Re: defaulting to new_with_mnemonic sucks

["Ross McFarland" <rwmcfa1 neces com>]

Guillaume Cottenceau said:
> "Ross McFarland" <rwmcfa1 neces com> writes:
>
> [...]
>
> > the new_with_label variants are always available and in such cases should be
> > used for clarities sake anyway. the new('string') is just a convince. if you
> > want something particular to happen there's nothing keeping you from using
> > the
> > new_with_label('string') or new_with_mnemonic('string') for that matter.
>
> Of course! That's what we do (hum do you think we're *that*
> dumb?).

not at all. in order to attempt to justify my posistion/opinoin explaining
that was required, as it wouldn't be obvious to anyone who hasn't run into it
or thought about it.

> > the question was asked long long ago (when gtk2-perl-xs was first being
> > worked
> > on) and the list didn't raise objections and/or thought it was a good idea.
>
> Ouch.
>
> > mainly it was to encourage the use of mnemonics and it didn't in any way
> > prevent you from doing anything you could of otherwise. the overloaded new's
> > have a well defined, documented and adhered to behavior, that doesn't suck.
>
> Ok.

i meant no harm and had no major fundamental reason in mind when i originally
made the switch to new_with_mnemonic being the default behavior and had anyone
raised an objection back then i could of been swayed. but with 1.0rc's
happening and quite a bit of documentation already stating the behavior it's
to late with out 'it's broken' as the reason.

-rm