Boost logo

Boost :

Subject: Re: [boost] [accumulators] broken on msvc
From: Edward Diener (eldiener_at_[hidden])
Date: 2014-06-16 23:55:45


On 6/16/2014 12:33 PM, Michael Caisse wrote:
> On 06/16/2014 09:23 AM, Edward Diener wrote:
>> On 6/16/2014 9:11 AM, Edward Diener wrote:
>>> On 6/16/2014 2:38 AM, Eric Niebler wrote:
>>>> On 6/15/2014 10:36 PM, Eric Niebler wrote:
>>>>> On 6/15/2014 2:20 PM, Andrey Semashev wrote:
>>>>>> On Sunday 15 June 2014 21:18:12 Jürgen Hunold wrote:
>>>>>>> The question remains: Who fixes this where?
>>>>>>
>>>>>> As far as I can see, is_empty.hpp does not use BOOST_PP_CAT when it
>>>>>> doesn't
>>>>>> include it, so the header is correct. Accumulators should include
>>>>>> cat.hpp
>>>>>> whenever it uses BOOST_PP_CAT, so at least accumulators_fwd.hpp has
>>>>>> a bug.
>>>>>
>>>>> I've added the missing #include on develop. Hopefully, that fixes
>>>>> things.
>>>>
>>>> Looks like it doesn't fix things. If Edward's is indeed the problematic
>>>> commit, I wonder if he'd be willing to look into this. Edward?
>>>>
>>>> This is the commit:
>>>>
>>>> https://github.com/boostorg/preprocessor/commit/c66ea5871fd44274f681f02323c4d46bdf394c7c
>>>>
>>>>
>>>>
>>>> *All* msvc accumulators tests are currently broken on develop by this.
>>>
>>> I have backed out all the changes I made to preprocessor so the
>>> accumulators problem should now be fixed. I will look at the issue of
>>> why the macro was failing with my VC++ changes when I have some time.
>>
>> I have not been able to do this correctly. I can do a 'git reset --hard
>> abranch' to the local branch I want to back out 'develop' to but as soon
>> as I try to push this to the remote develop I get:
>>
>> hint: Updates were rejected because the tip of your current branch is
>> behind
>> hint: its remote counterpart. Integrate the remote changes (e.g.
>> hint: 'git pull ...') before pushing again.
>> hint: See the 'Note about fast-forwards' in 'git push --help' for details.
>>
>> Is there some way I can back out the changes I have already pushed to
>> the remote develop without having to manually back out each change
>> manually on the latest 'develop' branch before I push it to the remote
>> develop ?
>>
>>
>>
>
> You aren't going to want to use reset on something you have pushed
> remotely. Take a look at "git revert". It will generate the changes
> required to back-out 1:N commits into a single commit. History is
> preserved and life is good.
>
> The command is quite powerful and you should be able to do whatever is
> needed to get the commit into the state you want before pushing.

Thank you ! You are very right.


Boost list run by bdawes at acm.org, gregod at cs.rpi.edu, cpdaniel at pacbell.net, john at johnmaddock.co.uk