Re: [boost] [xint] Fourth release, requesting preliminary review again

Subject: Re: [boost] [xint] Fourth release, requesting preliminary review again
From: Chad Nelson (chad.thecomfychair_at_[hidden])
Date: 2010-06-06 10:10:32

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

On 06/06/2010 04:14 AM, Christopher Jefferson wrote:

>> That was the idea -- to implement nothrow_integer without using
>> any exceptions at all.
> 
> Mant libraries, including libstdc++ (the standard library in g++)
> implement support for -fno-exceptions by just calling abort() instead
> of throw. I would advise just doing that, as from my experience
> anything else turns into an unreadable mess very quickly. It is also
> reasonable to make the function called user-defined and allow
> longjmping out, under the condition memory will leak, so users should
> aim to clean up and exit, not get exceptions repeatedly.

- From my testing, that looks like the only reasonable solution. It looks
like -fno-exceptions disables throw, try, and catch -- you can't even
compile a program containing any of those statements, and putting them
in a template class or function that's never instantiated doesn't
prevent that.

Can I assume (or insist) that BOOST_NO_EXCEPTIONS will be defined if the
user wants that behavior?
- -- 
Chad Nelson
Oak Circle Software, Inc.
*
*
*
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.10 (GNU/Linux)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org/

iEYEARECAAYFAkwLrFUACgkQp9x9jeZ9/wQPdwCfTS24h3iGs4YAGoCwmI/g5BVu
qOQAoOR3CyMJ6ApLoa4ovaknnfoXCWsy
=o4Yk
-----END PGP SIGNATURE-----