Included changes:
* targeted primary NixOS channel back at `nixos-unstable`
* //3p/chicago95: disable new broken symlink check
* //tvix: omit unnecessary lifetime clippy recommends
* //users/sterni/blipqn: wait a bit for data to arrive in test so it
succeeds under load.
* //fun/paroxysm: force pkg-config flag of pq-sys by adding a bogus
dependency on it. Otherwise, pq-sys will try to use pg_config
which does not work correctly in pkgs.libpq at the moment.
* //users/flokli/keyboards/dilemma: disable temporarily
Change-Id: I6d53bd7bca6886f3457e1f41505e97314f4cd191
Reviewed-on: https://cl.tvl.fyi/c/depot/+/13119
Reviewed-by: tazjin <tazjin@tvl.su>
Tested-by: BuildkiteCI
Reviewed-by: flokli <flokli@flokli.de>
Autosubmit: tazjin <tazjin@tvl.su>
Reviewed-by: aspen <root@gws.fyi>
Reviewed-by: sterni <sternenseemann@systemli.org>
git-rev-list(1) orders the printed revisions before printing. This is
quite complicated and I noticed that it behaves in unexpected ways
w.r.t. commits given individually: sometimes they are reordered,
sometimes not. This is problematic when interdependent commits are given
via the command line individually: If they aren't re-ordered by
git-rev-list(1), we'd be cherry picking them in reverse order.
To solve this, just call git-rev-list(1) on all arguments individually,
allowing ranges to be resolved, but always picking individual arguments
in the correct order.
Notably, this means that excluding commits via individual REV^ arguments
won't ever work, but I believe this wasn't possible previously
anyways (maybe due to --no-walk?).
Change-Id: If8e83e0b47a74baf37d77b6c4f68a55af944b366
Reviewed-on: https://cl.tvl.fyi/c/depot/+/13149
Tested-by: BuildkiteCI
Reviewed-by: sterni <sternenseemann@systemli.org>
Autosubmit: sterni <sternenseemann@systemli.org>
- Reduce visual noise, e.g. parens where it hurts readability.
- Separate parts of trains visually where it helps readability.
- Reduce amount of unnecessarily pointfree code.
Change-Id: Ie0ef03af7d344863a81bf0557b90a4dfb7974e84
Reviewed-on: https://cl.tvl.fyi/c/depot/+/13126
Autosubmit: sterni <sternenseemann@systemli.org>
Reviewed-by: sterni <sternenseemann@systemli.org>
Tested-by: BuildkiteCI
Passing a namespace as 𝕩 to •CurrentError is not allowed since it is
reserved for future use (not sure how it'll change the behavior
exactly). Since WithFlipdot always passes a namespace as 𝕩, this is a
relevant problem for blipqn. To work around it, we wrap the passed
argument in a list (and unwrap it for 𝔾).
Change-Id: I3382f754cbb64ad799e47b7d174d641c43a0f2e0
Reviewed-on: https://cl.tvl.fyi/c/depot/+/13125
Reviewed-by: sterni <sternenseemann@systemli.org>
Autosubmit: sterni <sternenseemann@systemli.org>
Tested-by: BuildkiteCI
Included changes/fixes:
* bumped all `wasm-bindgen` usages again
* regenerated protobuf files
* keycloak terraform provider has been migrated to new name
This also included a state migration in the bucket, which I've already
performed.
* tvix/boot: disable tests that are broken in CI
* users/aspen/yeren: avoid upgrading kernel to 6.12
digimend depends on a fix: https://github.com/NixOS/nixpkgs/pull/378830/
Change-Id: I657dcf5c4d0d08f231bfe30e37c8062bfcfaaa32
Reviewed-on: https://cl.tvl.fyi/c/depot/+/13098
Reviewed-by: aspen <root@gws.fyi>
Tested-by: BuildkiteCI
Reviewed-by: sterni <sternenseemann@systemli.org>
Autosubmit: tazjin <tazjin@tvl.su>
Otherwise, two distinct entries would link to the same location with the
older entry “winning”.
Change-Id: I73cce835c110fcce0161c9581cc415697a7d74ab
Reviewed-on: https://cl.tvl.fyi/c/depot/+/13113
Autosubmit: sterni <sternenseemann@systemli.org>
Reviewed-by: sterni <sternenseemann@systemli.org>
Tested-by: BuildkiteCI
Markup for the git backend is implemented as a lookup table to commands
that read markup on stdin and write it to stdout. Markdown uses lowdown,
Org pandoc and HTML doesn't need to be altered.
Future Work:
- Title extraction from the files instead of the file name. Since we use
the file name as ids in the git backend, it is currently impossible to
change the title without breaking links.
- Use emacs for Org. Unfortunately, it is really slow, so we may need to
do something drastic, like writing an sourcegraph style markup
conversion service or dumping an emacs image.
- Copy post dependencies, e.g. included images.
Change-Id: I88ad4cc18ada1d68d81fdecf685096f105e3fef3
Reviewed-on: https://cl.tvl.fyi/c/depot/+/13110
Tested-by: BuildkiteCI
Autosubmit: sterni <sternenseemann@systemli.org>
Reviewed-by: sterni <sternenseemann@systemli.org>
Markup conversion and metadata querying beyond date is not yet
implemented. Currently, plain text is rendered in a <pre>, but without
any escaping.
Change-Id: I74cd2f8a0eab23bea00ca15232e915c0a3c74138
Reviewed-on: https://cl.tvl.fyi/c/depot/+/13109
Reviewed-by: sterni <sternenseemann@systemli.org>
Tested-by: BuildkiteCI
Autosubmit: sterni <sternenseemann@systemli.org>
This implementation is still very basic. For reference, e.g. mozilla has
published a more fleshed out variant of such an algorithm:
<https://github.com/mozilla/unicode-slugify>.
Change-Id: Id3d76a4f9e976952dd3c98835df0de7c4f86c3bc
Reviewed-on: https://cl.tvl.fyi/c/depot/+/13108
Autosubmit: sterni <sternenseemann@systemli.org>
Reviewed-by: sterni <sternenseemann@systemli.org>
Tested-by: BuildkiteCI
I'll probably have this added to the BQN website's editor support page.
It'll probably be less confusing if the page people are sent to doesn't
have information on random, semi-related stuff.
Change-Id: I75e03afea1362e022923ef8781b35113bf26b2e9
Reviewed-on: https://cl.tvl.fyi/c/depot/+/13116
Autosubmit: sterni <sternenseemann@systemli.org>
Reviewed-by: sterni <sternenseemann@systemli.org>
Tested-by: BuildkiteCI
I've finally figured out what the problem was with •CurrentError:
Passing a namespace as 𝕩 to •CurrentError is not supported as it may
control some kind of special behavior in the future. The right thing to
do seems to check the •Type of 𝕩 and only attempt to use •CurrentError
if it isn't a namespace.
Change-Id: I012009d0f1b913e77d65776ee43062817199baab
Reviewed-on: https://cl.tvl.fyi/c/depot/+/13106
Autosubmit: sterni <sternenseemann@systemli.org>
Tested-by: BuildkiteCI
Reviewed-by: sterni <sternenseemann@systemli.org>
We can use ! in R to display the error instead.
Change-Id: Ib7f6e559014e3f99a9f99c159a31a45a7a0e65a5
Reviewed-on: https://cl.tvl.fyi/c/depot/+/13105
Reviewed-by: sterni <sternenseemann@systemli.org>
Autosubmit: sterni <sternenseemann@systemli.org>
Tested-by: BuildkiteCI
The array listing was pretty nice for debugging, but having arrays where
every element has a different meaning as an interface between different
parts of the code is not great. Refactoring will be less frustrating if
we just store everything using a key-value scheme in namespaces.
Change-Id: I806fea247b6f996dbdd93df24f3fdb762340f043
Reviewed-on: https://cl.tvl.fyi/c/depot/+/13017
Autosubmit: sterni <sternenseemann@systemli.org>
Tested-by: BuildkiteCI
Reviewed-by: sterni <sternenseemann@systemli.org>
We can actually get all the headers we are interested in just one call
of mhdr(1) since it can query multiple headers in multiple messages in
just one call. With BQNs reshape / two dimensional array facilities it
is easy to group the lines of headers by their originating message and
append the original path to the result array.
We still use a separate mhdr(1) call for the Date header since we rely
on the -D option for parsing the header.
Change-Id: Iaaf8f21fa22ab3e32e6900df4ae164e72c722cd3
Reviewed-on: https://cl.tvl.fyi/c/depot/+/13016
Reviewed-by: sterni <sternenseemann@systemli.org>
Tested-by: BuildkiteCI
Autosubmit: sterni <sternenseemann@systemli.org>
Currently, blërg is an incomplete reimplementation of
//users/sterni/mblog that uses mblaze(7) and //users/sterni/mn2html to
process the backing mail notes.
Now, why start from scratch again?
- mblog is depressingly slow at the moment which is due to
//third_party/lisp/mime4cl. The performance problems are
(probably) solvable, but this would require carefully
redesigning the library's use of streams. I'm not motivated
to look into that at the moment, unfortunately.
- I had an idea for intergrating additional storage backends
for entries into an mblog like software. I've found experimenting
with this in BQN to be quite pleasant as the backends can
easily be expressed as namespaces.
Change-Id: I90c8ff7b5f7235d6fd7b0619338b553be8742e49
Reviewed-on: https://cl.tvl.fyi/c/depot/+/13015
Autosubmit: sterni <sternenseemann@systemli.org>
Tested-by: BuildkiteCI
Reviewed-by: sterni <sternenseemann@systemli.org>
TL;DR: xwayland-satellite (which I use) doesn't sync the primary
selection between X11 and wayland. Since devdraw prefers the primary
selection over the clipboard if populated, it is impossible to paste
from wayland (i.e. from the clipboard) in devdraw windows after cutting
once on the plan9port side (which populates the primary selection).
Change-Id: Idc46b875f294bf90d477ee7fe138dcd17e755221
Reviewed-on: https://cl.tvl.fyi/c/depot/+/13102
Autosubmit: sterni <sternenseemann@systemli.org>
Tested-by: BuildkiteCI
Reviewed-by: sterni <sternenseemann@systemli.org>
I semi regularly look up the sunset time, so this is actually useful for
me. Time zone conversion seems to be broken which I'll have to debug.
The GMT sunset times it comes up with seem to be accurate, though.
Change-Id: If179f9e1bb56665d4d6f8b0f81272d3d74fe6cec
Reviewed-on: https://cl.tvl.fyi/c/depot/+/13101
Tested-by: BuildkiteCI
Autosubmit: sterni <sternenseemann@systemli.org>
Reviewed-by: sterni <sternenseemann@systemli.org>
Reimplement the MIME body extraction and HTML rewriting from mblog in
Rust so that shelling out to it becomes viable. The problem with
mnote-html is mainly that it – being written in CL – requires a ~300MB
executable and is a bit sluggish starting.
Change-Id: I5c1adc1a7ab5f3dde207f9a1f67ace685bd3f69f
Reviewed-on: https://cl.tvl.fyi/c/depot/+/13014
Tested-by: BuildkiteCI
Reviewed-by: sterni <sternenseemann@systemli.org>
Autosubmit: sterni <sternenseemann@systemli.org>
much easier to use than my typical inotify shell loop
Change-Id: Id9c260d7f57bdde5307149e982ace4eca960ecdb
Reviewed-on: https://cl.tvl.fyi/c/depot/+/13055
Reviewed-by: tazjin <tazjin@tvl.su>
Tested-by: BuildkiteCI
Autosubmit: tazjin <tazjin@tvl.su>
I've checked the new links manually.
- //web/tvl: changed the target path of links to the tvix docs since
they were moved in r/2378.
- //users/aspen/resume: Tvix is no longer //third_party/nix.
Change-Id: I419bae1a46bdccc7baa7327215aa2368ffc0f01c
Reviewed-on: https://cl.tvl.fyi/c/depot/+/13043
Reviewed-by: aspen <root@gws.fyi>
Tested-by: BuildkiteCI
Autosubmit: sterni <sternenseemann@systemli.org>
Reviewed-by: fogti <fogti+devel@ytrizja.de>
Reviewed-by: tazjin <tazjin@tvl.su>
I like project-find-regexp better than grep-find-file, but it's a little
annoying to use for arbitrary directories. With C-u, it unnecessarily
prompts for a file name pattern.
For convenience, add a variant of project-find-regexp which always
prompts for a directory, but doesn't require a project nor prompts for a
file name pattern.
Change-Id: I40f0bdcca4aa129f447f4f4790d732afcf76d3f7
Reviewed-on: https://cl.tvl.fyi/c/depot/+/13056
Autosubmit: sterni <sternenseemann@systemli.org>
Reviewed-by: sterni <sternenseemann@systemli.org>
Tested-by: BuildkiteCI
Archive to a single archive file that is ignored in org-agenda.
Archiving that “stays close” can be done using the archive tag.
Change-Id: Icd121a765e5befa0ab2f6df662d757483c65efa5
Reviewed-on: https://cl.tvl.fyi/c/depot/+/13049
Autosubmit: sterni <sternenseemann@systemli.org>
Reviewed-by: sterni <sternenseemann@systemli.org>
Tested-by: BuildkiteCI
This is convention in Plan 9 and e.g. the plumber will reformat the
rules this way when they are read back via `9p read plumb/rules`.
Change-Id: I53543d774c11789c8352e2e632678313037c5555
Reviewed-on: https://cl.tvl.fyi/c/depot/+/13048
Reviewed-by: sterni <sternenseemann@systemli.org>
Tested-by: BuildkiteCI
Autosubmit: sterni <sternenseemann@systemli.org>
This uses git show to display the revision same as the git commit rule.
Change-Id: Ibf518adb60d15cac2126b6ab0c7e12c4b4637c25
Reviewed-on: https://cl.tvl.fyi/c/depot/+/13041
Reviewed-by: sterni <sternenseemann@systemli.org>
Autosubmit: sterni <sternenseemann@systemli.org>
Tested-by: BuildkiteCI
context.org tracks ongoing things, e.g. code I'm working on or want to
get started on. Tracking means that it contains notes (“context”) on the
things in question. These will mainly be links or quick todo lists which
will hopefully reduce the amount of open tabs and unread emails (marked
as TODO in thunderbird) I have.
Having inbox.org open is not as useful since it isn't viewed really: For
adding tasks org-capture should be used and for viewing them org-agenda
is far superior.
Change-Id: I4acbaa33e4c2de40a92393376bb86251d9dceb02
Reviewed-on: https://cl.tvl.fyi/c/depot/+/13044
Reviewed-by: sterni <sternenseemann@systemli.org>
Autosubmit: sterni <sternenseemann@systemli.org>
Tested-by: BuildkiteCI
The wrapper allows me to pick and choose what to install and also to set
some extra environment variables if necessary. A separate wrapper
derivation prevents the hefty rebuild of the entire plan9port derivation
when making changes.
Change-Id: I3f96cc3e31baf3bd37ae419578c5789aa093e578
Reviewed-on: https://cl.tvl.fyi/c/depot/+/13038
Reviewed-by: sterni <sternenseemann@systemli.org>
Tested-by: BuildkiteCI
Autosubmit: sterni <sternenseemann@systemli.org>
To avoid the need for two different versions of plan9port in the CI
pipeline, just move the plumbing sanity check into installCheckPhase.
For local iteration you'd have a plumber running anyways, so the short
build time of the plumb derivation isn't really a necessity.
Change-Id: Ib0ba5fa2d8107530fb7d1144cf7395f4f5a3cd12
Reviewed-on: https://cl.tvl.fyi/c/depot/+/13035
Tested-by: BuildkiteCI
Autosubmit: sterni <sternenseemann@systemli.org>
Reviewed-by: sterni <sternenseemann@systemli.org>
plan9port completely ignores XCompose because it has its own compose
mechanism (which is mapped to the same key). The sequences are defined
in /lib/keyboard and need to be compiled in.
Support for the BQN unicode characters is achieved by generating the
necessary lines for /lib/keyboard from the .inputrc (for GNU readline)
that is part of mlochbaum/BQN (simply because that file is somewhat
parseable and stores the sequences in ASCII, contrary to .XCompose).
This is implemented by a small BQN script which is executed in
postPatch.
All usual sequences are supported except those that map to the second
ASCII character of the sequence. These exist to keep certain characters
typeable in other input system. Thanks to the explicit compose key,
Plan 9 doesn't have this problem.
Change-Id: I590c03fd69a2aae3cbbbd39ebcbce6cec0418b50
Reviewed-on: https://cl.tvl.fyi/c/depot/+/13034
Reviewed-by: sterni <sternenseemann@systemli.org>
Tested-by: BuildkiteCI
Autosubmit: sterni <sternenseemann@systemli.org>
