From: Neal D. Becker (ndbecker2_at_[hidden])
Date: 2004-05-06 10:41:48
Vladimir Prus wrote:
> Neal D. Becker wrote:
>> 1. Is it possible to expand --help so that it also gives the default
>> of options? Obviously, it's not likely to tell defaults that it doesn't
>> no about, but at least the ones specified by default_value()?
> I need to think about it. The biggest issue I see that that it would mean
> that if you have option of type T then T would be required to have
> operator<<, which might be too severe limitation.
>> 2. What was the resolution to the earlier problem of confusing a negative
>> parameter value with an option?
> I think I've fixed that. See the attached.
> - Volodya
It's a small point, but here's my opinion. You say something like: "If
option requires a value and following -<number> doesn't match an option,
then parse it as an argument".
I would suggest "If option requires a value then parse *any* following
string, including -<number> as a value, unconditionally".
Boost list run by bdawes at acm.org, gregod at cs.rpi.edu, cpdaniel at pacbell.net, john at johnmaddock.co.uk