Hemera :nixos: :ms_snow_leopard:

Computer Cat with a love for Rust, Linux and creating infrastructure!

Feel free to reply to all the posts you can see :D

Or send me a direct message!

Interests:
- Infrastructure Programming
- Gamedev
- Playing with the coolest tech (Rust 🦀 + Nix ❄️)
- Furry Stuff (Art, Fursuits, Games)

Feline fine! :ms_snow_leopard:

PFP by Shwonky

Header image done by DelKoral

Public Key

npub1xzfa3shpevdl74wwuvhq80ge24qpw34rlm93mfeh2yv265dn7rtq8l66cc

NIP-05 Address

Hemera@meow-social.mostr.pub

Profile Code

nprofile1qqsrpy7cctsukxll2h8wxtsrh5v42sqhg63lajca5um4zx9d2xelp4spz4mhxue69uhhyetvv9ujuerfw36x7tnsw43qa042tz

Publishing to

relay.ditto.pub

Author Public Key

npub1xzfa3shpevdl74wwuvhq80ge24qpw34rlm93mfeh2yv265dn7rtq8l66cc

Show more details

Published at

2026-08-18T06:58:22Z

Kind type

0 Metadata

Event JSON

{ "id": "26ef06d79c2de824fffd1dba5b9e854bb51f8a62c609ee8d77455bdc61c029c7", "pubkey": "3093d8c2e1cb1bff55cee32e03bd1955401746a3fecb1da7375118ad51b3f0d6", "created_at": 1787036302, "kind": 0, "tags": [ [ "emoji", "ms_snow_leopard", "https://medias.meow.social/custom_emojis/images/000/072/728/original/2e6da0c43aa84c59.png" ], [ "emoji", "nixos", "https://medias.meow.social/custom_emojis/images/000/297/583/original/15accda2176f0370.png" ], [ "proxy", "https://meow.social/users/Hemera", "activitypub" ], [ "client", "Mostr", "31990:6be38f8c63df7dbf84db7ec4a6e6fbbd8d19dca3b980efad18585c46f04b26f9:mostr", "wss://relay.ditto.pub" ] ], "content": "{\"name\":\"Hemera :nixos: :ms_snow_leopard:\",\"about\":\"Computer Cat with a love for Rust, Linux and creating infrastructure!\\n\\nFeel free to reply to all the posts you can see :D\\n\\nOr send me a direct message!\\n\\nInterests:\\n- Infrastructure Programming\\n- Gamedev\\n- Playing with the coolest tech (Rust 🦀 + Nix ❄️)\\n- Furry Stuff (Art, Fursuits, Games)\\n\\nFeline fine! :ms_snow_leopard:\\n\\nPFP by Shwonky \\n\\nHeader image done by DelKoral\",\"picture\":\"https://medias.meow.social/accounts/avatars/000/179/026/original/d90efc38d4c67e50.jpg\",\"banner\":\"https://medias.meow.social/accounts/headers/000/179/026/original/09d6caa5ed198a19.jpeg\",\"nip05\":\"Hemera@meow-social.mostr.pub\",\"fields\":[[\"Website\",\"https://hemera.systems/about\"],[\"Github\",\"https://github.com/TheNeikos/\"],[\"Species\",\"Snow Leopard\"],[\"Matrix\",\"@Hemera:matrix.hemera.systems\"],[\"Technologies\",\"#rustlang #nix\"],[\"Enby?\",\"Yes\"],[\"Pronouns\",\"Whatever makes you comfortable interacting with my pfp\"]]}", "sig": "0e9307713b9092015f315c8929f71efb844f182ad122de7bfcc1ce5b2748ecd909617be56b6fd018c77538dae7b6a7174590af8535bf7415c43c001c74dfbcaf" }

Last Notes

2026-05-28 09:34:55 UTC

- reply

@nprofile…mcwl
Woah, way better than usual for me!
www.maptap.gg May 28
98🎯 99🎯 92🎓 95🏆 95🏅
Final score: 951

2026-04-27 05:57:23 UTC

- reply

@nprofile…mcwl
www.maptap.gg April 27
95🏅 92🏆 76👏 51😔 45😨
Final score: 627
Ha!

2026-04-09 06:29:24 UTC

- reply

@nprofile…mcwl
www.maptap.gg April 9
97🔥 99🎯 44😨 68😁 41😕
Final score: 611
Hehe :)

2026-04-01 10:59:21 UTC

"Hey, this service is returning things that aren't in the spec"
"Yeah, that's still being worked on"
"Ok, but its in production?"
"Yeah, we are just slow writing the spec."
??????
:neocat_facepalm:

2026-03-30 13:17:22 UTC

The official #nixos branding guide says "Do not place the logo on a cake"
:dragnsad: 🍰
https://medias.meow.social/media_attachments/files/116/318/313/746/028/333/original/c23219a4f9a1cb46.png

2026-03-25 12:51:09 UTC

Seen: "Inconsistent precision on your display"
What was displayed: 0.0.17, the version number
😭 😭
How many decimals do you use for your version number :DDD

2026-03-10 12:20:10 UTC

- reply

@nprofile…mcwl 835 hehe :dragncool:

2026-03-09 19:05:59 UTC

- reply

@nprofile…mcwl 772 💪 Hehe

2026-03-06 18:48:00 UTC

- reply

@nprofile…mcwl Only 700 :C

2026-02-21 22:51:55 UTC

- reply

@nprofile…mcwl I mean, you can stream it yourself :)

2026-02-21 15:58:59 UTC

- reply

@nprofile…mcwl No this is 100% valid, SQLite can be faster than using files as well :D

2026-02-21 08:54:05 UTC

I wish SQLite had a 'modern dummy preset' that is versioned, such that it enables all the settings that one would want to have, but are not set due to backwards issues otherwise.
Stuff like references, but also more obscure settings like "unicode61" tokenizer wrongly handling diacritics unless you set some setting to '2' :D
One would create a connection, and send it as the first pragma. And wouldn't change it unless one is ready to accept the potential breaking change.
Oh well, maybe there is a resource somewhere that lists all the settings/pragmas one should set upon starting an SQLite connection, such that it behaves more like what one might expect?

2026-02-06 10:25:53 UTC

Nothing like actually typing out: systemctl enable backdoor
:dragncool:

2026-02-02 15:22:16 UTC

Why is rustdoc so slow? I feel like it wasn't that slow 'before'. (I don't have a cutoff, but I built docs for projects like these before, and it did not take half an hour lmao)
Sure it was 800 crates (Most of which are crates I don't build, so I assume they are crates for other platforms) but 25 minutes ?? And most of it seemingly single-threaded? On a 12-core processor!
This is on rustdoc 1.91.0
#rustdoc #rustlang #rust

2025-12-19 19:42:01 UTC

- reply

@nprofile…f7sz I am pretty sure you learned a lot in the process!

2025-10-14 12:52:03 UTC

- reply

@nprofile…f7sz Du meinst doch Motorradschuppen oder?

2025-08-24 17:33:17 UTC

https://github.blog/changelog/2025-08-22-premium-request-overage-policy-is-generally-available-for-copilot-business-and-enterprise/
Am I getting this right, that Github has decided that some of its customers weren't paying for something inadvertantly, so now its easier to just... 'use' more than you expected?

2025-07-23 06:35:22 UTC

Me vs the guy she tells you to not worry about
https://medias.meow.social/media_attachments/files/114/901/157/070/641/812/original/5daf76dbaf35e0b3.jpeg

2025-07-20 07:33:22 UTC

Every time I read an article about shortcoming of a tool, I wonder how much of that is ego and how much of it is an actual shortcoming?
Maybe it's a sign my perfectionism is going down, but different tools being imperfect in different ways does not make them bad or less useful?
Like, I just read an article about the #rust borrowchecker not being able to prove a valid program because it doesn't look 'through' functions. And this making it 'worse' in user experience.
And this feels fairly off to me. Sure the checker could try to do whatever you wish it did. But it doesn't, and so you have to adapt.
Like, not everything needs to fit what you expect from a tool I think.
It is fair to not use that tool (i.e. Rust) if you find that annoying, but I struggle hard with this mindset. Where rather than adapting to use a tool to its fullest, people just bang their head against it repeatedly saying "look how bad it is".
#rustlang

2025-07-17 14:29:57 UTC

I remember a time when I was happy to read something on Medium, but nowadays its just a really bad experience :c

2025-06-24 09:00:09 UTC

- reply

@nprofile…ss2n @nprofile…8j6a allowing myself to butt in, since I have been eyeing for annex to handle my media library. :neodog_floof_cute_tail:
Reading though git-annex-export it seems that it causes an “as is” file export from a remote? Aka something a “normal” application could interact with? (Aka a Jellyfin?)
Does it do this through physical files? Or could they be on-demand?

2025-06-20 12:39:59 UTC

Day Two #gpn23 Hope everyone is doing well!

2025-06-17 12:54:27 UTC

- reply

@nprofile…fe8t I don't see a difference in frequency either?
There's lots of small comments in the python code in the style of:
# Do the thing if empty
if foo is '':
do_thing()
Which I wouldn't consider to be prime examples for readability.
Don't forget also that you're looking at productive code, meant to be read by people who know Rust and the codebase.
The fact that Rust is also more expressive than Python when it comes to high-level constructs is ofc not helping if one doesn't know individual combinators by heart, but one can look them up!

2025-06-17 12:48:59 UTC

- reply

@nprofile…fe8t I don't see a big difference?
The rust code is internal, which doesn't have doc or normal comments. And the python code has doc comments, and none for the private code?

2025-06-10 15:19:21 UTC

- reply

@nprofile…fe8t It's in the 'Explore' Tab at the bottom!

2025-06-10 14:05:12 UTC

Just ran into an interesting #cargo #rustlang workflow problem, that I am not sure how it is best avoided?
A project I work on is depending on strum and strum_macros. Due to a cargo update multiple strum versions now got unified, and only a single version is built. That's good.
A dependency I was using had a pattern as such:
use strum::EnumMessage;
use strum_macros::EnumMessage;
Which is totally fine!
With the unification, the features now also got unified. Which in this case meant that EnumMessage is now present in the macro-scope in both strum and strum_macros.
Which means the above two lines cause an error with EnumMessage being imported twice! Fun.
So the situation is that strum is fine, as it is an additive feature. And the dependency is fine as well, as it itself doesn't use the 'derive' feature.
Another reason why this happens seems to be because we are using cargo-hakari to have less cache-misses in our workspace. But once our workspace-hack crate added the derive feature it broke the build.
The final piece seems that a transitive dependency both:
Had the derive feature of strum enabled
Did a 0.x.y release (aka minor? patch?) that had a major update of its dependencies in it. (i.e. 0.26 -> 0.27)
Is a proc-macro ? (potentially irrelevant, potentially the missing piece)
So to me the full reproduction would be:
Have a dependency tree that contains both foo v0.1 (derive feature) and foo v0.2 (no features)
cargo update
It still builds
cargo hakari adds for both 'normal' deps and 'build' deps foo v0.2 (derive feature)
This causes cargo to unify foo for both normal builds and proc-macro builds
The build fails due to duplicate entries
It feels like the use of cargo-hakari caused this situation, as without it, it would have been fine I guess? But adding a feature flag, shouldn't break downstream crates either 😿 so bleh

2025-06-09 08:39:42 UTC

For those that don't know, qalc is amazing!

2025-05-23 11:49:15 UTC

Today is release day for nixos!

2025-05-14 13:46:07 UTC

Oh #zed supports #vim mode?
But does it run in the terminal? 🤔

2025-05-08 10:38:29 UTC

- reply

@nprofile…ss2n @nprofile…8j6a I mean process substitution is not thaaat more complicated: diff (cmd1 | psub) (cmd2 | psub)

2025-05-06 13:03:12 UTC

German politicians have the potential to do something really funny twice in a row

2025-04-09 08:55:13 UTC

git push --force-with-lease truly is a lifesaver when working with multiple peeps on a branch

2025-04-08 10:38:32 UTC

The fact that XPath has support to serialize to JSON opens up a lot of amazing opportunities
👀
#xml #xpath #xee #rust #rustlang

2025-04-07 14:10:43 UTC

I love when people argue about C vs Rust and try to say "But I just want to be able to write down my code and not have to worry with all the baggage Rust forces on me" and I am always dumbfounded by it.
Am I the one interpreting it as too shallow? Because the way I see it, is that the moment you write non-trivial code most non-borrow-tracking languages will not be your friend when it comes to making sure that what you wrote actually does what you think it does.
Like, 'simple' Rust code is simple. And yes, a method being duck-typed is complex!! So of course your language should help you to make sure your actually articulate your constraints. It's the same with Typescript really. And also what people seem to dislike a lot about it.
To me, this is just intellectual laziness. Yes, you have to think about the interfaces of the different parts of your fundamentally maths-based creation!!
What do you think should be the case instead? If something is 'vibe coding' it should definitely be whatever these people think they are doing.
Like, it's the epitome of the bug-creation: You write a nice self-contained function, that reads some input and does things. Someone then adds features that make it non-threadsafe. Maybe sharing some references or w/e. This would just be a bug report in C, but of course Rust will complain that something is not Sync/Send and you should fix that! Does it suck you have to do that? NO cause you'd have to do it ANYWAY. Because it's just pure logic underneath.
I think that I should introduce Rust not as an alternative to C, but really what it is: "All those tools and analyses that people wish they'd have for C. That's why the syntax is somewhat adjacent."
Of course that's not true for either, but maybe it helps the framing, and one doesn't have to hear a complete BS reply of "Proper C with the correct tools is just as safe."

2025-04-02 15:45:18 UTC

It’s called C# because your project dies by a thousand cuts
Send toot

2025-03-28 15:51:40 UTC

- reply

Some example testcase errors, thanks to the awesome #miette crate from @zkat (who also authored #KDL btw!)
https://medias.meow.social/media_attachments/files/114/240/845/011/407/563/original/c298548296cb1e67.png
https://medias.meow.social/media_attachments/files/114/240/849/318/284/408/original/d5fc8ef549ec8ba9.png

2025-03-28 15:48:47 UTC

Alright!
I just released a first functional version of a new crate:
🔧 test-dsl at 0.1.0
🦀 https://github.com/TheNeikos/test-dsl/
To paraphrase the readme: test-dsl allows you to easily author tests by decoupling between the 'test harness', 'verbs' and 'conditions'.
It's role in your testing infra would be as boilerplate reduction, so that you can focus on just writing tests around the behaviour of your system. I'm using this pattern for example in a client-server context. So I have verbs like start_server 1, start_client 2, connect_client 2 1 etc...
Of course the meaning of these verbs is highly project-specific.
If a verb does fail, either by panicking or returning an error, you get a nice error output telling you exactly which file and which verb caused the error. Allowing you to more easily pin-point directly what went wrong, rather than potentially searching through a log.
It works great together with datatest-stable from the nextest project, but it works just as fine standalone.
It uses #KDL as the language to write your testcases in, so its fairly nice to read.
It's still missing more comprehensive documentation, but as a prototype it should work just fine! It actually sprung out from having used this pattern a few times, and I didn't want to re-write the boilerplate around it, and maybe re-use some code between projects.
#rust #rustlang

2025-03-28 10:35:16 UTC

- reply

@nprofile…8j6a Ich denke, dass die Bank das nicht als 'personenbezogene Daten' sieht. Und daher nicht mitmacht.
Ob das so richtig ist, weiss ich nicht :D

2025-03-25 20:08:57 UTC

I've recently grown super fond of using the command pattern to write quick to author, and easy to understand tests.
The gist is that you define a list of 'commands' that your application can logically perform that you find interesting to model.
So for example I'm currently writing an MQTT client, so 'connect', 'publish', 'server disconnect', etc... are all commands that make sense to have in some fashion.
This then allows you to really quickly write down tests whenever you encounter some weird behaviour, or are wondering yourself 'what happens if'.
I'm using lots of async, and having a single-threaded executor allows a lot of control over timings, and finding out bugs that can potentially cause dead-locks (by effectively dead-locking in single threading)
The main benefit really is just writing more tests overall! You can then go more crazy and for example permute difference sequences. This allows you to get more coverage of various code-paths.
In one application I was writing, I had multiple clients connect (a multi step process), send a message (another multi step process), and then disconnect.
This unearthed some locking badness, where I was holding onto DashMap read-locks for too long at the wrong places. (And part of the reason why I wrote a custom dylint-lint to detect that cf.: https://git.hemera.systems/Hemera/dylint-various/src/branch/main/lints/dashmap-ref/src/lib.rs )
Its of course not a panacea, but it is a nice tool! I'm trying to see how far it can go together with check-point testing 🤔
#rust #rustlang #programming

2025-03-14 08:41:36 UTC

🎶 ✨ A spoonful of clone makes the compiler errors go down~ compile errors go down ✨ 🎶

2025-03-11 15:23:12 UTC

"MQTT is the JSON of the application architect"

2025-03-10 15:23:36 UTC

Using #git on a lower level than just commit+push feels like magic 🤩
I got told to clean my commits (they were right... :blobcatcry:)
So git rebase -i to the rescue! Edit the commit, re-commit the good parts, commit the other parts as needed, recombine the order, fixup/reword as needed, and voila!
Luckily it only took a few minutes, and if I broke something, I could always --abort or go to the reflog

2025-03-04 10:17:53 UTC

I finally got around what I've been wanting to do for quite some time!
@nprofile…8j6a and I have written a nix library to use cargo-dylint with #nix :blobcat:
The idea is to have the same experience as the default dylint, just reproducible.
It already works, but is a bit rough. The idea is to now heavily dog-food it and get it working easily, so that anyone can just add new lints through their nix config
:blob_cat_melt:
We're looking for input and ideas!
Repo: https://git.hemera.systems/Hemera/nix-dylint
Github Mirror: https://github.com/TheNeikos/nix-dylint (PRs still welcome)
#nix #nixos #rust #rustlang #dylint

2025-02-28 06:27:43 UTC

- reply

@nprofile…ss2n @nprofile…vu7f the disabledModules needs to be a path to the module you wanna disable, not just the name. In the link I sent they show an example!

2025-02-27 10:34:54 UTC

- reply

@nprofile…ss2n In that case I'd blacklist the nixpkgs module you patched, and import it from your fork

2025-02-27 10:26:52 UTC

- reply

@nprofile…ss2n It depends on what patch you're interested in.
Did you change a package? If so, an overlay should be fine

2025-02-20 10:14:02 UTC

- reply

@nprofile…8j6a wazzup

2025-02-13 14:53:47 UTC

The Entry pattern in #Rust types is so incredibly powerful!
It's so fun to use
#rustlang

2025-02-13 08:45:07 UTC

When you just want to 'quickly' fix something in your #nix config
https://medias.meow.social/media_attachments/files/113/995/694/592/143/400/original/b24f2c484ca5440b.mp4

2025-02-11 11:52:56 UTC

Why is ING being such asses about their online access?
They now require a German phone number to login… which I don’t have nor needed ever.
It’s not my primary bank so w/e but???

2023-10-25 14:49:48 UTC

GOD, I hate hate hate #YAML. Total BS language. It's literally 'Footgun: The Language'
I'd love to just generate it and be over it, but gitlab doesn't support generating the yaml file on the fly :)))
I might do something with #nix that generates a gitlab yaml (since it should be just JSON really), and includes a check that runs before everything else to verify whether the generated file is up-to-date....
🤬 🤬 🤬 🤬 🤬

2023-08-15 12:53:50 UTC

Hmmmm, I was writing some stuff for an upcoming RP and stopped at "Sagittarius A*" and then wanted to look up how long a galactic year for the milky way was... This had then led to 20 new tabs on Wikipedia and three hours of reading!
At least now I know that the there is a concept negative energy, its gravity, and that the future looks a lot less boring than just the heat death of the universe.

2023-07-13 07:59:42 UTC

I should ban parentheses from my normal writing. (Because it's literally just an added sentence (like this one), and it usually works just fine without them)