submit this sequence?

[Marc LeBrun]

My personal opinions (NJAS & editors may differ, of course):

 > Worth to submit?

Yes, but I think the commentary should be made more understandable.

 > With similar seqs: ask on the list first?

No.  The OEIS's mission is collecting *all interesting sequences*,
not filtering them out.  Having a lot of sequences makes it more
likely that programs (such as superseeker) can find valuable relations
between them.

You've obviously put a lot of effort and thought into your submissions,
and I see no reason why you should hesitate sending them in.  (The people
that *should* hesitate of course never do!<;-)

 > Alternativly I'd like to have a field in the form that is a remark for
 > the editor

This sounds like a good idea.  I often [bracket] comments for the editor,
but I always worry these might get missed and leak into the entry instead.

 > (where I'd routinely say: "discard if not interesting enough")

That would certainly be polite, but also redundant, so why do it?

All serious submissions should get the benefit of the doubt.  Aside from
the few inevitable lunatics and sequence spammers, generally if someone
has been interested enough to take the trouble to make a solid submission,
then it demonstrates the sequence was interesting enough to at least that
person, and so by definition passes the *all interesting sequences* test.

 > Hope I did it right:

Very nice.

Please let me make a few suggestions about the commentary though:

 > Words that, as binary polynomials p(x)

It would be good to be sure what this phrase means, but I'd guess
that you mean the map ...5-->x^2+1, 6-->x^2+x, 7-->x^2+x+1,...
where if the binary expansion of n contains 2^k then the polynomial
contains x^k.

Unfortunately I know of no standard nomenclature or notation
for this, although the idea keeps re-arising.  This is why I
began to advocate a "rebase" notation, in which the above map
from integers to polynomials would be concisely written as

   2[n](x) (or even just 2[n]x)

or, if you mean (as I suspect) for them to be polynomials over GF2
you might write

   GF2[n](x) (or perhaps Z2[n](x))

to clearly indicate that.

 > are fixed points of p(x) --> p(x+1)

This is hard to understand.  Are the values themselves fixpoints of
polynomials or are the polynomials they represent fixpoints or what?

Again, some notation would help: GF2[a(n)](a(n)) = a(n), perhaps?

And, most of all, a well-chose example or two would be a huge help
(eg it would directly show the mapping and arithmetic you had in mind).

 > Joerg Arndt <a href="http://www.jjj.de/fxt/#fxtbook">fxtbook</a>,
 > section "Invertible transforms on words" in chapter "Bit wizardry"

Unfortunately that link leads to confusion: I wound up on the page
near the lines:
...
   <http://www.jjj.de/fxt/doc/bits-doc.txt>bits-doc.txt bit wizardry
   <http://www.jjj.de/fxt/doc/bpol-doc.txt>bpol-doc.txt binary 
polynomials and arithmetic over GF(2**n)
...
but neither of these led to pages with the word "Invertible".
Then I noticed elsewhere the line
...
   Much of the low level bit-magic code is shown on the 
<http://www.jjj.de/fxt//bitwizardry/bitwizardrypage.html>bit wizardry page
...
which I followed to find the line
...
   <http://www.jjj.de/bitwizardry/files/bittransforms.h>bittransforms.h 
(invertible transforms of binary words: blue-, yellow-, cyan-, and red code)
...
which link I followed, but I had trouble relating the code on that 
page back to your entry.

I'm sure it's all quite coherent and interesting, but it was too hard 
to follow up!

I would suggest that when people include links or similar pointers 
they try following them
themselves to make sure they lead where they think they do, and that 
there are enough clues
there for the reader to actually figure out where to look.


By the way, if you're not already familiar with the site 
www.hackersdelight.org you might
enjoy checking it out, and perhaps want to correspond with Henry Warren.