[Top][All Lists]
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: Header cleanliness [was: removing the 'make install'-->'makeall'dep
From: |
Alexander Malmberg |
Subject: |
Re: Header cleanliness [was: removing the 'make install'-->'makeall'dependency] |
Date: |
Sun, 22 Dec 2002 13:24:15 +0100 |
Nicola Pero wrote:
[snip]
> > * A header must be self-contained.
>
> Yes.
>
> > * A header should include as few other headers as possible. Use @class
> > instead.
>
> Yes.
>
> I mean, very good habits ... at least for the core libraries. :-)
>
> Very good habits ... hmmm .... maybe not for all. If you are not worried
> about compilation speed but need to cut down development time, or if you
> have precompiled framework/library headers, for example, you may just want
> to do
>
> #include <Foundation/Foundation.h>
> #include <AppKit/AppKit.h>
>
> this definitely doesn't follow your rules :-) it compiles slowly on
> GNUstep, yet it can be acceptable if you don't care about compilation
> speed.
True, and a person using GNUstep libraries should be free to make that
decision. However, (since I have a slow system :( ), compilation speed
matters to me. Anyway, I think the decision should be left to users of
the libraries. It means extra work for us (as library developers), but
that's ok; as library developers, we're supposed to do extra work so our
users don't have to :).
It should be noted that most GNUstep headers already follow this to a
large extent. I just want to make sure this is what we want so I can fix
the remaining places. :)
> If we get precompiled headers on GNU, that might become the preferred
> solution even to cut down compilation time, because usually you will have
> a precompiled version of Foundation/Foundation.h and of AppKit/AppKit.h,
> while you will not have precompiled versions of the specific Foundation/*,
> AppKit/* headers if you include them separately.
Why not?
- Alexander Malmberg
- Re: Problem with autogsdoc and GSFusedSilica.h, (continued)
- Re: Problem with autogsdoc and GSFusedSilica.h, Chris B. Vetter, 2002/12/19
- Re: Problem with autogsdoc and GSFusedSilica.h, Nicola Pero, 2002/12/19
- Re: Problem with autogsdoc and GSFusedSilica.h, Chris B. Vetter, 2002/12/19
- Re: Problem with autogsdoc and GSFusedSilica.h, Nicola Pero, 2002/12/20
- Re: Problem with autogsdoc and GSFusedSilica.h, Chris B. Vetter, 2002/12/20
- Re: Header cleanliness [was: removing the 'make install'-->'makeall' dependency], Alexander Malmberg, 2002/12/16
- Re: Header cleanliness [was: removing the 'makeinstall'-->'makeall' dependency], David Ayers, 2002/12/17
- Re: Header cleanliness [was: removing the 'makeinstall'-->'makeall'dependency], Alexander Malmberg, 2002/12/17
- Re: Header cleanliness [was: removing the 'makeinstall'-->'makeall'dependency], David Ayers, 2002/12/17
- Re: Header cleanliness [was: removing the 'makeinstall'-->'makeall' dependency], Nicola Pero, 2002/12/19
- Re: Header cleanliness [was: removing the 'make install'-->'makeall'dependency],
Alexander Malmberg <=
- Re: Header cleanliness [was: removing the 'make install'-->'makeall'dependency], Nicola Pero, 2002/12/22
- Re: Header cleanliness [was: removing the 'make install'-->'makeall'dependency], Markus Hitter, 2002/12/23