[Bioperl-l] Bio::Tools::Primer3.pm bug: code or doc, you decide???
George Hartzell
hartzell at kestrel.alerce.com
Tue Jun 14 13:48:48 EDT 2005
Brian Osborne writes:
> George,
>
> You're presenting a good argument that it's the code that should change.
>
> Brian O.
>
>
> On 6/12/05 6:09 PM, "George Hartzell" <hartzell at kestrel.alerce.com> wrote:
>
> >
> > The pod for Bio::Tools::Primer3::number_of_results says for Function:
> >
> > "Retrieve the number of primers returned from Primer3"
> >
> > and for it's Notes:, it says
> >
> > "Returns the maximum number of primers returned from Primer3."
> >
> > In fact, it's returning the maximum offset of the array of results,
> > also known as one *less* than the number of results that primer3
> > returned.... (see Bio::Tools::Primer3::_separate() and it's use of
> > $maxlocation and $self->{'maximum_primers_returned'}).
> >
> > Given the method's *name*, I figured it'd tell me how many primers I'd
> > gotten back from primer3.
> >
> > Given the method's Function doc, ditto....
> >
> > Given the method's Notes, it's a bit fuzzier, but it still evokes a
> > count.
> >
> > I'd like to fix *something*, so that I'll feel better about wasting an
> > hour walking the primer3 C code and pestering the authors about an off
> > by one bug.....
> >
> > So, should I change the method to add one, or change the documentation
> > to specify that it's the maximum offset into the results array (even
> > though the names then misleading).
> >
> > I'd rather fix the code, but *I* don't have to be backward
> > compatible....
> >
> > Comments?
Here's a patch that fixes it.
It changes the behaviour, code that used to work will now have an off
by one error.
There should be a "Headsup! We just made a non-backward-compatible
change" warning if when this gets committed.
g.
-------------- next part --------------
A non-text attachment was scrubbed...
Name: patch-Primer3.pm
Type: application/octet-stream
Size: 942 bytes
Desc: not available
Url : http://portal.open-bio.org/pipermail/bioperl-l/attachments/20050614/c27aff49/patch-Primer3.obj
More information about the Bioperl-l
mailing list