From: Jeff Garland (jeff_at_[hidden])
Date: 2004-01-12 11:53:05
On Sun, 11 Jan 2004 21:02:17 -0800, Stephan T. Lavavej wrote
> > Also, I think you are probably aware of some [shadow
> > warnings] in tokenizer which is included by date_time.
> I encountered one shadow warning in tokenizer which I pointed out, after
> which it was promptly fixed.
> I am working on a codebase that uses date_time, and I am beginning
> to use regex. Whenever I encounter warnings I post them to the
> Boost list. So, there are no warnings I am aware of that I haven't
> mentioned, but there are probably lurking warnings that my code
> doesn't trigger yet.
Well, I'm running the entire date_time test suite which instantiates all the
various templates and I can assure you there are others. I'll try post some
of the non date-time ones later...
> > but there are several others I haven't addressed yet. I'll
> > try and gradually clear these out as I have time...
> That would be great. It would be best for me as a user if all the shadow
> and unused parameter warnings were cleaned out before 1.31.0. I
> would be dismayed to encounter warnings with the final version of
> 1.31.0 and have to turn back to Boost CVS again. (Of course, it'll
> probably happen ANYWAYS, but I can dream.)
While I agree with the sentiment I think the time for this effort is long
past with 1.31.0. The release should be imminent and we don't want to break
things just to fix a 'meaningless' warning. What I mean by 'meaningless' is
that the code executes and passes tests which is the true test (I'm not
suggesting we ignore the warning, it just isn't helpful in this case)...
Boost list run by bdawes at acm.org, gregod at cs.rpi.edu, cpdaniel at pacbell.net, john at johnmaddock.co.uk