If you’re doing it right, containers are less like VMs and more like cgroups. If orchestrated correctly it uses less system resources to run lots of services on a single system/node.
That said, I’m a devops/infrastructure/network professional and not a developer, so maybe I’m missing something from the dev experience… But I love containers.
Docker does kinda suck now, though. Use podman or another interface instead if you can help it.
I’ve seen many a terrible containerized monolithic app.
I’ve seen plenty of self-hosters complain when an app needs multiple containers, to the point where people make unofficial containers containing everything. I used to get downvoted a LOT on Reddit when I commented saying that separating individual systems/daemons into separate containers is the best practice with Docker.
Separate containers works like a dream when one app starts shitting the bed, gets auto-cycled, and everyone else just chills. Not surprised on the Reddit downvotes though. That place is so culty, especially now.
That’s one of my dilemmas. Due to using BSD and relying on jails I have a hard time using lots of possibly nice apps being released nowadays because they only offer the docker way of installing.
I don’t like containerisation because it leads to bullshit like atomic distros. I don’t want a spicy Android.
Steam OS has some cool elements like the menu, the in-game side panels and the game mode/desktop dichotomy, but incremental rolling release is utterly deranged from my POV as an Arch user (btw).
Say Plasma “breaks” - a wiki will not help fast enough.
I’m not trying to defend rolling release for a gaming console, but give me at least the option to decide for myself whether I’m ok with breakage or not. There is this kernelspace NT driver that I wanted to try, but I couldn’t because pacman is locked.
I’m currently trying Fedora Kinoite and from the get go the hassle of getting a proper Firefox+codecs to watch online videos feels like a major step back.
Then you have the issue of installing software in flatpack (is: vscode, texmaker) that are either not fully working of need to have their access tweaked. Atomic distros appeal is to “just work” it doesn’t seem like it does.
Doesn’t it steal control flow? More like a break point, except you define where execution continues.
I wonder if it’s a compile error to have multiple conflicting COMEFROM statements, or if it’s random, kind of like Go’s select statement.
How awesome would it be to be able to steal the execution stack from arbitrary code; how much more awesome if it was indeterminate which of multiple conflicting COMEFROM frames received control! And if it included a state closure from the stolen frame?
I'd say it's more like setting up a handler for a callback, signal, interrupt or something along those lines.
Function declarations by themselves don't usually do that. Something else has to tell the system to run that function whenever the correct state occurs.
That doesn't account for unconditional come-froms.¸but I expect there'd have to be a label at the end of some code somewhere that would give a hint about shenanigans yet to occur. Frankly that'd be worse than a goto, but then, we knew that already.
Its like if subroutine bar could say its going to execute at line N of routine foo. But if you were just reading foo then you’d have no clue that it would happen.
You can simulate this effect with bad inheritance patterns.
A function will be called by code and go to that point in code. To implement functions, you store necessary things to memory and goto the function definition. To implement that with comefrom you’d have to have a list of all the places that need to call the function as comefroms before the function definition. It’d be a mess to read. We almost never care where we are coming from. We care where we’re going to. We want to say “call function foo” not “foo takes control at line x.”
IPv6 is already backwards compatible though. There’s a /96 of the IPv6 space (i.e. 32 bit addresses) specifically for tunneling IPv4 traffic, and existing applications and IPv4 servers Just Work™ on IPv6 only networks, assuming the host operating system and routing infrastructure know about the 6to4 protocol and are willing to play ball.
Oh nice. Does your system FINALLY provide enough addreses for every Planck volume in the observable universe? It’s been frickin amateur hour, this internet thing.
I’m enjoying the idea of someone looking at the picture, looking at their cords, looking at the plugs, and then working this out as a solution rather than sending a picture of the plug end and port and asking for clarification.
Like this person probably felt really stupid needing to sort out what’s going on, because it was just unclear, and I enjoy where they ended up all the same. It’s totally wrong, but it’s creative problem solving for sure.
Reminds me of every fuckin PR I do for the Indian contractors that were sold to us as “senior devs” but write code like a junior and you better believe every other line has the most obvious fucking comment possible
Better than writing beginner level crap that is at the same time super cryptic and not documenting at all. We have a bunch of that in our codebase and it makes me wonder why these devs are writing extension methods for functionality already built into the standard libraries.
Better than writing beginner level crap that is at the same time super cryptic and not documenting at all. We have a bunch of that in our codebase and it makes me wonder why these devs are writing extension methods for functionality already built into the standard libraries.
programmer_humor
Top
This magazine is from a federated server and may be incomplete. Browse more on the original instance.