Unfortunately given how hard reproducible builds are they aren’t done much, and aren’t talked about much. A vicious cycle. A nice short video would indeed be helpful for understand and awareness.

Yeah that site is pretty good. There’s a lot of information though. I think a good starting point is maybe this page: https://reproducible-builds.org/docs/env-variations/

An important point to add for someone who hasn’t heard of reproducible builds before: The key difference to a normal build process is that it is 100% deterministic i.e. it produces exactly the same output every time.

You might think that most built processes would be like this by default, however this is not the case. Compilers and linkers usually have some non-deterministic values that they put in the final binary such as timestamps. For a build to be deterministic these sources of variation must be disabled or mocked to be a repeatable value (i.e. not based on the actual compile time).

Would a game that is essentially a micro Linux distro count? I feel that should be pretty doable as a bootable USB stick or CD.

If you did it that way you’d have to bundle the Linux kernel plus graphics drivers at a minimum. But I wonder how much of the OS you could avoid having. Certainly you wouldn’t need a Desktop Environment. I wonder if you would need something like X or Wayland or if you could get away without that (to run games built in a normal-ish userspace way). I guess finding the minimal environment for SDL would be a good starting point. That sounds like an interesting exercise for sure.

Although something like that probably isn’t as pure as you’re looking for, it would be pretty cool to do anyway. Maybe we should start a club.

I found this good video by Trace Dominguez. He gives a good overview and also mentions a bunch of new studies that are being done https://www.youtube.com/watch?v=dpjGLLbWZJ0

Memes of the past!? Isn’t this meme like 4 years old. I feel old now

This is bizarre. Snap has improved a tiny bit over time, but it continues to not be that great. Meanwhile, flatpak is miles ahead. Things are generally just smoother and less annoying, even when Snap is working as intended.

Personal anecdote: I was having no end of trouble with Inkscape, it was just not working, very unreliable, all sorts of very odd issues. It got worse and worse over time to the point where it didn’t even seem to understand paths to open files anymore, if it even felt like opening that day. I tried reinstalling, clearing the config, all sorts of things. I suspected maybe the version of Inkscape Snap was giving me might have a bug in it so I was looking around for alternative ways to install an older version and then for some reason I tried Flatpak. It was like some kind of magic. Totally night and day. All of a sudden Inkscape had absolutely none of the issues that the Snap version had. It just worked. After that I realized that it hadn’t been a bug in that version of Inkscape at all, it was just Snap.

I haven’t had any issues with any other Snaps, but that incident really opened my eyes to just how bad things can get if a program isn’t packaged correctly.

Something I never seem to hear explained: What IS long COVID. Is it damage to lungs, is it a change in the behavior of the immune system, is it something that happens in the cells? Where in the body is it hiding? Is this something we just don’t know yet?

It would make sense that eventually you could do both real-time and after-the-fact calculations depending on whether real-time communications is available. Presumably it will depend on the specific application

In the hamburger menu on the left you should find an “Add account” button. IIRC it says the word “Anonymous” and you have to tap it to show the add account button.

Some things in Jerboa are a bit confusing for now. Once you get to know where things are it’s pretty good though.

Mm indeed. Types to the rescue.

Are there any projects/tools that turn the types into Java-style web html documentation? If you need to know something while you’re working on a non-typescript project that could be more convenient.

The issue I see a lot in the JS ecosystem is laying out documentation like a reference guide, but then not including all parameters or functions. These types of documentation are very helpful if what you need to know is included, because they have nice friendly explanations and examples. But eventually you will run into a parameter that is mentioned on Stack Overflow, or is in a code snippet in the documentation, but then has no further explanation in the documentation, as if it doesn’t exist.

Projects where the README is the only documentation seem to suffer from this problem the most. They give examples of the most common parameters and functions, and then that’s it.

In JS this is a big issue because there may be no way to know a parameter even exists, or what values it accepts, unless it is documented.

A lot of documentation in the Java ecosystem has huge auto-generated monstrosities with absolutely no explanations. In Java this is usually not useful because that information can be found in the types. But in JS it would be incredibly useful. Unfortunately it isn’t as easy to automatically generate that type of documentation for JS.

Python in my experience has the best of both worlds. It has the friendly explanations and examples. But also has all of the parameters, even if explanations for some are a bit less detailed. And all of that is combined into a single place.

Indeed. It seems this wouldn’t be useful for applications where real-time position is needed. You would most likely do the calculation at a later time like in the Miikshi video. It’s a little confusing from the article, but the video actually does a good job of explaining this limitation.

I think the most amazing part about this is the video at the bottom of the article: Miikshi: Cosmic Rays (4K). The caption calls it a “charming fictional animated video to explain their muon-based systems”. But I cannot emphasize enough how much this undersells it.

It’s like a weird charming mashup between Thunderbirds, Muppets and a real muon science team. You really have to just watch it.

If you had trouble understanding what the muon positioning system from the article might be used for or how it works there is a short explanation from Professor Tanaka at the end of the video.

For a Reddit/Lemmy app though I would still prefer images to load in the app itself, because it’s more seamless. For now Jerboa handles images and GIFV just like any other link.

RIF let’s you configure this, which is nice. For Jerboa I think the default should eventually be to load them in the app itself (RIF calls this “Native”).

Edit: Jerboa does already load the OP image in-app when you open the comments. I think that behavior shouldn’t change. It’s just when you open the image from the news feed that it opens it like a link currently.

When you open a link from Jerboa it should open a mini in-app version of your main browser. If your default browser app is set to Firefox it will say “Powered by Firefox” and if it is Chrome it will say “Running in Chrome”. UX wise it is very similar to having a bundled in-app browser. But the rendering is handled by your main browser.

I’m not up-to date with the story behind this, but my understanding is that new apps that want in-app browers should be implemented in this way. This seems to be the modern way of doing it. The reason for it I believe is mostly security. Your main browser app should be up-to-date on security patches and features, and if apps can just piggie back off that then they don’t need to worry about shipping and patching their own in-app copy of the browser. Also it respects the user’s choice of default browser. So if your default browser is Chrome it will use that for the in-app browser for all apps that work this way (which is quickly becoming most apps).

Another advantage of doing it this way is that when you use the “Open in Firefox” button it seems to just move the tab from one app to the other rather than reloading it, so it happens almost instantly.

I’ve been using Jerboa so far. It’s pretty nice. At first I was a bit sceptical because it seemed a little half baked. But it’s growing on me. Also now that I’ve used it for a while I think it’s relatively well rounded feature wise. It is quite minimal, but I think a lot of people probably like that. I think long term if other apps come up that are more similar to the existing 3rd party Reddit apps I will probably switch though.

My main issue so far is not being able to follow links to communities inside the app. I imagine this will probably be more difficult to implement than on Reddit because of federation, but hopefully apps can get that working eventually.

Also it can be difficult to add new communities from other servers even via search. Which is also a federation issue.

I also miss the features from Reddit Is Fun which let you choose which content is displayed in-app vs in a web browser. But that’s only a minor annoyance so far.

I find I am getting used to the visual style of comments. I initially didn’t like it, but it’s growing on me.

I think the metaphor is great. Calling an adult cat dead would indeed be extreme. But calling a perfectly working piece of hardware dead is also extreme. I wouldn’t call my toaster dead because the people who made it don’t make toasters anymore.

We’ve become far to accustomed in this day and age to accept that because something has software on it that it’s expiration date is decided by the company that made it. Cats don’t work like that, toasters don’t work like that, and neither should game consoles.

In practice that means modding is very important. And unfortunately many companies make that very difficult or impossible. But we should celebrate any product where that is possible, even at launch and its estore equivalent is still around.

Edit: dead cat not alive cat