[seqfan] Re: Excessive b-file, perhaps?

[Tom Duff]

Quoth David Wilson <davidwwilson at comcast.net>:

> Perhaps A000959 has an unduly large b-file (200K terms).  It bogs the
> graph, and even downloading the b-file is noticeably slower than
> instantaneous.
>
> I don't mind having the file around, but aren't 10K terms sufficient for
> a b-file?

It is obviously useful to have as many terms for each sequence as we can 
afford to store.  The b file for A000959 is only 2.8 MB.  If the average
b-file were that long, it would only take about 600 GB to store them all. 
Certainly it would be useful to return shorter results to casual users 
(including the graphs.)

A reasonable compromise might be for the server to only give you the first 
few thousand terms if you ask to see http://oeis.org/b000959.txt but have 
other means to get the whole thing. http://oeis.org/b000959.txt?all might 
be an idea.  I am uninterested in the particular syntax, but you could, 
for example allow asking for ranges of terms: 
http://oeis.org/b000959.txt?10000-1000000 and even other, more complex 
queries.  Depends on how much the person that implements it is up for.

While we're asking for implausible features, I would love to be able to 
get a bulk download of all the b files for exploratory work. (Maybe 
there's an easy way to do that already and I just don't know about it.  I 
can certainly write a short shell script that does the job, though I would 
be loath to beat on the server that hard.)

-- 
Tom Duff.  Two red 10' stepladders used to be behind the Castle.