On Mon, 2005-11-07 at 13:22 -0500, Robert Story wrote:
> On Mon, 07 Nov 2005 16:52:37 +0000 Dave wrote:
> DS> One of the concerns expressed when I first submitted my new code,
> DS> was that people who were using the original implementation
> DS> might want to continue using that version.
>
> But you've claimed that there are no backwards-compatibility issues,
Not true.
If you look at the message where I first reported my new implementation,
that explicitly states that there *are* some backwards-compatibility
issues. Mostly fairly minor (e.g. -i vs -I), but they are there.
And the behaviour is subtly different as well (e.g. what notifications
are sent when an explicit event is specified).
> and that the new code works perfectly, right? ;-)
Well, I haven't spotted any bug reports as yet :-)
> Wes> It seems wiser to me to move the old header to old-disman-event.h
> Wes> or something so people have to explicitly request it.
> DS> That sounds good to me.
> DS> But I'd also suggest that we revert to 'disman/event.h' for the
> DS> new implmentation, and use 'disman/event-mib.h' to issue a warning
> DS> about the existence of the two alternatives.
>
> I don't particularly like this option, but will bend to the will of the
> people. I would argue that the header name for the old code should indicated
> it is the old code.
Like 'disman/old-event-mib' does, you mean?
> Otherwise how are the clueless going to know which is better,
> event or event-mib?
But the choice wouldn't be "event" vs "event-mib"
It would be "event" vs "old-event-mib"
Trying to use "event-mib" would throw an error, and force them
to choose between these two.
On Mon, 2005-11-07 at 09:30 -0800, Wes Hardaker wrote:
> Dave> Do you mean that the difference between the two header file
> Dave> names should be more distinctive, or that there should only
> Dave> be one header file?
>
> Meaning we should use the *old* name (the one people know about)
> instead of creating a new one.
But that implies (to me) that the new implementation is an
*exact* replacement for the old one. That's not quite true
in this case (see above).
> You're not going to easily get
> people to use the new code if it isn't turned on easily
> in the fashion they're used to.
But the new code is being included by default. They don't
*need* to turn it on - it'll be provided automatically.
> Dave> But I'd also suggest that we revert to 'disman/event.h'
> Dave> for the new implementation
>
> I think we shouldn't have 2. As I said, IMHO, we should have
> the following:
>
> disman/event-mib.h -> new code
> disman/old-event-mib.h -> old code
Eh? But that *is* two!?
I'd prefer to use:
disman/event -> new code
disman/old-event-mib -> old code
(with event-mib throwing an error)
This has two advantages, IMO:
a) It avoids silently moving someone from
the old code to the new code without warning.
b) It's also in line with the naming scheme
used for the other DisMan code
(schedule & expression)
> But, I'm a lone voice and not the implementer so I won't win ;-)
Don't worry, Wes. Being the implementer doesn't seem to make
a great deal of difference in this discussion :-(
Dave
-------------------------------------------------------
SF.Net email is sponsored by:
Tame your development challenges with Apache's Geronimo App Server. Download
it for free - -and be entered to win a 42" plasma tv or your very own
Sony(tm)PSP. Click here to play: http://sourceforge.net/geronimo.php
_______________________________________________
Net-snmp-coders mailing list
Net-snmp-coders@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/net-snmp-coders
