[Bioperl-l] ->alphabet documentation wrong here there and everywhere :-)

[Jason Stajich]

The documentation is correct we cannot assume that all implementers of the
interface will be mutable.

An interface does not assume that you can set value while a implementation
may choose to allow it.  Consider a DBPrimarySeq which is a read-only
representation of the db - it implements the PrimarySeqI interface but
does not allow alphabet to be updated while Bio::PrimarySeq does.


On Fri, 9 Jan 2004, Mark Wilkinson wrote:

> Does anyone object to me fixing the documentation for PrimarySeqI et al.
> with respect to the 'alphabet' function?  It claims to not accept any
> arguments, but every implementation of it that I can find in the
> sequence hierarchy actually does, so... the docs are wrong all over the
> place.
>
> Alternately, I could leave the Interface documentation alone, and simply
> update the documentation for all of the implementing objects who
> actually DO allow you to set the alphabet explicitly after Seq object
> creation...
>
> Which do you prefer?  Was it intentional that the ->alphabet interface
> not be get/set, or is that historical cruft?
>
> Mark
>
>
>

--
Jason Stajich
Duke University
jason at cgt.mc.duke.edu