CuriousCosmic

> but have they change something where "pointer to an array[5]" is a meaningful distinction to draw

This has been a meaningful distinction since at least C99.

> "a pointer to an array of X" is simply "a pointer to X"

They aren't actually the same. The former can be decayed into the latter but they aren't actually equivalent and you'll get type mismatches if you try to treat them as the same thing.

> who has families of functions with the same type signatures that they want to point to

An example is callbacks or handlers for responding to interrupt requests. A lot of hardware interface code relies on typedef-ed function decls because very often you are passing user side functions through your interface so that you can stash those function pointers somewhere and invoke them when some event occurs.

The trick is to just assert equal types. Most compilers have extensions that allow you to easily compare type equality (and C23 actually standardizes the `typeof()` operator).

So you basically take your ugly type, put it in a #define and then create a static assertion that matches the type against said ugly type.

Now the compiler will throw a shit fit if the types don't match. Have fun breaking the type up into smaller pieces until you have something legible.

the function signature certainly should be type def-ed. i.e.

    typedef char (*fn())[5];

and then you have the original as

    fn x [3];

That's a current wave going through. There's a bit of discourse going on where some users were rage baiting by constantly posting twitter screenshots over to bluesky and people who left bluesky are pushing back about how they don't want to see that content.

That discourse comes in waves each time a major migration happens from twitter to bluesky but it settles down fairly soon after each time. Give it a few days and people will have moved on to the new topic of the day.

Also it's worth noting that the "Discover" feed is trained specifically per person so while the defaults aren't great, if you use the "show more like this" and "show less like this" options on posts (under the triple dot) then pretty quickly it tunes in towards content you care about vs content you don't.

If you are looking for stuff from big publishers then yeah it's gonna be a bit slow/empty but there are a lot of small/indie devs on bluesky and it's awesome being able to interact with them directly without all the spam that usually came with posts on twitter.

Obtanium exists for a very specific use case.

1. You have an app you want to use.

2. That app isn't on the google app store or you don't want to/can't use google services.

3. The app is not open source so it can only be built and packaged by the first party.

4. You don't want to manually update the app by downloading a new APK every time.

5. You don't want to give a black-box closed-source app you downloaded from the internet permissions to install new apps (and therefore grant them certain new permissions as well).

My example of this is WhatsApp. I hate the app. I think it's scummy as shit. However if I want the version of WhatsApp that doesn't package google services, I either have to download a 3rd party app store, update the app from their web page manually, or grant the app permission to update itself. I obviously don't want to install a (often closed source) 3rd party app store just to install this app without granting it keys to the castle. So instead as I already use F-Droid, I can install the FOSS build of Obtanium and pin my trust on F-Droid. Then I use Obtanium to manage my WhatsApp updates.

Technically this also extends to open source apps where you trust the first party enough to use the app but not enough to let it update itself and where you want to be able to just download updates from github releases.

> Why isn't the usual Markdown (or the like) approach enough?

Because you can't use markdown to design CSS or SVG animations. I want to be able to design an animation with code that describes how the elements move/interact and then compile that down into CSS or SVG keyframes so that I don't have to manually declare however many arbitrarily complex keyframes.

I understand the purpose of a static site generator and I'm not looking for anything to replace that. Rather I want to be able to write the templates for my static sites without 1. writing exceptionally tedious HTML & CSS and without 2. relying on client side JS to do those tedious things.

Eh to a certain degree. The architecture of the models is very much control theory and then the training is control system tuning (which of course is an optimisation problem like you said).

I would definitely agree that optimisation fits the definition in part but I find really only control theory covers that entire field of signal processing, optimisation, and decision making systems.

And importantly, because ML in some amount touches on all of those, control theory tends to fit better as it focuses so heavily on providing a comprehensive framework for reasoning about all of those elements together.

Yep. Again, I'm not looking for a perfect global solution.

Just some domain specific language for writing sane low-to-no-js web pages or parts of web pages without having to manually fiddle with HTML or CSS in any real amount.

Static site generators honestly get me a lot of the way there but those are unfortunately template based which means any significant customisation requires dealing with the HTML and CSS rather than being able to just describe the layout and behavior I want.

Honestly I'd just write the compiler/lang myself if I didn't hate frontend so much.

> As to HTML + CSS: it goes other way around...

Yeah that kind of stuff is closer but most of those types of frameworks seem to lean heavily on mapping to JS because they try to be turing complete rather than just being easy to work with DSLs.

> - Browsers don't "talk" any other language but JS

I'm not looking for browsers to talk with any language. I'm looking for some DSL that directly maps to the 3 layout languages that browsers understand (i.e. HTML + CSS + SVG). Not anything turing complete at runtime but rather a sane way of describing a webpage layout with fancy styling, UI elements, transitions, and animations but without dealing with the pain that comes with actually writing in the native browser layout languages.

> You can't do "fancy web flourish" without manipulating DOM.

There is a lot of web flourish you can do without manipulating the DOM. It's not actually terribly unperformant to do but writing that code (mostly HTML + CSS, occasionally SVG) feels like peeling your eyelids with an unwashed lemon zester.

> ...

And for the rest of that, again I'm not looking for anything that actually executes in the browser. Just a sane, modern layout language that compiles down to static HTML and CSS with no JS or WASM (unless you explicitly ask for it).

Yep. I'm well versed with static site generators but every one I've ever worked with has been heavily template based rather than being an actual layout engine that map onto HTML+CSS+SVG. i.e. They all require you to still write in HTML+CSS+SVG rather than being a generalised way of writing HTML+CSS+SVG without dealing with the warts of those languages.

FWIW Retr0id is one of the big community bluesky/AT protocol developers and importantly the person who more or less figured this out (or at the very least brought it to the community's attention).

Point being that they were just clarifying the situation as they are probably one of the most informed people on the situation (with the other reply in this subthread being a first party bluesky dev and therefore one of the other SMEs on the topic).

Is there a reason why there seemingly aren't any frameworks to do all the fancy web flourish in a normal language and then have it compile down to plain HTML5 + CSS + SVG?

> not yet fixed, still being exploited

It is actually fixed. Tweets have been showing up with their "Indexed On" date instead of the actual authoring date. The thing that isn't implemented is showing both dates for tweets with separate dates.

Yeah machine learning is more or less just very complex application of control theory techniques and notably it is usually done by people without formal control theory backgrounds.

Super useful for control applications but obviously you really want to know control theory so that you aren't just using ML to throw darts at a wall.

I have to massively disagree here. I'm far more confident about code written in the modern style of C++ than I am in old school C++.

Modern C++ with RAII, smart pointers, constexpr, concepts, and heavy instrumentation is so much easier to reason about. It's actually close to Alexander Stepanov's idealised C++.

NOTE: This is not libcrux the formally verified rust crypto library (and one of the underlying crypto primitives libraries for Firefox's NSS).

The problem with Teslas almost universally is with lemons. Which is to say if you get lucky your experience will be great. The problem is that there are a fuck ton of lemons and cars that have no problems turn into lemons right around the time that the first few problems show up (which means more lemons as they age).

That's essentially just a QA and reliability design failure which is indicative of broader issues with Tesla but it also means that until someone's vehicle starts having problems they will have nothing but a positive experience which can leave them blindsided by said problems.

I don't think so. My passport is somewhere around the house so I can't just grab it to check but last time I came in they just put it on a scanner and took a photograph of me when I went through immigration.

Plenty of hackers with substantial social media presences but they live in pockets that don't cross over into mainline social media too much.

In most countries you can get what is called a notarized or certified "true copy" of identifying documents (passports being one of them). The intent is for you to submit them with applications so you don't have to submit the original copy. Now how you get a true copy depends on the country. Canadian passports for example can only have true copies made by their embassy or immigrations offices.

Note that this doesn't include your passport stamp pages but Japan hasn't issued passport stamps for several years now and they just look up your passport in a registry instead. So for that purpose, a true copy should be effectively the same thing.

That's where it comes down to what is essentially a technicality but given they carry an embossed seal and signature with the words "TRUE COPY" on them, they look very official and officials are very rarely going to push back on it even if whether they are to be treated as a full substitute for a passport (for identification purposes) is technically a grey area.

> what's at the root of this contradiction?

capitalism.

> what kind of change could incentivize different behavior?

worker ownership.

Sorry if that seems terse but really this is the root of the problem. Shareholder value is the end goal and shareholders are generally seeking immediate short term profits (as there's little cost to liquidating a share position and moving to a more profitable company/asset instead). If you want to eliminate this problem, you need the shareholders to be the workers. They need to control the company in a meaningful way because while shareholders generally only care about profits, your workers are going to care about continued employment and the stable maximization of wages and benefits (which as a long term target is only pursuable via real long term growth).

Yep. You can also do most of it or even all of it with a number of other tools as well.

IMHO the main sell is that you can do "all of the above" regardless of what domains you are working with without having to switch tools or negotiate different formats, etc.

So a systems engineer can model, simulate, and build a test framework for a project for all the bits (including the inside of the hardware and the physical world around it) without changing tools.

I have a non-commercial license that I use personally. The appeal of mathematica is that it has a library already built into it for just about everything which lets you quickly prototype and reason about fairly complex cross domain problems and then easily visualize them.

Mathematica unfortunately has a lot of warts and rough edges clustered in specific fields (low level cryptography and binary manipulation heavy algorithms) which makes working with it a pain in the ass if you touch those fields regularly but outside of those specific fields it's fantastic.

And I can claim praise doubly so if you get the System Modeler/Modelica license as well. With those two together you can model a project using analog circuits, digital circuits, software running on hardware/RTOS, pneumatics, hydraulics, CFD, multibody physics, etc all running in a unified environment. And then of course from a high level you can tune parameters that drive all the parts of your model and test out swapping in different parts or running under different conditions.

You can do all the stuff that Mathematica and System Modeler are good for with other software packages cobbled together but that takes a lot more time to do and the integrations tend to not be nearly as clean. The only other product package on the market that compares would be MathWork's MATLAB + Simulink. The main difference being that Mathwork's products are more comprehensive in what all they cover but they tend to be less pleasant to work with.

Japan still uses a lot of physical games media. Lots of little game stores still exist (or at least a lot existed before COVID, I know a lot of specialty stores closed up during that) and doujin circles have historically made the bulk of their game sales (and generally gained their fame) from physical sales made during large events like comiket in their doujinsoft sections.

And while in the west this sounds kind of like a niche thing, a substantial number of the bigger games that have come out of Japan (excluding those from the AAA studios) either started their distribution with physical sales at conventions or their developers started off that way before they got enough of a following to just publish digitally and avoid the hassle.

So there's not really a clear cutoff between "studios" small enough to avoid the ire of the law and those large enough to afford getting all the ratings fees, etc paid for.

It's far less effective because you have to hit an extremely small target to actually kill a python. Far easier and more reliable to just get up close and personal with it and either bludgeon, impale, or decapitate it after you get it pinned down.

That's not what happened. He said quote "which would allow RPs to block you, and something that I have previously rallied against".

This is something that has been proposed that Tim fought against but mentioned in the thread to provide context of the types of kneejerk reactions the spec authors have had to push back against.

> Sweating does not make you stinky. Sweat is not stinky. It is bacteria that causes the stinkiness. If your skin is (relatively) clean, there would not be any (food for) bacteria and you won't stink.

As much as I agree with your general point, this isn't strictly true.

For a sizable chunk of the population, sweat doesn't contain high concentration of compounds that when digested by bacteria produce body odor.

However, despite being a sizable population, people lucky enough to have this trait are in the minority. I don't know the actual percentage but among European populations it's as low as 2% and among east Asian populations it's as high as 50%. Either way less than half the population.

The rest of the population has variations of that trait and their sweat produces moderate to extreme amounts of amino acid based compounds that when digested by bacteria produces the VOCs that make up the infamous body odor smell.