> On Oct. 20, 2014, 10:56 a.m., Niklas Nielsen wrote:
> > 3rdparty/libprocess/3rdparty/stout/include/stout/os.hpp, line 1283
> > <https://reviews.apache.org/r/26903/diff/2/?file=725229#file725229line1283>
> >
> >     How about flatten it to os::libraries::expand() ?
> 
> Kapil Arya wrote:
>     It's not clear what expand() would mean but expandName() is clearer.  
> expandLibraryName() would have been the best, though.

library is already a part of the namespace, so that would become redundant?
I mostly suggested it because you won't have other conflicting 
expand(std::string name) in that namespace and it would be more concise.

Anyway, it is not a biggy; feel free to drop the issue if you want to keep 
expandName


> On Oct. 20, 2014, 10:56 a.m., Niklas Nielsen wrote:
> > 3rdparty/libprocess/3rdparty/stout/include/stout/os.hpp, line 1271
> > <https://reviews.apache.org/r/26903/diff/2/?file=725229#file725229line1271>
> >
> >     I looked through your patch set - it doesn't look like you use this 
> > function anywhere?
> >     
> >     It is a bit unclear to use 'ld library path' and 'path' 
> > interchangeably. How about sticking to 'path' or 'paths' (it is a path list 
> > right?)
> >     
> >     For example:
> >     
> >     // returns ':' separated list of library paths.
> >     os::libraries::paths()
> >     
> >     // Resets library variable to ':' separated library list.
> >     os::libraries::setPaths()
> >     
> >     // Appends library path to library list.
> >     os::libraries::addPath()
> >     
> >     Let's think about the path() overload for a little bit. I think for 
> > getting and setting, we should call out explicitly what we do (
> >     
> > http://google-styleguide.googlecode.com/svn/trunk/cppguide.html#Function_Overloading).
> 
> Kapil Arya wrote:
>     We use ldLibraryPath() to get the corresponding environment variable 
> string ("LD_LIBRARY_PATH"/"DYLD_LIBRARY_PATH").
>     
>     I like the paths(), setPaths() and addPaths() function names and will 
> update the diff accordingly.

Sure - but are you using it anywhere else? We expose two kinds of 'paths' - a 
ldLibraryPath (which is not a path but a env variable name) and 'path' which is 
a ':' separated directory list.
You should be able to hide ldLibraryPath with your new abstraction, so that 
would make the api cleaner IMHO.

Ben, what is your take?

I am not strongly attached to this, so if you think they are distinct enough - 
feel free to drop.


- Niklas


-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
https://reviews.apache.org/r/26903/#review57372
-----------------------------------------------------------


On Oct. 18, 2014, 12:04 p.m., Kapil Arya wrote:
> 
> -----------------------------------------------------------
> This is an automatically generated e-mail. To reply, visit:
> https://reviews.apache.org/r/26903/
> -----------------------------------------------------------
> 
> (Updated Oct. 18, 2014, 12:04 p.m.)
> 
> 
> Review request for mesos, Benjamin Hindman and Niklas Nielsen.
> 
> 
> Repository: mesos-git
> 
> 
> Description
> -------
> 
> Also added a utility function to expand a library name to a filename
> by prefixing "lib" and adding the default extension.
> 
> 
> Diffs
> -----
> 
>   3rdparty/libprocess/3rdparty/stout/include/stout/os.hpp 
> 63bda7a87e24e1696f3e7bb0206c551ce39dfe27 
> 
> Diff: https://reviews.apache.org/r/26903/diff/
> 
> 
> Testing
> -------
> 
> Ran make check.
> 
> 
> Thanks,
> 
> Kapil Arya
> 
>

Reply via email to