> On 2011-03-02 09:21:38, Steve Reinhardt wrote:
> > I'm willing to be convinced on this one, but I'd argue that the 
> > inconsistency is that VERBOSE should be lowercase, not that update_ref 
> > should be uppercase... I think I originally made update_ref lowercase to 
> > distinguish it from the sticky args that control the build itself, where 
> > the non-sticky vars are more like --foo command-line options (conceptually 
> > if not syntactically).
> > 
> > Just some historical perspective from the guy who introduced update_ref.
> 
> Gabe Black wrote:
>     Since I originally made this patch I ran into the "default" scons 
> variable which sets what file in build_opts to draw from if you're using some 
> new name that's not already in there. This is inconsistent too, but then it 
> seemed like making it capital might be confusing since it's selecting what to 
> use to set variables -except- itself and potentially others. It's not a huge 
> inconsistency, but when I was thinking about it it felt like something 
> somebody could trip over.
>     
>     I also thought about the syntactical inconsistency between VERBOSE and 
> --color and --no-color. Both are changing something about how the build 
> output looks, both are transient, but one is a scons argument and the other 
> is a command line option. I remembered somebody saying scons already had a 
> --verbose command line option, but looking for it just now I didn't see it.
>     
>     So how about this. All non-sticky options should be in the form of 
> command line options in lower case. That would be default, color, no-color, 
> update_ref, and verbose as far as I can tell, but default was hiding from me 
> for a while so there may be some other one lurking in there. Then all the 
> sticky options will be CAPITALIZED and left as scons arguments. That seems to 
> be consistent as far as capitalization, and input mechanism relative to 
> stickiness.
> 
> Steve Reinhardt wrote:
>     So you're suggesting converting them to real options, like --verbose?  
> That makes a lot of sense to me.  And while we're on a consistency kick, it 
> should probably be --update-ref rather than --update_ref.

Ok, great. I'll make these changes and make it --update-ref instead of 
--update_ref. I'll update the diff once I get a chance.


- Gabe


-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
http://reviews.m5sim.org/r/522/#review922
-----------------------------------------------------------


On 2011-02-28 04:54:41, Gabe Black wrote:
> 
> -----------------------------------------------------------
> This is an automatically generated e-mail. To reply, visit:
> http://reviews.m5sim.org/r/522/
> -----------------------------------------------------------
> 
> (Updated 2011-02-28 04:54:41)
> 
> 
> Review request for Default, Ali Saidi, Gabe Black, Steve Reinhardt, and 
> Nathan Binkert.
> 
> 
> Summary
> -------
> 
> SCons: Clean up some inconsistent capitalization in scons options.
> 
> 
> Diffs
> -----
> 
>   SConstruct c6ba129c2764 
>   src/mem/protocol/SConsopts c6ba129c2764 
>   tests/SConscript c6ba129c2764 
> 
> Diff: http://reviews.m5sim.org/r/522/diff
> 
> 
> Testing
> -------
> 
> 
> Thanks,
> 
> Gabe
> 
>

_______________________________________________
m5-dev mailing list
m5-dev@m5sim.org
http://m5sim.org/mailman/listinfo/m5-dev

Reply via email to