[seqfan] Re: The OEIS mentioned on Quomodocumque

Olivier Gerard olivier.gerard at gmail.com
Sun Dec 5 10:36:03 CET 2010


On Sun, Dec 5, 2010 at 06:01, Charles Greathouse <
charles.greathouse at case.edu> wrote:

> >>      fergal on December 2, 2010 said:
> >>      If the old licence permits then it should be possible to take the
> last dump under the old licence and fork.
> >>      Just the threat may do the trick.
>
>
>
Here is my personal reaction, as I am not a member of the foundation, just
an old-timer of the OEIS.

The old "licence" as <fergal> calls it (but he does not know much
apparently) was very simple: it was copyright by AT&T, employer of Neil
Sloane.

Any download of the encyclopedia's files was for private and informal
purpose only and not to be reproduced or transmitted as a whole.

Only quotations were allowed and encouraged (in a suggested format).

(In fact most people downloading the whole OEIS were fans and editors
wanting to find mistakes or extend the encyclopedia, in coordination with
Neil).

The interest of the Encyclopedia lies in its life, continuous expansion and
correction.

Without a community of contributors and an infrastructure to maintain and
display, any clone or "fork" of the OEIS is without interest.

I believe that the OEIS foundation should define a way for an individual to
download a full snapshot of the OEIS for research purpose in a mutually
beneficial way, but I am not shocked that it is currently not permissible by
default as it could represent a large load on the infrastructure of the
OEIS, especially if it is not done correctly and also as it could be
construed as an authorization of anyone to crawl and clone the OEIS without
attribution. I feel it would be normal for the foundation to be able to keep
track of the individuals who downloaded such a large set.

I remark that the current 5% ratio (let's convert it to 10.000 sequences) is
large enough for interesting sequence families.  For instance there are
about 6000 sequences using the "walk"  or "path" words anywhere. There are
169 core sequences and 6300 nice sequences.

Cheers,

Olivier



More information about the SeqFan mailing list