On Wed, Dec 05, 2001 at 04:13:18PM +0100, V. T. Mueller wrote:
> On Wed, 5 Dec 2001, Ollie Cook wrote:
> > I understand you objection to the first suggestion I made, but the
> > second one gives you the same redundancy, without the waste. I may be
> > missing what you're getting at though, so could you explain it a
> > different way?
>
> Well, ok. I have a very pragmatic opinion regarding software syntax
> changes.
There's absolutely nothing wrong with that in my opinion, and I
understand your motivation for not wanting to see a change in syntax.
In that case, perhaps there could be a 'mysql_servers_new_format' for
those that want to use it; there are ways around the problems that
you can foresee.
> I really don't think that Philip's possible implementation of the
> feature you requested would provoke him to break anything. It would,
> however provoke either failures or extra cost/effort on a certain
> percentage of sites that have to deal with different syntax.
If not use a new configuration directive, as I mentioned above, I'm sure
that Exim could be intelligent enough to notice that if the configuration
doesn't contain the new 'identification', that the old meaning will be used,
at least for a few releases, to give people the opportunity to get themselves
up to date.
On the other hand, there have been occasions when Exim's behaviour has
changed (output of -bv changing, or require_files, for example) or defaults
changing between releases, so it's not as if a change of syntax is much
different to that - unexpected things may happen if you don't read the
ChangeLog.
The transition between 3->4 would be the perfect opportunity to change the
syntax, since so much else is changing as well. Although I have a feeling
that Philip has "locked out" changes for the initial 4 release.
> For example: we'd have to changes all our internal documentation and
> backup configs. You state that you'd rather see a little tweak in software
> to be done instead of buying hardware. On the other side you say that
> you're of course going to do thoroughly implementing and testing the new
> syntax. And I'll bet any amount that, all in all, the second sums up to
> be more expensive.
It's not the expense that's the issue - it just seems to me, personally,
that Exim could be improved to behave more 'nicely' by not wasting time
doing things that the administrators know are going to fail.
I can go ahead and achieve my task without the change to Exim. But during
testing I realised the way that Exim is treating the multiple database
servers, and thought I'd suggest a way that it could be improved.
I'm sure the our database servers will be able to cope with the erroneous
queries - but what's the point in making a server handle things which there
is no need for it to waste resources on?
Anyway - I've made my case, so it's now up to Philip to look into it
if he wants to.
Yours,
Ollie
--
Oliver Cook Systems Administrator, ClaraNET
ollie@??? 020 7903 3065
