Re: -getRed:green:blue:alpha: not defined for the NSColor

[Keith David Bershatsky]

Thank you, Paul, you hit the nail on the head with your first educated guess.  
The problem is present with 9624f609493da7c08016ba00d6895bad0fe26a0e, and the 
problem disappears when reverting to the previous commit 
cb367c8e0d4a881acf8255e1ff6bdd6442efb538.  You saved me a lot of time -- 
greatly appreciated!!!

I will reset to *26a0e and revert each component of that commit one by one in 
an effort to isolate the problem.  When I have an update on my progress, I'll 
post another comment to the mailing list -- it may require a good night's sleep 
though.

;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;

> Date: [07-12-2019 16:56:23] <12 Jul 2019 16:56:23 -0700>
> From: Paul Eggert <address@hidden>
> To: Keith David Bershatsky <address@hidden>
> Cc: Emacs Devel <address@hidden>
> Subject: Re: -getRed:green:blue:alpha: not defined for the NSColor
> 
> Keith David Bershatsky wrote:
> 
> > it would likely take me a very long time to locate the commit by trial and 
> > error; e.g., building various versions of Emacs from 04/08/2019 to the 
> > present until the culprit is located.
> 
> You can use binary search, so this is only O(log N).
> 
> You could also take a look at commits 2019-04-02T20:49:36Z!address@hidden
> (9624f609493da7c08016ba00d6895bad0fe26a0e) and the flurry of commits near
> 2019-04-14T02:41:31!address@hidden
> (5d8b0fadeec373cd2861328aeb1cb4293cbc9ded) as they seem relevant. But for 
> this sort of thing I've found binary search is often best.