[new feature]

Joe Crump joecr at microsoft.com
Mon Dec 4 20:53:22 CET 2000

One possibility is to add a voting feature to the reported sequences,
where the browser can report whether a sequence was helpful or not.
(I.e. two variables in terms of response NumPositive, NumNegative)
Then, sort the list by NumPositive-NumNegative.

Reasoning: If a sequence has been historically more helpful than others
then it should probably be reported first.

(Of course to protect against a mother continually voting for her baby,
there'd have to be some sort of IP-logging as well I suppose unless we
will just trust the users of EIS)

- Joe Crump (joecr at microsoft.com)

-----Original Message-----
From: Christian G.Bower [mailto:bowerc at usa.net]
Sent: Monday, December 04, 2000 1:32 PM
To: seqfan
Subject: Re: [new feature]

"N. J. A. Sloane" <njas at research.att.com> wrote:
> Dear Seq Fans,
> When you use the standard lookup page,
> www.research.att.com/~njas/sequences/,
> the matching sequences will now be listed 
> with the "best" ones first.
> To make this work better, many more sequences should
> be given the "nice" keyword.  Suggestions will be welcomed.

I remember that when this discussion appeared earlier, the problem was
with the most popular sequences (e.g. Fibonacci) where the EIS contained
many derivative sequences of the main one. Thus, I suspect, the vast
majority of sequences that need tags to make this feature work have
already been tagged. No tag is needed on a sequence that becomes unique
after a few terms. Perhaps we seqfans should send Neil a note if we
look something up, get many replies and the most important sequence isn't

Also, as an author of many sequences, I think we should proceed with
caution when soliciting "nice" and "core" sequences. Asking a sequence
author if his sequence is "nice" is like asking a mother if her baby is

> PS If there is enough demand, I could also add this feature
> to the email server, sequences at research.att.com

Is there a reason not too?

Get free email and a permanent address at http://www.netaddress.com/?N=1

More information about the SeqFan mailing list