Subject: Re: [Boost-build] Python port development
From: Steven Watanabe (watanabesj_at_[hidden])
Date: 2010-07-09 16:14:56
>> From: Steven Watanabe <watanabesj_at_[hidden]>
>> How is this particularly tricky? I can see that it would get tedious
>> is the size is not known to be one of a few possibility (thus requiring
>> a binary search), but it seems like this is rather the obvious solution
>> when running a compiled program isn't possible.
> It is not tricky when you know the solution :-).
> But as you can see neither your nor Gevorg, did this right.
Well I actually thought of both solutions and dismissed this
one as more difficult and a waste of resources. IMHO, the
most important part is realizing the problem with running
a compiled program.
> BTW: why do you need there *2? In order to get negative value as some
> compilers accept 0 as valid array size.
It would probably be clearer to use
char test[(sizeof(wchar_t) >= 4)? 1 : -1];
Boost-Build list run by bdawes at acm.org, david.abrahams at rcn.com, gregod at cs.rpi.edu, cpdaniel at pacbell.net, john at johnmaddock.co.uk