can you guess what these sequences are

I've run into a weird problem with parallel ports. I have been making devices that connect to the parallel port using EPP mode since 2001. Although a few PCI cards and multi-I/O chips had problems, it all seemed to settle down by 2006, and they all worked. I've been using Dell motherboard ports and 3 different generations of SIIG PCI ports for years with no trouble.

Then, a customer had a Startech PCIe parport that didn't work. This used the Oxford OXPCIe952 chip. I got a SIIG PCIe card (used the same chip) and got the same result.

What happens is that as long as the control register has bit 5 low, for write mode, everything works as expected. But, if I put ctrl reg bit 5 high, for read mode, and read the EPP data port with an INB instruction, it does NOT generate the DATASTB/ pulse. So, it appears there is no way to read data back from the target device in EPP mode.

I got a Saba PCIe card from Amazon, and it worked immediately. I assume various printers and scanners read data back in EPP mode to sense toner and error conditions. It is hard to believe that an error this big could go into production. The EPP spec is really poorly written, there are timing diagrams with no numbers anywhere on them, and some multi-I/O chips even violate the implied order of signals, like producing the strobe signals before the data lines or WRITE/ are asserted. But, this chip just seems to not work at all when trying to read in EPP mode, and I can't find anything that I'm doing wrong.

Thanks in advance for any ideas!

Jon

fallback to SPP mode? :)

FWIW, none of them are in OEIS. When you get the formula, they'll probably be interested in it.

So how much money do I win if I describe the sequences?

Tim

-- Seven Transistor Labs, LLC Electrical Engineering Consultation and Contract Design Website:

I've got a feeling that these sequences would pass whatever statistical tests you'd need to conclude "the sequence of natural numbers with some type of random noise thrown in" pretty well. ??

Hi,

They aren't totally random:

sequence1 has no values that are divisible by 6 except for 6.

sequence2 has no values that are divisible by 6

sequence 3 has no values that are divisible by 6 except for 12.

Sequence3 values are always +-2 from multiples of 6, except for 12

mod6 is a big part for all three of the sequences.

Other than that it is probably pretty random, I would be surprised if there is an easy formula.

cheers, Jamie

Hi,

I've broken sequence2 down into three different sequences now:

original sequence2: (name: class1)

new subset sequences from original sequence2 above:

sequence1: 5,9,15,21,39,45,69,81,99,105,111,129,...

sequence2: 14,26,34,40,44,64,86,104,106,110,...

sequence3: 22,32,46,92,116,130,...

Probably not much help..

cheers, Jamie

I just checked oeis and the new sequence1 is in there:

"Numbers n such that both n+2 and n-2 are prime"

or

"Midpoint of cousin prime pairs. The only prime is 5. All other terms are multiples of 3."

The other two new sequences I made aren't in oeis.

cheers, Jamie

sequence2: 14,26,34,40,44,64,86,104,106,110,...

The above sequence are the numbers n such that both n+3 and n-3 are prime (midpoint of sexy prime pairs)

sequence3: 22,32,46,92,116,130,...

The above sequence are the numbers n such that n+(n-3) is prime and (n+(n-3))-n is prime

cheers, Jamie

That sequence2 from above of sexy prime midpoints, is actually a subset sequence of the midpoints of sexy primes (it isn't all the +-3 sexy prime midpoints). The additional criteria for sequence2 that makes it a subset sequence is that there are no other n-x and n+y where x=y and x and y are both prime and n-x is prime.

The full sequence of sexy prime center points is a combination of sequence2 in addition to another new sequence called "Z(count)=2" sequence:

sequence2 (from above):

and the "Z(count)=2" sequence: (for numbers where one of the two Z(count)'s has an x value of 3, all, of the numbers have a single n+-3 sexy prime pair, which is the whole sequence except for the number 12 which has an x value of 5, and is the center point of a 5x2=10 prime pair, not a 2x3 sexy prime pair)

giving the full sexy prime midpoint sequence:

Why is this interesting?!

It is interesting because the full list of sexy prime pairs is composed of the set of sequence2, which is restricted to having only a single primal pair, combined with the Z(count)=2 sequence, which is restricted to having only two primal pairs, one of which has an x value of 3, and a y value of n+(n-3)

So the two Z(x y) pairs of the "Z(count)=2 sequence" are:

Z(n-3,n+3) Z(3,n+(n-3))

So all sexy prime pairs have a midpoint that meets the condition of having either just a single primal pair, or at most two primal pairs, with one of them having an x value of 3.

Restated in (more normal?) math terminology:

There are only two types of sexy prime pair midpoints:

midpoint type1: for a midpoint n, there is exactly one x where n-x and n+x are both prime, and where n-(n-x) is prime (x=3)

midpoint type2: for a midpoint n, there are exactly two x where n-x and n+x are both prime, and where n-(n-x) is prime x=3 and x=n-(n-x)

midpoint type1 examples:

midpoint type2 examples:

These two sequences combined make up the full sequence of sexy prime midpoints, ie:

Here is an example of a type2 midpoint n = 20:

All type2 sexy prime pair midpoints n have a prime pair with midpoint n, with the first element = 3, and where n-3 is prime.

cheers, Jamie

Have you noticed that no one cares? Really, Jamie, do you want everyone to killfile you?

No but what makes you think I want to read what you wrote?

cheers, Jamie

Sometimes you need to hear what you don't want to hear. I'm just stating a fact. Everyone tired of you "primes" nonsense *long* ago.

I guess you didn't see what I wrote, so we are in the same boat, but at least I am learning about math while you are complaining when you could just as easily follow through with your meaningless threat to killfile me, which for some reason you think I care about, in fact I prefer anyone who spends their replies complaining or attempting to impersonate Bill Slowman etc would killfile me to save thoughts for better purposes.

cheers, Jamie

No, you're not but that's irrelevant anyway.

Idiot. I threatened no such thing. I don't threaten or even post when I do. I'm stating a fact.

It's obviously you who can't read.

I was kind of just hoping you'd killfile me, maybe there should be a reverse kill file, where you can select who your posts don't show up for.

No, I can just ignore you like Slowman and other SED fruitloops. Have fun playing with yourself.

uess what the next numbers in each sequence will be and

With a good math handbook, you could post these things indefinitely> What's the point?

Whats the point of 1+1?

sequence3: 22,32,46,92,116,130,...

The above sequence are the numbers n such that n+(n-3) is prime and (n+(n-3))-n is prime

For n=22, from that sequence:

3+22 isn't primal, but (3,41) is a primal pair of 22.

and 3,41,19 are all prime

See the updated graph showing primal pairs Z(counts) on Y axis and numbers 0 to 100000 on x axis, showing bands in the Z(counts):

cheers, Jamie