Phobos - breaking existing code

[H. S. Teoh via Digitalmars-d]

On Mon, Dec 01, 2014 at 03:51:48AM -0800, Jonathan M Davis via Digitalmars-d wrote:
> On Monday, December 01, 2014 11:21:23 bearophile via Digitalmars-d wrote:
> > Jonathan M Davis:
> >
> > > but it's still better than leaving something as deprecated
> > > permanently.
> >
> > Right, sorry, I meant disabled.
> > But I don't like to keep those things disabled permanently, they
> > eventually should be removed.
> 
> I agree. I do think that we should strive for stability and backwards
> compatibility, but if we're getting rid of something, then I think
> that we should actually get rid of it at some point. That's part of
> why we have the deprecation process that we have.
[...]

I agree that if we're gonna remove something, we should actually
*remove* it, not just leave an empty husk behind forever cluttering the
code. That's why I suggested lengthening the deprecation cycle. But
OTOH, how long can you drag out the deprecation process before it
essentially becomes needless cruft that we're forced to carry around
"almost forever"?


T

-- 
The computer is only a tool. Unfortunately, so is the user. -- Armaphine, K5