|
Boost : |
Subject: Re: [boost] [tti] Moving TTI to release
From: Edward Diener (eldiener_at_[hidden])
Date: 2013-05-19 14:17:37
On 5/19/2013 1:45 PM, Daniel James wrote:
> On 19 May 2013 18:08, Edward Diener <eldiener_at_[hidden]> wrote:
>> Do I have to do anything further to move TTI from the trunk to the release
>> branch for Boost 1.54 ?
>>
>> If so, please inform me of it as I am happy that TTI can be part of the 1.54
>> release, and I believe that the cutoff date for new libraries is tomorrow.
>
> Go ahead. There are a few test failures, but nothing that should stop you.
What is the normal procedure of moving from 'trunk' to 'release' ? Is it
merely adding the files to the release branch or is there a better
subversion technique ( I guess I should already know this ) ? Is the
release branch 'https://svn.boost.org/svn/boost/branches/release' ?
>
> If you don't think you can fix the vacpp failures, you could mark them
> up in status/explicit-failures-markup.xml.
The vacpp failures do not trace the error call stack any further than
the use of function.types by TTI. I also notice that function.types has
failures for vacpp, so I expect it is related. So I think I should mark
them in status/explicit-failures-markup.xml. Unfortunately I don't have
a computer running AIX to test this out.
>
> As far as I can tell from the config results, the Visual C++ 9 tester
> that has odd template template errors is running an older version of
> Visual C++ 9 (teeks99-03b-win7-64on64's version is 150021022, DrWeb's
> version is 150030729). I think we can only really support fully up to
> date versions, and the difference can't be marked up, so if that can't
> be fixed, there's nothing to do there. Not being a windows programmer,
> I don't really know much about the various Visual C++ versions though.
I did try a syntax notation update for those VC9 failures in the test
results, which simply affects the structures I created for the tests,
and not the TTI code itself. It is in revision 84337 on the trunk, but
none of those VC9 tests have cycled to use that revision yet. In my own
tests for VC9 I am not seeing the failures in any situation.
Once I move to release do I assume that any coding changes are forbidden
but that documentation changes are allowed ?
Thanks very much for your help.
Edward Diener
Boost list run by bdawes at acm.org, gregod at cs.rpi.edu, cpdaniel at pacbell.net, john at johnmaddock.co.uk