you questions as raised:

1. I'm not going to do it, since I can't do much of the 11 checkpoints that you do, and I also got the impression that it was the HTML part that irritated you, but since that is not the case, I'm not going to do any further on that subject.

2. This is almost obvious, that I'm not going to update the HTML, I only offered to do the initial and the main HTML, and nothing further. I can still do that, if you wan't to, however it sounds like it is the most easy part of the maintenance.

3. Regarding the trivil factors, I do understand, but I may not have been as clear as you would have liked. According to what you've previously told me, all k's where b-1 mod primebase = 0, has following trivial factors:

on the riesel side: k = = 1 mod primebase (primebase)
on the sierpinski side: k = = primebase-1 mod primebase (primebase)

so in total it would have been a question of doing only 168 search and replace on the sierpinski code, once you have the completed Riesel code, where 1 is replaced with primebase-1. If the above is wrong, then others might have the trivial factors wrong also, cause that was the explanation I got in a thread a long time ago.

4. No, no GFN's

6. No I would also be listing the HTML coding for the none proven bases, such that you should just copy+paste it to the HTML file that you shows the conjectures on.

7. I've never had any luck on getting the site you referred to for checking for algebraric factors to work, so NO I would not be checking for those. Also I remember a while back, that it was stated that it was only the Riesel side that had algebraric factors. This sound like it has changed, since at least S63 had algebraric factors on 2 of its k's.

8. I'm checking Roberts list for both the conjecture and the correct covering set, so it should be okay and good to use.

9. And YES I would of course be sorting them such that the highest prime is listed first and the smallest prime is listed last.

Hope this got it all. But to sum up, I think you'll be better of letting Ian do the work, because I only offered to do the initial HTML coding and since I only have internet connection a few hours each week, it will be a problem to make it through your 11 checkpoints without people having to wait weeks or months to see their work listed on your site. So in short terms, I'm not continuing the HTML creation nor am I going to work on any HTML or base additions. So I'm sticking with the 3 reservations I have for now.

Take care

Kenneth

[henryzz]

reserving riesel 258
4 ks remaining at 2.5k

[MyDogBuster]

Reserving Riesel 367 as new to n=25K

[unconnected]

I'm working with R500 (done to 50K, continuing to 100K).
107*500^30954-1 is prime!

[MyDogBuster]

Riesel Base 268
Conjectured k = 1344
Covering Set = 5, 17, 269
Trivial Factors k == 1 mod 3(3) and k == 1 mod 89(89)

Found Primes: 865k's - File attached

Remaining: 17k's - File attached - Tested to n=25K

Trivial Factor Eliminations: 457k's

MOB Eliminations: 3k's
536
804
1340

Base Released

[henryzz]

Originally Posted by henryzz

reserving riesel 258
4 ks remaining at 2.5k

that 4 was including the conjecture k so 3
i have found two primes since:
14*258^2624-1
22*258^8471-1
so just k=6 remaining
edit:
i am at n=27k