Apply scrollback reflow patch ?

[ShellCode33]

Hey, thanks for your work ! Love the custom dmenu stuff. Would you mind applying the scrollback reflow patch ?

Apply the following patch on top of the original scrollback patch (=>0.8.5) to allow column and row reflow.

Very useful when you resize windows or switch to fullscreen.

Take care

[LukeSmithxyz]

How is this any different from the current functionality?

[ShellCode33]

Let's say you have a small window in which you perform a command with long lines :

For whatever reason you change the layout of your WM, or you resize the window, or you go fullscreen with that window.
The content's width of that window will remain the same :

The otherway arround is more problematic because some part of the content is not shown anymore :

becomes

Expected behavior for

would be to become

This is a poor example but you get my point. Maybe this is the behaviour you want though.

[LukeSmithxyz]

Okay. Yeah I'll try to add it, but looks like there are a bunch of little incompatibilities with other patches I or someone else will have to figure our first. I'll get to it if I can.

[ShellCode33]

Yep, many incompatibilities indeed. At first I was willing to submit a PR but I gave up when I saw the amount of lines in the rejected diff. Ended up doing my own fork of st, which I'm pretty sure saved me some time. Even for you I wonder if it wouldn't be faster to reapply all your patches from scratch using a clean base of st 0.9. Many of the fixes you have here on this repo are probably also fixed upstream and some others are now available as patches (such as glyph_wide_support).

Anyway, good luck to anyone reading this and willing to put in the effort

[LukeSmithxyz]

The main issue is probably getting this, boxdraw and ligatures all to work in one place. I'll try to repatch from scratch, but I think it will have to grapple with this same problem.

[ShellCode33]

I feel like the boxdraw patch is not worth it if you're already using a "good" font. I use Fantasque Sans Mono for my terminal, and it looks pretty much the same with and without boxdraw. At least the difference is not noticeable enough to me to add boxdraw. I added NotoSansMono as comparison, it doesn't render well indeed. It's up to you of course but for me I choose to rely on Fantasque Sans Mono instead of messing up my code even more using boxdraw. Maybe you should consider getting rid of it.

NotoSansMono font on my fork of st, without boxdraw (pretty ugly I agree) :

NotoSansMono font on your fork of st, with boxdraw :

Fantasque Sans Mono on my fork of st, without boxdraw (pretty good right?) :

Fantasque Sans Mono on your fork of st, with boxdraw :

[LukeSmithxyz]

Yeah it's not that bad. I had added boxdraw from public outcry, but if it's either or, I might go with this instead due to the wider functionality.

[speedie1337]

The original redraw patch that was mentioned was suggested in the st-flexipatch repository but never added because it is so massive and conflicts with a lot of patches.

There is a much smaller patch that achieves nearly the same functionality. See this patch. I have this in my own build and it works great!

Only functionality this version seems to miss is reflow for the top, but text does seem to redraw for example when you're changing mfact in dwm.

This version written by nimaipatel does not include scrollback but it can be applied on top of scrollback without issues!

The main issue is probably getting this, boxdraw and ligatures all to work in one place. I'll try to repatch from scratch, but I think it will have to grapple with this same problem.

Patch boxdraw first and apply the boxdraw version of ligatures on top.

The only difference between the original diff and the boxdraw specific diff is

diff -up a/hb.c b/hb.c
--- a/hb.c	2022-11-18 16:05:28.833868505 +0100
+++ b/hb.c	2022-11-18 16:05:56.260771013 +0100
@@ -102,6 +102,9 @@ hbtransform(XftGlyphFontSpec *specs, con
 		if (codepoints[i] != specs[specidx].glyph)
 			((Glyph *)glyphs)[i].mode |= ATTR_LIGA;
 
+		if (codepoints[i] != specs[specidx].glyph)
+			((Glyph *)glyphs)[i].mode |= ATTR_LIGA;
+
 		specs[specidx++].glyph = codepoints[i];
 	}