Aw damn, thanks for the information though!

screen is about to have truecolor/direct color support, preview like this:
gnu screen 4.99 master (screen 5.xxx has not been released yet)
compile with bool truecolor=1 instead of 0
~/.screenrc truecolor on (did not find where truecolor is set in code at this point)
TERM=xterm-256color
COLORTERM=24bit

blink, overline and undercurl did not work as applied with test in khamer's comment above. (bash)

awk rainbow test works with dash and bash.

I think weston-terminal does not support true color.

It still doesn't in version 8.0.0 that was released this year, unfortunately. Worse, it doesn't parse and ignore \e[38;2;<R>;<G>;<B>m sequences (or map them to the allowed color space like Linux terminal does) but sees everything after 2 as a new SGR parameter.

Code here: https://github.com/wayland-project/weston/blob/master/clients/terminal.c#L1581

IOS terminal apps that support true color:

• Secure ShellFish (my favourite)
• Blink Shell

IOS terminal apps that do NOT support true color:

• Prompt (a.k.a. "Prompt 2")

mintty: delimiter semicolon or colon; also supporting CMY and CMYK colour models

Right now have been tested on KDE Plasma's Konsole, and it supports true color.

cmd.exe, Windows PowerShell, and PowerShell Core are using same Console. These apps are just shells, the actual console window is conhost.exe.

Thank you very much for sharing this @XVilka, it's really helpful ! 👍

Would you consider adding delta (a syntax-highlighting diff-viewer/-formatter) to the list of apps supporting 24bit colors ?

I would to add mine in this list too.

A small nit: please remove the word Introducer from the following:

… it is possible to use both ';' and ':' as Control Sequence Introducer delimiters

A Control Sequence Introducer is just the initial 1B 5B or 9B bytes of a whole Control Sequence. It is the latter which may contain delimiters.

(Hide this comment once that's done.)

Another nit: "delimiter" is related to "limit" rather than "meter", and is spelled accordingly. (This seems to have started with commit 507d2c2.)

Or just pull the changes from my repo clone of this gist.

You might also like to pull patches from the forks by CaninoDev, JohnMorales, and tizee.

(Although there doesn't seem to be a way to do this using the gist web interface, it's certainly possible using a local git repo on your own machine: you can pull from multiple gists as "remotes", and you can push to your own gist with changes merged from someone else's gist. See the instructions for cloning a gist as a repo for details.)

(As before, hide this comment once that's done.)

Edex-UI seems to support TrueColor - oculd it be added to the list of supported terminals?
https://github.com/GitSquared/edex-ui

It's kinda a joke terminal emulator, but it's still managed to do it

urxvt Debian users may want to manually downgrade rxvt-unicode to 9.22-5 as it contains 24bit color patch. It was reverted in next version as it caused bugs as mentioned in 9.22-6 changelog. So use it with caution:

rxvt-unicode (9.22-6) unstable; urgency=medium

• Revert the 24bit colour patch. Though no issues seem to arise when using
  the default resource values, it seems to cause many issues if changes are
  made to resources related to colour.
  (Closes: #922268, #922289, #922297, #922298, #922299)

-- Ryan Kavanagh [email protected] Thu, 14 Feb 2019 12:00:54 -0500

@anpavlov The original patch to rxvt was pretty buggy:

• It crashes when run as urxvt -pe tabbed
• It doesn't cope with the terminal being in reverse video mode (it hides the cursor)
• It introduces several other cosmetic defects

The first item on its own is a good enough reason to revert the patch. Clearly there needs to be more work done before it's ready for inclusion in a large-scale distribution.

It seems this gist has become moribund, as @XVilka has neither commented nor applied any patches since October 2019.

It's become big enough that it really ought to be an actual project that accepts pull requests and has a bug tracker.

To that end, I suggest any further follow-up occur at https://github.com/kurahaupo/about-terminals/commits/kurahaupo/TrueColour.md

I'm happy to grant push access as required, especially to XVilka if they return.

@kurahaupo: I made repo some time ago - feel free to send PRs, issues, etc https://github.com/termstandard/colors

@XVilka Great, I'm glad not to have to maintain my own.

(And this also explains why things have been so quiet here.)

People will keep coming here because this gist has 2200 stars and 147 forks - twenty times more links than your repo - so search results are naturally going to rank it higher than your repo.

I searched for the URL you just posted and can now see that it's referenced in the top line "Most updated version is always available", which is great, but it's very easy to miss or ignore: at first glance it just looks like the kind of link that's on every GitHub.io website, that simply points back that the very page you're already reading; it's not obvious that it's pointing at a different repo without carefully reading the URL. (Indeed it's so unobvious that I went ahead and made my own repo, thinking that none existed.)

This gist also hundreds of comments, so it's not easy to find the other links to your repo either.

To make sure that people do actually go to your new repo, instead of commenting here, I suggest starting with something like:

THIS GIST IS OBSOLETE

See termstandard/colors instead.

And to make it even more obvious, remove all the other content.

I updated the repository, thanks for the @kurahaupo's PR, also syncronized this gist with the repository, and added the link.

xterm no-longer approximates truecolor rgb to a palette, afaict.

About Querying the Terminal: the response should actually look like ...;48:2::1:2:3m - note the double colon which allows a position for the optional color space value.

I want to know why MacOS built-in terminal not support true color?

echo -ne \e[1;3;4:3;5;53;38;2;255;127;0;58;2;0;48;255;48;2;255;0;{0..255..8}mX \e[0m\n

windows terminal was not a fan of this one

@khamer @JoshMerlino what is the 4:3 supposed to do in the middle of that escape sequence, before the 38 and 48 "extended" colour codes?

@mintty @wendajiang @JoshMerlino - please note that this gist is no longer maintained. Please add your comments to the standardterm repo instead.

@XVilka - this is why I suggested that the entire content of this gist should be deleted except for the link to the repo. It would also help to mark all the comments as "hidden".

After my confirmation. Xshell6/7 support truecolor. But you must set "Tools-Options..-Advanced" check the "Use true color*"

@coolmian would you mind please adding your notes as an issue on the standardterm repo ? (Preferably create a PR with a fix.)

Reminder to future readers who want to make a comment but didn't notice the warning at the top of this page.

It's now a full-blown git repo so please write your comments there, or clone the repo and suggest changes via a Pull Request:

git clone git+ssh://github.com/termstandard/colors truecolor

THIS GIST HAS MOVED
so any comments made here will be ignored!

@kurahaupo yes, should have made this a long time ago. Now it's done.

@coolmian would you mind please adding your notes as an issue on the standardterm repo ? (Preferably create a PR with a fix.)

OK, I've done it now

Just repeating this so that it's visible at the bottom of this conversation; please don't reply - the whole point is that this should be the last comment, to make it plainly visible when someone is about to hit "reply".

Reminder to future readers who want to make a comment but didn't notice the warning at the top of this page.

It's now a full-blown git repo so please write your comments there, or clone the repo and suggest changes via a Pull Request:

git clone git+ssh://github.com/termstandard/colors truecolor

THIS GIST HAS MOVED
so any comments made here will be ignored!