[Top][All Lists]
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: Handling deprecated features
From: |
Dirk Herrmann |
Subject: |
Re: Handling deprecated features |
Date: |
Tue, 8 May 2001 16:14:10 +0200 (MEST) |
On 8 May 2001, Marius Vollmer wrote:
> Dirk Herrmann <address@hidden> writes:
>
> > On 29 Apr 2001, Marius Vollmer wrote:
> >
> > > Therefore, what do you think of providing this interface to the user:
> > [...]
> > > #ifdef SCM_ENABLE_DEPRECATED
> > > ...
> > > #endif
> >
> > Do you think that SCM_ENABLE_DEPRECATED should replace
> > SCM_DEBUG_DEPRECATED, or what is the idea behind the SCM_ENABLE_DEPRECATED
> > macro? Maybe some warning message system like for the scheme level?
>
> Oops, I think I was confused about names. I wanted to use the
> existing system with SCM_DEBUG_DEPRECATED, but I had to think about
> the polarity of SCM_DEBUG_DEPRECATED every time I used it. Like, "I
> want to remove this code when deprecated features are not enabled,
> that is, when the user wants to debug his code for use of this
> feature, that is, when SCM_DEBUG_DEPRECATED is true, that is I have to
> check for SCM_DEBUG_DEPRECATED == 0 in the `#if' to include the code
> in the normal case." While SCM_ENABLE_DEPRECATED would have fitted
> better with the way my brain is wired. No double negations and stuff.
Yes, I think in a similar way. IMO, we could rename this without problems
if it makes things easier. When I introduced that identifier, we had no
clear concept yet about how it would work, and it was at first meant to be
a debugging feature for those who wanted to try if their stuff uses
deprecated definitions. Now, I think, we are close to a very complete and
helpfull handling of deprecated features, and thus it appears that
SCM_ENABLE_DEPRECATED is a better name.
Nevertheless, a system for issueing warning messages when deprecated stuff
from the C level is used would also be nice...
Best regards,
Dirk Herrmann