Re: [boost] [utf8_codecvt_facet]

Subject: Re: [boost] [utf8_codecvt_facet]
From: Robert Ramey (ramey_at_[hidden])
Date: 2013-01-06 13:24:17

Marshall Clow wrote:
> On Jan 5, 2013, at 10:53 PM, Robert Ramey <ramey_at_[hidden]> wrote:
>
>> The change made in the trunk addresses an issue which has been
>
>> in the code for years and only produces a warning.
>
> For people who build with "warning as errors', this is a showstopper.
>
>> I would prefer that we take our time and address it in a more
>> definitive manner. To try to fix it on short notice at this point
>> isn't worth the risk
>> of imposing a setback on the release.
>
> This is not "short notice";  I've been trying to get this bug fixed
> for over a year.
> I first proposed fixing this for 1.49.
>
> This is now an issue for Visual Studio users as well, since they have
> changed their codecvt to be standards compliant in VS 2012. Boost is
> now in a position where we have a slower reaction time than Microsoft
> on this issue.

I'm not disputing any of the above.  I'm just not convinced that this way
of addressing the issue will actually address the source of the problem
which I believe is more subtle than it might first appear.

Right now we have  a simple choice - merge in the change which
eliminates the warning but trips some error on one of the tests
of the serialization library or don't do it which leaves the warning
and permits the serialization library test to pass.

I'm actually indifferent to which decision everyone wants to make.
>From my view either option leaves the utf8_codecvt where
it has some unknown behavior.

I DO think it's a bad idea just to tweak the serialization library
test just so it passes the test.  I would prefer to consider the
warning at face value.  This warning indicates that we have
named a function in utf8_codecvt which is also implemented
in the base class.  This in turn means we should actually
look at the base class header and determine what is really
going on.  Given the fact that there is a lot of variation
among libraries - this requires some time and effort.

Robert Ramey

>
> -- Marshall
>
> Marshall Clow     Idio Software   <mailto:mclow.lists_at_[hidden]>
>
> A.D. 1517: Martin Luther nails his 95 Theses to the church door and
>        is promptly moderated down to (-1, Flamebait). -- Yu Suzuki
>
>
> _______________________________________________
> Unsubscribe & other changes:
> http://lists.boost.org/mailman/listinfo.cgi/boost