[PATCH] Fix string to integer conversion

[Jeremy Allison]

On Fri, Jan 25, 2019 at 07:20:12PM +0100, swen wrote:
> > 
> > Yep, and that's why I'm going NAK on this.
> > I showed you the right way to do this, even
> > wrote the wrapper for you, and you decided
> > to ignore that advice. 
> Hmmm, not sure if this is worded ... really fair !?

It's accurate from my point of view. I told
you I have strong opinions on the right way
to do this.

> Oh, ok. 
> I interpreted your position on this a bit more passive I guess
> as in "I won't support but won't block either".
> 
> As it stands I would have needed 2 supporters anyway,
> with or without your NAK, therefore I was a bit surprised by 
> you emphasizing it.
> 
> The chances were small to get it in.
> But now, I'm afraid, they're pretty much zero.
> 
> Anyhow, as we both stated already, not important enough to bikeshed.

Well you could always just rework the code
to do it the 'right' (by my definition of course :-)
way. That's part of the collaborative process,
learning to work with others. As I said, I
have strong opinions on the correct API for
this. If you really think I'm wrong, you
can get two other engineers to agree with
you.

We're already bikeshedding this, so we might
as well finish the paint job :-).

Doing this the right way would have expended
much less effort than this argument. I guess
I don't understand why you're adamant that
doing a half-assed job of fixing the code
(by explicitly messing with errno inline)
is better than using a wrapper function
that does it correctly (and avoids multiple
direct manipulation of errno). Remember,
errno is a global variable - messing with
it is just nasty.

Learning to work with others, even if you
disagree, will stand you in good stead in
the future, more than just abandoning
something as 'too hard to get in'.

Jeremy.