Posts by goldstein@im-in.space
(DIR) Post #AQyAvAd5NWi39BCdxg by goldstein@im-in.space
2022-12-25T15:29:38Z
0 likes, 0 repeats
@goldstein It was accepted! cargo-mommy is now a part of nixpkgs.I love NixOS
(DIR) Post #AQyV4BAwiQ6CWwc7pQ by goldstein@im-in.space
2022-12-25T19:28:03Z
0 likes, 0 repeats
@mo It’s possible (and has been done multiple times, with different tradeoffs), but getting package into nixpkgs makes it easier to install and provides a binary cache for it.
(DIR) Post #AQyVvsv8nq5HYEpWVs by goldstein@im-in.space
2022-12-25T19:37:49Z
0 likes, 0 repeats
@mo If you look at the actual package, it’s just calling a “generic Nix function” with a few hashesrustPlatform.buildRustPackage rec { pname = "cargo-mommy"; version = "0.1.1"; src = fetchCrate { inherit pname version; sha256 = "sha256-p1SAYUQu1HpYJ6TbLJ3lfA9VlKHvB7z5yiFXmTQOCXA="; }; cargoSha256 = "sha256-5RidY+6EF23UNzz1suSdA4LL59FalipaJ+ISSsmiCXM=";}it’s possible to write a package without manually specifying SHAs via flakes + naersk/cargo2nix, so it’ll be even shorter
(DIR) Post #AR02mX3mCbvon4ye8m by goldstein@im-in.space
2022-12-26T13:19:45Z
1 likes, 2 repeats
git push --force-with-leash
(DIR) Post #AR8qm1TGCT2A9Wx59s by goldstein@im-in.space
2022-12-30T19:04:26Z
0 likes, 1 repeats
normalize meowing at people to remind of your existence
(DIR) Post #ASEMl06aXG8Mj2SuOm by goldstein@im-in.space
2023-02-01T08:54:42Z
0 likes, 1 repeats
I’m non-binaryto be precise, I’m balanced ternary, so I can express negative numbers natively, without needing additional encodingI use this negative numbers to represent my mood
(DIR) Post #AUFXd6mWKxM3CPSJyy by goldstein@im-in.space
2023-04-02T18:13:11Z
1 likes, 2 repeats
so, you know how there is std::mem::zeroed() in Rust if you want to get some 0x00 bytes? but did you know that there’s a function to get you some 0x01 bytes if that’s what you want?it’s called std::mem::uninitialized(). https://github.com/rust-lang/rust/blob/5b08c9f39754039ef9c6cbde157ac9eb8c252a58/library/core/src/mem/mod.rs#L688-L692buy uninitialized memorylook insideinitialized
(DIR) Post #AUTt0bLkWo8x0SKZZQ by goldstein@im-in.space
2023-04-09T16:22:21Z
2 likes, 7 repeats
The Rust Foundation is currently planning to set an extremely strict trademark policy, essentially prohibiting everyone for using the word “Rust” for the project names and domains (affecting project names like rust-crypto, names of books about Rust etc) and also from modifying Rust logo (affecting e.g. Actix). They also managed to trademark the word Clippy, despite taking it from Microsoft.They’re currently accepting community feedback. I think that proposed policy, if accepted, would greatly hurt Rust-related projects for no clear benefit. If you also have an opinion about that, please, give them your feedback.https://nitter.cz/rust_foundation/status/1644132378858729474
(DIR) Post #AXuQsyHcDN37pLuCci by goldstein@im-in.space
2023-07-21T05:27:08Z
0 likes, 0 repeats
@mo @hkc so competitively feminine?
(DIR) Post #AYLuoWtRtURi2BC1lA by goldstein@im-in.space
2023-08-03T11:38:58Z
0 likes, 1 repeats
TIL the latest (20230722) GNU Parallel version has the codename “Приго́жин” for some reason.https://savannah.gnu.org/news/?id=10493I tried to search for more info on why @GNU_Parallel would do that, but didn’t find anything substantial. parallel’s mailing list just announces it like it’s a perfectly normal name without any reasoning behind this choice.
(DIR) Post #AZ94Mt4FrBFIuXDyUa by goldstein@im-in.space
2023-08-26T21:22:38Z
0 likes, 1 repeats
https://gitlab.freedesktop.org/xorg/app/xeyes/-/merge_requests/6TIL xeyes supports multiocular eyes with the -biblicallyAccurate flag
(DIR) Post #AcnPC6M2AtiWg8Etou by goldstein@im-in.space
2023-12-13T19:18:22Z
0 likes, 0 repeats
I solved this day of #AdventOfCode in #Rakulang, which still looks fun, until I try to actually write it and discover once again that basically every operation has some weird footguny semantics. nevertheless, I’m quite happy with the final result, which looks nice and concise.https://github.com/GoldsteinE/aoc2023/blob/master/day13/code/main.raku
(DIR) Post #AcnPC79JDh8H8xIFcm by goldstein@im-in.space
2023-12-13T21:02:11Z
1 likes, 0 repeats
some random examples of weird footguny semantics in Raku:my @arr = (2, 3);reduce &[+], 1, @arr;is 3 (list gets reinterpreted as a scalar). you need to use |@arr for it to work. reduce &[(&)], ();(identity element for set intersection) is an empty set for some reason. Set.new((1, 2), (3, 4)) (&) Set.new((3, 4), (5, 6))is an empty set (sets use comparison by identity, so all the lists are considered different).set((1, 2), (3, 4))is a set of four numbers 1, 2, 3, 4, set flattens its arguments, while Set.new does not. sequences like x...^y and (x..^y).reverse are either not actually lazy, or just really slow, not sure which one.
(DIR) Post #AdcL5j7g44qBM1w7Zg by goldstein@im-in.space
2024-01-07T16:11:05Z
0 likes, 0 repeats
@zabbeer @mgorny freedom to link software with other software without doing dumb slow workarounds because GPLv2 is not GPLv3 compatible.https://eighty-twenty.org/2021/09/09/perf-addr2line-speed-improvementin our era of software that never actually gets “distributed”, GPL is worse than useless for most projects: it doesn’t stop corporations from using software internally (and they only ever use it internally anyway, since everything is web-based) and it hurts good actors by being incompatible with other open-source licenses, including itself.“just always use GPLv2+”: this + is a statement that I’m willing to distribute under any terms that FSF brands as GPLv4, which requres a lot of trust in FSF, which I don’t have. it also doesn’t solve the problem of linking GPLv2-only libraries with GPLv3-only libraries.“just always use AGPL”: it’s less useless than GPL, but is still easily circumvented by keeping AGPL code in its own microservice, never seeing the outside world, so never “distributed”.
(DIR) Post #AdcL5kBy5brofcx3tw by goldstein@im-in.space
2024-01-07T16:18:50Z
0 likes, 0 repeats
@zabbeer @mgorny if you want to stop corporations from using your software, just use a non-commercial license, it’s that easy.
(DIR) Post #Ait6VUWArsGUAtPo5Q by goldstein@im-in.space
2024-06-13T14:21:26Z
0 likes, 0 repeats
@feld @drq cc @mpldr nope, still not a subset, YAML 1.2 uses different unicode escapes: 𨭎 is "\ud862\udf4e" in JSON, but "\U00028B4E" in YAML. Also YAML requires keys to be shorter than 1024 “unicode characters”, while JSON doesn’t impose this restriction (although it allows implementations to impose restrictions on key length, so it’s a weaker argument).there’s also a problem of “there’re basically no conforming YAML implementations” (see https://matrix.yaml.info/), so e.g. Python’s ruamel-yaml breaks down on '{"a"\n :"b"}', which is totally valid JSON (PyYAML doesn’t support YAML 1.2 and has similar bugs).
(DIR) Post #AqGT2BIcYolervZkGG by goldstein@im-in.space
2025-01-17T17:43:23Z
1 likes, 0 repeats
every rabbit hole that leads to ISO-managed stuff seems to end with extremely unsatisfying “the answer exists, but obtaining it is not feasible due to ISO being ISO”there’s this standard, ISO/IEC 7812, which describes payment card numbers. in theory, the first digit of the number identifies the industry card belongs to, with 4, 5, and 6 being reserved for banking. but a bunch of cards, notably Mir and also some new MasterCards start with 2, which is reserved for airlines, “financial” and “future assignments”. so one of the following seems to have happened:maybe ISO ran out of 4-6 prefixes and started giving out 2-prefixes to banksmaybe these 2-prefixes were not meant for banks, but were acquired somehow bypassing ISOor maybe this “future assignment” happened and now 2 is a valid prefix for banks.ISO’s contractor CUSIP has the full list of assigned prefixes, but it’s not public, because ISO. I’m kinda wondering if they’ll answer to my email if I ask anyway though.
(DIR) Post #AqGT2SYUP4YuO489r6 by goldstein@im-in.space
2025-01-17T17:46:07Z
0 likes, 0 repeats
UzCard seems to have 8600, which is in theory reserved for “healthcare, telecom, and future assignments”, and again, there’s no way to check how that happened, despite information obviously existing somewhere in the guts of ISO bureaucracy. maddening.
(DIR) Post #Arqa4OLBCi6Mwut1zk by goldstein@im-in.space
2025-03-08T14:11:12Z
0 likes, 1 repeats
so I’m kinda looking for a jobmy current position is ok, but it’s trading-related and doing trading-related stuff is slowly killing my soul, so I would like to do something more useful to societymy CV is at https://goldstein.lol/cv.pdf, but TL;DR: my primary language is Rust, but I’m also willing to work in basically any functional language or maybe modern C++. I’m not very proficient with either of those, but I’m willing to learn, so if knowing the language perfectly isn’t the main requirement for the position, I’m still interested.if you know a company that might be interested, please, send this their way!my residency situation is weird, so I’m looking for a global remote position. I’m not a EU or US resident currently, and I’m not willing to relocate. I’m very much open to discuss how to formalize a remote agreement so it’s convenient for both sides though.
(DIR) Post #AxSfklSmL81xd636cC by goldstein@im-in.space
2025-08-23T12:34:14Z
2 likes, 1 repeats
PSA: you can disable native #xwayland support in #sway and replace it with https://github.com/Supreeeme/xwayland-satellite, which properly supports scaling so X11 apps don’t look blurry