[Top][All Lists]

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

RE: [ft-devel] Re: const problems

From: Turner, David
Subject: RE: [ft-devel] Re: const problems
Date: Tue, 7 Jun 2005 11:25:40 +0200

I think you're misunderstanding me, but my post wasn't
that clear after all:

- I have no problem with introducing 'const' in 2.2.0. 

- I also think that keeping the non-constness of the
  current 2.1.x line is _not_ a big problem.

- I don't want a FT_CONST macro in the next release,
  (2.1.10). That's ugly as hell, and the kind of
  "temporary" band-aid that will haunt us for a very
  long time

nobody's speaking about putting const everywhere. I'm
just objecting using a macro. As I said, we'd all be
better with const where it is due, but this topic isn't
important enough to introduce yet another ugly hack.

Hope this helps,

- David Turner
- The FreeType Project  (

> -----Message d'origine-----
> De : Steve Hartwell [mailto:address@hidden
> Envoyé : mardi 7 juin 2005 10:53
> À : Turner, David
> Cc : address@hidden
> Objet : Re: [ft-devel] Re: const problems
> on 6/7/2005 12:57 AM Turner, David wrote:
> >Contrary to popular advice, putting "const" everywhere is
> >_not_ important. It's just semantic sugar, and a hint to
> >developers that can also be placed in the documentation of
> >a given function. Of course we'd all be better if it was
> >there in the first place, but they have strictly no impact
> >on performance or portability.
> >  
> >
> I'm sorry to find a reason to contradict you, but I don't think that 
> Werner is proposing to put const "everywhere", just the 
> places where the 
> contract of not modifying the contents of parameters passed 
> as pointers 
> makes sense, such as in outline callbacks.
> If you're worried that this will disrupt the current minor 
> release API, 
> please note that Werner has already reverted the recent 
> changes, and is 
> now seeking a nondisruptive solution leading to the next 
> major release.
> I can't believe I'm pleading the case of simple const 
> declarations to a 
> man who describes himself as "a bit anal on API abstraction".
> Steve Hartwell

reply via email to

[Prev in Thread] Current Thread [Next in Thread]