| Age | Commit message (Collapse) | Author | Lines |
|---|
|
I have really poor discipline when it comes to /tmp.
In particular, almost every boot, I create a /tmp/t, and then I pollute
it.
With this, I can make auto-cleaned-up directories that are arbitrary
(`tt`) that's for the quick experiments.
And then there's a helper for longer term stuff (via `t`) that's
equivalent to what I already do, but nicer.
|
|
|
|
|
|
|
|
The interest is mostly in the code actions.
|
|
I didn't like that you couldn't keep mixing and matching, so I made it
so you can.
In the process, I also implemented even?, odd?, filter, mapcat, flat,
and concat.
Some of them didn't end up needing to exist, and the entire :toast.
namespace does need to be cleaned up.
Then again, it's not like I'm selling this like an alternative stdlib.
I should really try and port all of clojure.core though, it'd be funny.
|
|
Also, simplify the opts.
Since I no longer want to require a mini component for the config,
I don't have to wrap it in an fn.
Also, I should just do a `#{}` next time anyway.
|
|
Also I removed the workspaces stuff.
I don't use them.
|
|
Yeah so the default is the specific attach, not the default one I wrote
up.
|
|
There's still some improvements to be made, but this is cleaner.
We lose global caps defaults, but those made no sense to begin with.
We do also lose caps overrides, I'll add them back in if I need them.
In the process, we gain `every?` and `executable?`.
There's also a misc comp print vim.inspect to vim.print change.
|
|
|
|
The plan from now on will be "single big file" unless there's more
complex things going on, such as with lsp.
|
|
This also introduces a neovim-level `.chezmoiremove`.
This will be used from now on to cleanup files.
It might grow over time but I can start trimming the top lines once it's
been "long enough" anyway.
|
|
|
|
|
|
|
|
I have to touch it at work, as troubling as that is.
|
|
|
|
|
|
Thank you Olical for updating fennel!
Move to `(set (. t k) v)` for overwriting non-insertion list elements.
Also don't bother specifying a `k` for the insertion case, seeing as we
know for a fact it's just the final element.
|
|
I swear I'm not succumbing to the APL, I just couldn't find a better
symbol.
|
|
Lmao oops, tfw you don't function without `dec` anymore.
|
|
I'm not using my stdlib because I want the final .lua file to be usable
drop-in (as long as you have mini.base16 of course).
|
|
It performs significantly better and I have no idea why.
|
|
Oops I was debugging don't mind me.
|
|
Yeah I'm lazy but these are my dotfiles.
A quick summary:
* plugins.lsp.attach is now ported
* toast.core is reorganized
* new nil? and number? checks
* new assoc function
* new toast.table.unpack function (cross lua compat)
* toast.macros.tbl now uses assoc instead of from-pairs and group
* apparently I had forgotten to add {pre,post,init}.fnl before, so
that's done now
Still todo: ftplugin, colors.
|
|
Until now you couldn't do `(tbl 1 2)`.
Now you can.
|
|
One file isn't done, and there's also the question of ftplugin and co.
One step at a time though, eh?
|
|
|
|
|
|
|
|
|
|
I've been having to work with jsonc and xml recently.
Long enough to get some preferences.
Potentially to be expanded upon later.
|
|
Ok so this is a deep cut.
The problem is that when you have diff.external set, it overrides
*everything*.
So if you want to have, for example, a textconv diff filter… you just
can't do that.
Even passing --no-ext is questionable.
Instead, just require the use of `dt` / --ext-diff.
It's tragic :(
|
|
It honestly makes no sense, why did I do this?
It breaks on older fish and is really unnecessary.
|
|
To my knowledge there's no advantage to the builtin over `z`.
Maybe syntax highlighting?
Anyway I forget to type `z` pretty often, so just alias cd.
|
|
|
|
|
|
I increase/decrease quote levels a lot, so I figured I'd write a small
thing for doing that more easily in visual mode.
This isn't without problems. Known bugs:
* In normal-mode, it's a bit slow (waiting for more inputs? possible,
but if you give it more inputs it just does it twice).
* You cannot increase/decrease indentation with <</>> with this on. This
is mostly a bummer for code blocks.
* Doesn't work with indentation-based quoting, rather than >.
But it's good enough for me, eh?
|
|
1. We default to $PATH now, but allow having an explicit $DOTPATH.
This is important because the UX of source-ing /bin/ls is bad.
2. Explicit handling for real paths (/*, ./*, ../*).
|
|
|
|
It's debatable but ultimately it *is* arch-independent and ro.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
Chloé Vulquin