[seqfan] Re: The OEIS mentioned on Quomodocumque

[Charles Greathouse]

Attribution-wise, this is just William Stein's views, familiar to
NMBRTHRY subscribers.

I'm quite sympathetic to his efforts to further 'open' the OEIS.  I
see many times when this would be useful: creating an improvement to
the superseeker, a new browsing tool like Robert Munafo's "Sloandora",
finding and correcting errors in the OEIS, and so forth.  I know that
I have used stripped.gz or cat25 many times for various projects over
the years, and surely others have creative uses I haven't even
imagined.

This is a complex issue which deserves study; I wouldn't want to rush
to a decision like 'no, the OEIS is open enough'.

On the other hand, I think Stein is wrong to conflate issues of access
and legality.  Before, there was no licensing agreement allowing
distribution or use of the OEIS data, but it was easy to download
certain data (e.g., the sequence values in stripped.gz).  Now there
are legal ways to use some data.  Using data as complete as
stripped.gz are on the same shaky legal grounds as it was before the
OEIS End-User License Agreement.  But now there is no access.

So I see these issues before us:
* Determine the legal issues surrounding changes to the OEIS End-User
License Agreement.
* Determine competing needs for openness and closedness, and an
appropriate level of openness based on thise
* Make appropriate data available, or minimally tools to allow as much
freedom to our users as possible.

If for reasons I don't currently appreciate it is impossible to make
the data as open as I would like, we should at least build tools that
allow common and "good" uses to whatever extent we can.  For example,
if a project like Sage wanted to be able to do analysis with the OEIS,
building an API that it could use would be a minimal courtesy if we
could not (for whatever reason) sufficiently open the data.
Similarly, for data validation, building an interface that would allow
searching the OEIS with regular expressions would allow at least a
third of the common cases (better than nothing).

Of course these approaches would cause more work for the OEISF
(probably Russ) and would cause increased server load, rather than
offloading both to outside developers like Stein.  This is another
argument in favor of openness, albeit a small one.

I hope we can avoid hasty conclusions on this complex issue.

Charles Greathouse
Analyst/Programmer
Case Western Reserve University

2010/12/4 Benoît Jubin <benoit.jubin at gmail.com>:
> Dear all,
>
> The OEIS is mentioned on the blog Quomodocumque, together with some
> criticism on the new license agreement.  I'd be interested to know
> what the OEIS users, contributors and editors think about it.
> Link: http://quomodocumque.wordpress.com/2010/12/02/licensing-changes-at-the-online-encyclopedia-of-integer-sequences/
>
> Regards,
> Benoit
>
>
> _______________________________________________
>
> Seqfan Mailing list - http://list.seqfan.eu/
>