Notices by linear cannon (linear@nya.social), page 11

i have outright deleted a major patchset i wrote for a project under freedesktop.org stewardship, which someone else is probably going to write again in a year or two, because i realized the project had a real-name policy, and decided it wasn't worth it. i then lost motivation for the cool thing i was working on that needed me to write that patch

this is not the intended effect of a "real-name" policy, but it is the actual effect. and, as the cool kids say, "the system is what it does".

there is no such thing as a "real name". the concept of a "legal name" is fraught, and most certainly is not what you think it is, or what you are looking for, if you are a software developer. many assumptions you have about what a "legal name" is probably are not true.

consider this: the name on my birth certificate is different than the name on my drivers license, and that is different from the names i am called by my friends. those names are all different from what is likely to be on my passport when i get it, and all of those are different than the name i publish my open source projects under. all of these, in different jurisdictions, might or might not be something you could consider a "legal name". which one do you want me to use when i submit a major feature to your library? are you going to turn me away if i try to submit it as "linear cannon"? why? if i have a website and contact information under that name, why does this matter? how is it substantially different than an author of fiction novels publishing under a pen name? does it change if i produce a piece of government-issued documentation with that name on it? why, or why not?

if your real name policy does not answer these questions adequately, then there's a very good chance i'm just going to assume that you're going to turn me away, as has happened to me several times already

i think it should be illegal for car dealers to use chatgpt to send me emails pretending to be a real person, that asserting that i've apparently talked to them over the phone several times over the past few months about my interest in purchasing a new motor vehicle

i feel a bit cheeky submitting a pull request to an open source intel project that could be summarized as "hey, maybe don't assume that any cpu architecture that is 'not arm', is x86_64"

but i'm also correct

i would like to see extensions to C/C++, supported by major compilers, that allow for SPMD programming a-la intel's ispc compiler

it would be nice to just be able to put in some extra directives and keywords to tell the compiler "hey, this next function has some different semantics, please make it go fast now"

though it would also be nice if folks who are using arch-specific simd intrinsics right now, instead used portable vector extensions supported in gcc/clang

programming language that literally doesn't have a concept of scalar values and everything (yes everything) is a simd vector

no but seriously give me a language where a for loop looks likefor (i: u32 from 0 to 999) { ... }and the semantics of the language mean, assuming you're compiling with avx512, that "i" becomes a 16-wide vector of u32s - 0 to 15 in the first iteration, 16 to 31 in the second, etc. except on the last iteration, which reduces the vector size and/or masks operations appropriately

and it's not just implicit. i is actually a vector of u32s here at the language level. you can get the length of i if you want to. all the operations you do apply to the whole vector by default. operation masks are a first class thing in the language so you can mask out certain elements for operations without having to use branching control flow. etc etc etc

that said, i'm not sure how i would structure a self-sustaining project to be resistant to this without resorting to a "bdfl" model. i don't think this is the right answer for a lot of large-scale collaborative projects, but i don't know what is, that would remain resistant to infiltration

open source software projects and standards organizations need more people willing to say "no. fuck you" to corporate entities, sponsors, and other bad actors putting on a polite face.

i think the reason linux works so well is precisely because the maintainers will happily tell a corporation off and reject huge amounts of work out of hand, if they aren't actually making things better. much as they may try, you cannot buy your way into making linux do something for you. you can buy insurance that it will continue to exist, and you can buy labor to submit improvements that benefit you, but your money will not afford you any lenience, and it will not direct attention to your own goals

i wish i could see more of the same elsewhere

other girls: "would you still love me if i was a worm"

me: "would you still love me if i was right next to you and i suddenly turned into a spoon. like. a regular-ass spoon you use for eating. except i'm made of pressurized solid atomic radon, and over the next few milliseconds i am likely about to violently depressurize and give you a dose of who knows how much radiation

@lanodan@queer.hacktivis.me i'm paranoid and care very deeply about privacy concerns, i use weirdass computing setups and encrypt my data, i use vpns, i've been known to use tor, etc. but i don't like. do anything interesting, and if they started tailing me i think they'd lose interest quick

because behind the curtain i'm just an anxious transgirl who grew up in an environment that made me value controlling the availability of information about myself

@lanodan@queer.hacktivis.me if you're using vga or dvi, you only have one monitor, and i know approximately where in your home your computer is, font size would likely not be an issue for me (once i figured out your resolution and refresh rate)

confound any of those, and it probably would be. i got enough out of an HDMI stream to recognize my desktop, but not read text in my terminal. but i'm also just a random transgirl with a cheap SDR who was doing a personal risk assessment with open source tools plus something i hacked together myself, in order to sate my curiousity and light paranoia/ocd

i'd say there's a few ways of looking at it, and where i sit is: the fact that i can get that far without much effort is worrying, and a three letter agency has a lot more time and resources they could devote to things like "writing a proper hdmi/displayport decoder with an sdr backend" or "figure out the signal processing bullshit you need in order to separate two display streams". simultaneously, i would be baffled to learn that they were using that against me.

that said, i don't think i'm a particularly interesting target

if there's a three letter agency tailing me, i'm pretty sure they're having a very slow day and/or have made deductions about me that are simply untrue

not me being smug about firefox not being vulnerable to the gpu browser timing attack, while the fbi agent in the "catering" van at the bar across the street is TEMPESTing my monitor

jokes aside, i've used an SDR and TEMPESTed my own setup just to see if it was plausible

it's noisy, it's monochrome, but it's good enough, and if an fbi agent wanted to know what was on my computer screen for some reason, a van across the street with a cover story and a fancy antenna array sure could do it

misskey update made me re-do account setup and, in the process, turned off follow requests, and auto-accepted all pending ones

and i am NOT happy about it

@jamey@toot.cat i think this is the first time ive ever run into a problem where a DSL is the right solution

i think ill have fun building it, but i also think ill need to be careful

i dont want this to grow outside the very specific thing i need it for

it' s always fun running a big build/update on a machine with hundreds of cores

it's kinda fucked up that we live in a world where i can spin up a blob of statistical data on my own personal computer and have a coherent conversation with it