bug#81639: Please add alias from "ghostty" to "xterm" in faces.el

On 2026-08-18 19:46, Richard Stallman wrote:
> [[[ To any NSA and FBI agents reading my email: please consider    ]]]
> [[[ whether defending the US Constitution against all enemies,     ]]]
> [[[ foreign or domestic, requires you to follow Snowden's example. ]]]
> 
> I know where to find the ncurses developers.  Can you tell me
> how to reach the ghostty develpers?

Ghostty developers mostly do discussions on GitHub and Discord, 
specifically https://github.com/ghostty-org/ghostty and 
https://discord.com/invite/ghostty, respectively.

The motivation for setting TERM to xterm-ghostty is well documented at 
https://github.com/ghostty-org/ghostty/discussions/8268, which can be 
viewed without signing in with GitHub accounts. For people's 
convienience, I've copied pasted the relevant section below:

> After some discussion among the team, we're reversing our position 
> regarding the system terminfo distribution — that is, we won't be 
> waiting for systems to come with Ghostty's terminfo definition 
> preinstalled. People should always use one of the manual or automated 
> ways to install it from their local client, or find some way to install 
> an extra terminfo package (e.g. ghostty-terminfo) from the system 
> package repository if one exists.
> 
> The main reason behind this is that the terminfo authors have 
> deliberately chosen to ship their own version of the terminfo 
> definition under a different name (ghostty instead of xterm-ghostty), 
> with their own modifications that make it substantially different from 
> our own terminfo definition, so it wouldn't even work out-of-the-box 
> like what we had expected. Now, it is understandable from their point 
> of view as custodians of the wider terminfo database, since Ghostty is 
> not de jure a xterm variant, so it shouldn't receive the xterm- prefix. 
> In practice, however, we've found out that a lot of legacy programs 
> rely on TERM containing the xterm substring to enable a lot of features 
> that originated with xterm (extended colors support, most OSCs, etc), 
> while ignoring terminfo completely. We, much like other innovative 
> terminal emulators like Kitty, found it easier to just include the 
> xterm- prefix out of de facto pragmatism, diverging from the philosophy 
> of the terminfo maintainers.
> 
> If you feel like this is unreasonable/dumb/etc. and none of this should 
> be the way this is, we don't blame you. Terminal emulators have had a 
> long history and with it come generations of technical cruft. You're 
> more than welcome to simply set TERM=xterm-256color and move on with 
> your life. What I am trying to say is that this is a problem not just 
> isolated to Ghostty, and the only true way to "fix" this once and for 
> all is to reduce dependency on TERM itself by using alternatives like 
> XTGETTCAP or capability-based queries at runtime. For now there are 
> only workarounds, automated or not.

They're correct that setting TERM to xterm-ghostty as a pracitcal 
tradeoff generally makes things better for users, specifically because 
ghostty is trying to be xterm behavior compatible.  For example, Emacs 
behaves more correct when TERM is set to xterm-ghostty specifically 
because Emacs then assumes that the terminal is an xterm variant, using 
term/xterm.el for initialization.

   -- MJF