[seqfan] Re: Three lines, or 260 characters?

Olivier Gerard olivier.gerard at gmail.com
Mon Jul 2 16:44:04 CEST 2012

On Mon, Jul 2, 2012 at 7:23 AM, Matthew Vandermast <ghodges14 at comcast.net>wrote:

> Hello SeqFans,
> How long should the standard sequence be? It seems there may be a conflict
> between 1. the "three lines" standard and 2. the "260 characters (including
> spaces)" standard.

Longer answers:

The principal requirement for a sequence is a *minimum* of specific terms.
 The 260 characters for display, is a good rule of thumb (including all,
comma, signs and spaces) and was before the generalization of b-files but
after the suppression of the Axxxxxx reference in database lines.

Now we have b-files.  You still enter about this amount (260 characters)
when editing the sequence, and if you can, you add a b-file with as many as
you have or can compute (1000 terms is great). The more terms the better.

b-file are becoming the reference for the sequence itself, whereas value
lines used to be the only reference.

The site builds automatically a default b-files for sequences which have
and the search engine is able to search inside the bfile for matches.

> 1. The Style Sheet's "How many terms do we need?" section implicitly
> requests "enough terms to fill three lines on the screen, about 200
> characters..."
> https://oeis.org/wiki/Style_Sheet#How_many_terms_do_we_need.3F
> I believe that when Neil wrote this, he meant "digits plus commas" by
> "characters." (In my experience, the old limit used to be about 70
> digits+commas in each of the internal format's three lines.  Other people
> may have had different experiences.)
200 characters total was a good heuristics for the old cardboard format.

The OEIS Mathematica package used to format whole lines at 78 characters to
avoid formatting troubles (remember that
many editors, screens, mail systems, etc. used to format in 80 characters)
: the prefix
%Z A000000 takes 11 characters so it left around 67 characters, including
signs, by line. So 200 characters in total.
Without the A reference, it is more around 230 characters, and with the
current layout of the normal, 260 is ok.

See also my comment above.

> 2. The programs on the Sequence Tools page (
> https://oeis.org/wiki/Sequence_Tools) cut down a sequence to 260
> characters, but at least some of those programs count spaces as characters.
> So there's not always enough sequence to fill three lines.  For example,
> the "U" line here is about half empty:
> https://oeis.org/edit?seq=A212174&internal=1.
Sequences are currently stored internally without the spaces when you
submit them but they are displayed in the normal format with spaces after
each comma. The number of characters is easy to customize in the tools

> Sometimes there is less empty space than this.  In other words, apparently
> some of the 260-character programs don't produce output that is a constant
> length in the database.
See my comment immediately above about spaces. What is important is not
what is stored internally for the first three lines but how it looks in the
normal format. If you look at it for the sequence you took as an example

 https://oeis.org/A212174 <https://oeis.org/edit?seq=A212174&internal=1>

it fits nicely three rows.  But you can see than there is no real, strict
and enforced limitation as


has three and 2/3 of a row.

Charles rightly points out that it allows to have a specific feature of the
sequence in direct view, especially if you put a comment about it.

> I'm just wondering if either of these effects were consciously intended,
> or if perhaps there was a miscommunication somewhere. (For example: Is it
> possible that somewhere, an administrator recommended 260 characters as the
> new standard length for a sequence, but *without* counting spaces as
> characters?) Thanks for any clarification.
I am afraid everything here is a mix of old technical decisions and
limitations, evolution and taste.

Thanks for your care in entering and editing sequences.


More information about the SeqFan mailing list