On Sat, 2005-10-01 at 11:38 -0700, rwilcox wrote:
> --- Robert Story <[EMAIL PROTECTED]> wrote:

> > Knowing full well that it is a waste of my time to
> > do so, I proposed a mega search/replace of all ints
> > and longs with the new int32_t, with cleanup for
> > int64_t where needed.
> > 
> 
>  I am all in favor of making this modification.  Is
> there a process in place to enact change of this
> magnitude?

The reason that Robert phrased his response in the way
that he did, is that he knows things aren't quite as
simple as that.

  Replacing all int/long declarations with suitable
int32_t/int64_t (or unsigned equivalents) would be
a tedious but fairly straightforward process.  The
difficulty is that it would blow a massive hole in
the principle of backward compatability.  *All*
existing programs that use the Net-SNMP libraries
would need to be changed to match, and that's not a
step we could undertake lightly.

  It's something that we're going to need to address
eventually, but would probably require a completely
new top-number release (i.e. Net-SNMP 6.x).  And there
are several other significant changes that should
probably be tackled as part of this move.

  That doesn't meant that nothing can be done in the
meantime.  It would be perfectly possible to clean up
any *internal* declarations within library or agent
routines.  Or annotate each public API with the
relevant types that ought to be used.  Actually
changing the public API calls would need to wait,
but patches along these lines should be safe enough.


Dave


-------------------------------------------------------
This SF.Net email is sponsored by:
Power Architecture Resource Center: Free content, downloads, discussions,
and more. http://solutions.newsforge.com/ibmarch.tmpl
_______________________________________________
Net-snmp-coders mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/net-snmp-coders

Reply via email to