> On Jan. 9, 2016, 12:42 a.m., Joris Van Remoortere wrote:
> > 3rdparty/libprocess/3rdparty/stout/include/stout/internal/windows/dirent.hpp,
> >  lines 193-199
> > <https://reviews.apache.org/r/39019/diff/10/?file=1175210#file1175210line193>
> >
> >     `free` is meant to be safe for `NULL/nullptr`.
> >     Why not follow this pattern?
> >     asserting `directory != NULL` seems like it might surprise people?
> 
> Alex Clemmer wrote:
>     Well, you are the C++ programmer, so I'll trust your judgement on the 
> `assert`s here.
>     
>     But, since this question about `assert` comes up 3 times in your review, 
> I'll answer your question, and justify the switch. I initially chose to do it 
> this way because I think internal APIs are easier to reason about if all the 
> functions assume the data in non-`_OUT_` params is fully initialized before 
> you call. I just used `assert` to make this assumption explicit, which I 
> think is a good way of broadcasting to internal users what the contract is.
>     
>     Of course, (1) if you think this is less clear, then my thesis is clearly 
> wrong, and (2) the actual difference in this case is not really a substantial 
> amount of code/complexity, so the cost of switching to your model is not 
> prohibitive, so if I were to advocate staying it would be for purely 
> religious reasons.
>     
>     Therefore, we shall transition to fit your suggestion.

I ended up changing my mind on the other two `assert`s. See below.


- Alex


-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/39019/#review113576
-----------------------------------------------------------


On Dec. 23, 2015, 6:44 a.m., Alex Clemmer wrote:
> 
> -----------------------------------------------------------
> This is an automatically generated e-mail. To reply, visit:
> https://reviews.apache.org/r/39019/
> -----------------------------------------------------------
> 
> (Updated Dec. 23, 2015, 6:44 a.m.)
> 
> 
> Review request for mesos, Alex Naparu, Daniel Pravat, Artem Harutyunyan, 
> Joris Van Remoortere, and Joseph Wu.
> 
> 
> Bugs: MESOS-3441
>     https://issues.apache.org/jira/browse/MESOS-3441
> 
> 
> Repository: mesos
> 
> 
> Description
> -------
> 
> Windows: Added dirent compat code for non-Unix systems.
> 
> 
> Diffs
> -----
> 
>   3rdparty/libprocess/3rdparty/stout/include/Makefile.am 
> b2dea9b93adfa3ea0415a0b5c81a369dd29b6cfe 
>   
> 3rdparty/libprocess/3rdparty/stout/include/stout/internal/windows/dirent.hpp 
> PRE-CREATION 
>   3rdparty/libprocess/3rdparty/stout/include/stout/windows.hpp 
> d46e262e0fd1c2de36f3bf19d8bd693c23bf58cd 
> 
> Diff: https://reviews.apache.org/r/39019/diff/
> 
> 
> Testing
> -------
> 
> `make check` from autotools on Ubuntu 15.
> `make check` from CMake on OS X 10.10.
> Ran `check` project in VS on Windows 10.
> 
> 
> Thanks,
> 
> Alex Clemmer
> 
>

Reply via email to