Please Base 3

[PawnProver44]

Nothing Big, this is just A quick request that someone should consider adding a base 3 project. Base 5 sounds stupid to me (when compared with base 3). Base 3 is easier than base 5, so how come people are choosing to do Base 5 instead?

[LaurV]

We do both.
This is the time when I miss RDS.

[VBCurtis]

Originally Posted by LaurV

This is the time when I miss RDS.

OP-
You ought to consider perusing all the subforums on this site. You might find what you're looking for. And why would you start a new thread for this when you not only already requested this in another thread a few days ago, *AND* were directed to the link to find the project you requested? Argh

[LaurV]

Originally Posted by VBCurtis

OP-
You ought to consider perusing all the subforums on this site. You might find what you're looking for. And why would you start a new thread for this when you not only already requested this in another thread a few days ago, *AND* were directed to the link to find the project you requested? Argh

Not only, but he made a mess in other threads too (srsieve) refusing to read help files after indicated so by many users, and asking silly questions like why his system can't find a file created by himself few minutes before (most probably typo in the name). He really channeled my inner RDS, I had the feeling that he is doing it on purpose, and only being very busy here at the office kept me from posting an angry reply there. I think we helped him enough, it is the time to help himself.

[axn]

Originally Posted by PawnProver44

Base 3 is easier than base 5, so how come people are choosing to do Base 5 instead?

What makes you think base 3 is easier than base 5? AFAIK, there is not even a proper conjectured k value in base 3 (but I haven't looked into the matter in a while, so could be wrong).

Not having 3 available as a divisor really reduces the efficacy of would-be covering sets.

(If you didn't understand any of that, you have a lot of learning to do).

[PawnProver44]

Base 3 is simply smaller. For example 3^262399 is 125197 digits, whereas 5^262399 is 183410 is digits. That saves 58213 digits and is MUCH easier to compute. Also, factors of 3 are used WAY more often than factors of 5 in everyday life. Just simply learn Base 12.

[gd_barnes]

Originally Posted by axn

What makes you think base 3 is easier than base 5? AFAIK, there is not even a proper conjectured k value in base 3 (but I haven't looked into the matter in a while, so could be wrong).

Not having 3 available as a divisor really reduces the efficacy of would-be covering sets.

(If you didn't understand any of that, you have a lot of learning to do).

Yes there is:
On the Riesel side it's k=63064644938
On the Sierpinski side it's k=125050976086

[LaurV]

Well, wrong again
Which one is smaller, \(3^{63064644938}\), or \(5^{346802}\)?

(why those values?)

[gd_barnes]

Originally Posted by PawnProver44

Base 3 is simply smaller. For example 3^262399 is 125197 digits, whereas 5^262399 is 183410 is digits. That saves 58213 digits and is MUCH easier to compute. Also, factors of 3 are used WAY more often than factors of 5 in everyday life. Just simply learn Base 12.

PawnProver, quit being so stubborn and read people's previous posts instead of posting the same thing over and over. You've been directed where to go already and still don't seem to get the hint. See responses in the previous thread:
http://mersenneforum.org/showpost.ph...1&postcount=16
http://mersenneforum.org/showpost.ph...3&postcount=19

See the Conjectures 'R Us project at http://mersenneforum.org/forumdisplay.php?f=81.

For a status of all bases see:
Riesel: http://www.noprimeleftbehind.net/cru...onjectures.htm
Sierpinski: http://www.noprimeleftbehind.net/cru...onjectures.htm

For statuses of base 3 see:
Riesel base 3: http://www.noprimeleftbehind.net/cru...e3-reserve.htm
Sierpinski base 3: http://www.noprimeleftbehind.net/cru...e3-reserve.htm

The reason that base 5 was started before base 3 was because the conjectured k-value for base 5 is much lower than base 3. Base 3 is a tremendous amount of work. See my last post. Just because it is easier to find primes for base 3 than base 5 does not mean that base 3 is less effort. On the contrary it is far more effort than base 5 because so many more k's must be searched. For the past 8 years, Conjectures 'R Us (CRUS) has been working on all bases <= 1030 that are not being worked on by other projects.

A large portion of Riesel base 3 has been tested to n=25K.
For Sierpinski base 3 only k<=1G (k<=10^9) has been tested.

Do not expect any further help from anyone on this unless it is help beginning your search in the CRUS subforum.

[gd_barnes]

Originally Posted by LaurV

Well, wrong again
Which one is smaller, \(3^{63064644938}\), or \(5^{346802}\)?

(why those values?)

And wrong one more time.

It's 63064644938*3^n-1 and 346802*5^n-1. The conjectured k-value has nothing to do with the exponent. Base 3 is more effort because there are more k's to search not because the exponent is higher. The exponent is the n-value.

Wow, that's a lot of misinformation flying around here in just a few hours.

[axn]

Originally Posted by gd_barnes

Yes there is:
On the Riesel side it's k=63064644938
On the Sierpinski side it's k=125050976086

Have all the k's below them been ruled out as not having a covering set? IOW, do we have actual reason to conjecture that these are in fact the smallest S/R Base 3 k's?