[PATCH] color: enabled color support for export command (issue1507)

Ankur Dahiya ankurd at fb.com
Mon Sep 3 03:01:26 CDT 2012


I think if there exists a 'color' option, it should be consistently supported across all commands. That's the most intuitive user experience.
As regards to the 'TTY detection' risk, a similar risk exists for hg diff and all other commands which support color. If a detection bug surfaces in the future, scripts will break and we will have to fix it!

Regards
Ankur dahiya

On 3 Sep 2012, at 06:22, "Gilles Moris" <gilles.moris at free.fr> wrote:

> On Sunday 02 September 2012 03:29:32 pm Martin Geisler wrote:
>> I think we need to decide if we trust the TTY detection or not. Being
>> afraid of unclear risks is not constructive -- either we have a bug or
>> we don't.
> 
> This is not only about now, it is also about the future. How can you ensure 
> that there will not be a regression introduced in a month or a year?
> 
> Activating the color in export introduce a possible dormant error, that can 
> not be present if color is not present by design.
> 
> This is why I say it introduces a "risk" for something we already have a 
> workaround for.
> 
> Regards.
> Gilles.


More information about the Mercurial-devel mailing list