Against deprecating aliases

[Jonathan M Davis]

On Thursday, September 29, 2011 09:09 Jacob Carlborg wrote:
> On 2011-09-29 08:36, Jonathan M Davis wrote:
> > On Thursday, September 29, 2011 08:22:41 Jacob Carlborg wrote:
> >> On 2011-09-28 21:56, Jonathan M Davis wrote:
> >>> Having toUTF and toUTFz is a marked improvement in many cases -
> >>> especially for generic code. They're being added regardless of what the
> >>> deal with toUTF16z is.
> >>> 
> >>> The reason that toUTF16z is being removed is essentially because Andrei
> >>> is very much opposed to having functions with specific types in their
> >>> names and thinks that they should all be generic. Personally, I'm not
> >>> opposed to keeping toUTF16z as an alias or wrapper to toUTFz. It's
> >>> Andrei that seems to feel stongly about it. So, if enough people really
> >>> want to keep toUTF16z, then I think that that can happen.
> >> 
> >> Seems a bit arbitrary what gets deprecated with the old API left in
> >> place and what gets deprecated with the old API removed.
> > 
> > I'm not sure that I understand. Very little gets deprecated with the old
> > API immediately removed, and _everything_ which gets deprecated will be
> > eventually removed (or it wouldn't be deprecated). What are you thinking
> > about in particular?
> 
> Lately if feels like when new functionally (or things were renamed) was
> added the old was just removed without keeping aliases or wrappers.
> 
> I base that on what I've read here lately, people are complaining that
> functions are just gone. But I might be wrong or I've missed something.
> Or maybe people just complaining about functions are about to be removed.

I'm not aware of functions just being gone. Certainly, if I caused that to 
happen anyway, I screwed up. I think that the complaints are either about how 
deprecation messages have been being handled or that the change is happening 
in the first place. For instance, for what Andrej is doing, he wants to keep 
toUTF16z, so he's unhappy about it being scheduled for deprecation. And given 
the cost of making toUTFz appropriately generic, he has a good argument for 
it, and we may very well end up keeping it. But I believe that the complaints 
about stuff going away or being replaced have been because stuff was scheduled 
for deprecation, not because it actually went away. The closest that I can 
think of to a function actually going away is a case or two where a function 
was further genericized, and it ended up taking a range rather than an array, 
so it stopped working with immutable arrays, and to my knowledge, all of the 
functions in std.array have been fixed so that that is no longer an issue.

- Jonathan M Davis