On Mon, Sep 17, 2007 at 12:25:40PM +0100, Chris Webb wrote:
> Szabolcs Nagy <nszabolcs_AT_gmail.com> writes:
>
> > On 9/16/07, Julien Danjou <julien_AT_danjou.info> wrote:
> > > Except for people applying multiple patch on the now same and uniq file.
> > > Everything will be fuzzy.
> >
> > that's why you should put your modifications in a separate .c file if
> > possible
>
> Hmm. Speaking personally, I think including a .c file in a header file is
> a stylistically hideous hack---something I could never bring myself to do
> in public! Patches which aren't just additional layouts or convenience
> functions will probably non-trivially touch existing code anyway.
It's definately a stylistic hack. But I can't think of a more
simple way to allow access to all dwm-stuff without the need to
hack together yet another .h file and patch config.mk
accordingly - and to declare all variables and functions as
extern's - which does the same in the end, but in a more
complicate way.
> However, I suspect maintaining a local tree descended from a single
> upstream dwm.c will be nicer than descending from multiple file dwm. In
> particular, noise from variables moving from static to extern (e.g. in
> taglayouts and pertag where previously local structures suddenly need
> initialising) in patches goes away.
>
> One question: why do you explicitly declare all dwm.c functions static
> given that you only have a single source file anyway?
Yeah, that's unnecessary. I remove that - it will also consume
less disk space then ;)
Regards,
-- Anselm R. Garbe >< http://www.suckless.org/ >< GPG key: 0D73F361Received on Mon Sep 17 2007 - 16:28:38 UTC
This archive was generated by hypermail 2.2.0 : Sun Jul 13 2008 - 14:53:32 UTC