Re: turning "Octave:broadcast" warning off by default

[Philip Nienhuis]

Carnë Draug wrote:
> automatic broadcast was introduced in 3.6 and at the time we enabled
> warnings by default because it was very new.  This is not true anymore.
> Should we turn it off by default for 4.0?
>
> The reason is that I am starting to turn off the warnings a lot (at the
> moment, there's 8 instances in the image package).  The "local" option
> for warning makes things easier but still, should Octave be issuing
> warnings for making use of syntax features that are being used the
> way they were meant to be used?
>
> Not only is turning off the warning more verbose than calling bsxfun
> explicitely, it also gives the idea that automatic broadcasting is
> something not recommended (warnings usually means that you're probably
> doing something wrong, and turning off the warning means "I know this is
> probably the wrong thing, but do it anyway").
>
> Ideas?

IIRC the warning is also there because Octave strives for Matlab 
compatibility. Automatic broadcasting isn't ML-compatible AFAIK.

As Octave has several other ML-incompatible things for which there are 
no warnings (viz. endif, unwind_protect, double quotes, etc.), turning 
off the warning doesn't sound unreasonable.
Maybe the broadcast warning should be part of a broader warning setting 
that is triggered at any ML-incompatibility? like 
"maximum-braindamage...." or whatever it was called.

Such a setting would surely aid me - at work I use Octave "in a sea of 
Matlab".

Philip