Re: [PATCH v2 4/6] gitk: use more natural language for labels of color preferences

From: Mark Levedahl <mlevedahl@gmail.com>
Date: Aug 13, 2026, 20:11 UTC
Message-ID: <6da65615-0870-402d-8630-9b486b512947@gmail.com>
In-Reply-To: <d6754ded-89f6-48e9-9755-24c45c90845f@kdbg.org>

On 8/13/26 3:35 PM, Johannes Sixt wrote:

Show 17 quoted lines

> Am 13.08.26 um 19:00 schrieb Mark Levedahl:
>> On 8/12/26 9:16 AM, Johannes Sixt via GitGitGadget wrote:
>>>          markbg      markbgcolor {} \
>>> -                    [mc "Marked line bg"] \
>>> +                    [mc "Marked line background"] \
>>>                      [mc "marked line background"] \
>>>
>> Not a new issue, but "marked line" is not a good description. What this actually controls
>> is the background color of commit message text found by search criteria in the gui, not
>> lines of text, and (as far as I can tell) never any text found in the patch text.
> Actually, it is the background color of lines found via "Show origin of
> this line" in the diff panel.
>
> When reading this code for the first time, I thought it has to do with
> the "Mark this commit" in the commit list, but it doesn't.
>
> -- Hannes

Sorry, yes you are right, I confused myself. 
- Maybe "Marked line background"  should be something including "origin" hinting at the
menu item triggering it, e.g., "Origin line background"?

The highlight color associated with the problem I noted (gitk highlights text found in the commit message, but not in the commits), is "foundbgcolor" in the config file, and is not on the config page. There are many colors in the config file but not on the config page: as long as no one is complaining, I'd leave that as-is.

Mark