[seqfan] Re: Policy on near-duplicates

[Peter Luschny]

Max Alekseyev wrote:

> I support this idea, assuming that the main entry in an "equivalence
> class" of near-duplicates is clearly indicated.
> I think that all information that is applicable to every sequence in
> the class should be concentrated in the main entry only (that would
> simplify editing/updating/extending this information in a single
> place).
> Ideally, the other (non-main) entries in each class should be kept
> minimalistic and marked with the "dead" or "dupe" keyword (indicating
> that they are permanent in their current form) - since their purpose
> is only referencing to the main entry.

So what is it good for then? This occurs to me as
a waste of time and space.

The only place an adaption is necessary is in the
search-algorithm which must be able to give the
correct answer in both cases.

This, however, necessitates a small change only in the
already very powerful search-algorithm implemented
(as far as I can judge from the outside).

Cheers Peter