OEIS's Adjacent sequences? (OEIS XML again)
Marc LeBrun
mlb at well.com
Mon May 23 20:46:57 CEST 2005
"...would require extra ordinary amount of programming..."
"...probably one line...in a language of my choice..."
There's nothing stopping us from doing this for ourselves today.
Anyway there is no reason Neil &co should, or could, support every possible
style of interface anyone will want--either user interface or programmatic
interface.
Trying to find a constructive resolution for such debates is what led me to
start thinking about an XML-based OEIS interface (not the also
perhaps-interesting but different subject of the database's implementation
itself).
All the OEIS really need do is provide a well-defined format to output (and
input) the basic raw data over the web. Then, whether due to taste or
function or both, if anyone wants to do things differently than the
"native" way they can write their own--even offer it to a sub-community of
like-minded users.
In fact right now today anyone who wants to can do this, simply using the
"internal format" for the main sequence data and doing a little processing
to strip off the presentation matrix that it's embedded in.
Of course nowadays such information is commonly formatted in XML wrappers,
and the functionality is exposed as web services. So I concluded that
would probably be a good "medium" to target, rather than some ad hoc solution.
It seems that, instead of advocating this feature or that to add to "the
ever-growing to-do pile", we in seqfandom could actively help out by
collaborating on the actual development of some prototype standard formats
and web services.
Presumably one or more of these would eventually make sense for the OEIS
itself to support. But we could perhaps hack together initial
approximations, or substantially more if we want, without impacting the
considerable ongoing effort required to maintain the OEIS.
I apologize for my usual intermittent facilitation of this effort, but I
will continue to pursue it as I'm able, and hope others will as well on
their own initiative. I think this is really a viable and exciting
additional way that the community can directly contribute to the growth and
usability of the wonderful resource we enjoy.
At this stage proposals for formats, interesting new applications,
customized (or customizable) interfaces would all be cool.
Hugo has kindly set up a web area that can be used for further
discussion. Looks good and easy; I just signed up myself!
http://community.novacaster.com/index.pl?section=45&n=5012
More information about the SeqFan
mailing list