Hacker Newsnew | past | comments | ask | show | jobs | submit | ranger207's commentslogin

A lot of scheduler experimentation has been enabled by sched_ext: https://lwn.net/Articles/922405/

IMO Apple and Valve are taking the opposite approaches but on a different axis than the article discusses: Apple is continuing to increase their lock in and remove choice, while Valve continues to add choice. You can argue that Steam being a nigh-monopoly means there isn't a lot of choice, but I'd argue that's not correct. For one, Steam rarely censors games (it does happen! A notable case happened this month! But it happens rarely) and doesn't have requirements for games to use Steam's platform technology to be on the market. In fact, you're allowed to offer direct competitors to Steam features in your game without penalty (some games I play have both Steam Workshop support and the game dev's own mod platform support). For another Steam doesn't try to nudge you towards their solutions constantly either (eg like in the recent article on passkeys where the user had to click half a dozen times to not use Touch ID with the Touch ID option being on every page of those clicks). And of course, there's the "Add a non-Steam game or app" button in Steam that just asks you "where's the executable" and then it gets all the non-platform features Steam offers, like the overlay, screenshots, Steam Input (I think it even supports community input profiles for non-Steam games; I'm pretty sure I've seen community profiles for Primehack on my Steam Deck), etc. Of course the Steam Deck (and now Steam Machine and Steam Frame) are constantly advertised as "it's just a PC and you can do whatever you want with it". There's no lock in; you can install competitors' stores on those devices easily.

The reverse playbook then is that Apple is trying to make every option other than staying in the Apple ecosystem a bad choice, while Valve is trying to make Steam the best option in every scenario. The difference in base philosophy is the important part.

(Of course as a profit-seeking corporation there's no guarantee they'll stay this way, particularly after gaben leaves, but I'll appreciate it while it's here at least.)


Gundam is several universes that are unrelated except they have giant robots colored in primary colors. The main universe is called Universal Century (UC), and started with the original Mobile Suit Gundam show in 1979. It's retroactively called 0079 ("double-oh seventy nine") after its in-universe date. It's good but has significant lumps, like the haphazard animation and villain of the week fillers. I'd recommend starting with one of the 90s OVAs: War in the Pocket (0080), Stardust Memory (0083), or 08th MS Team, as they're fairly short, self contained, and have fantastic animation. Alternatively a modern UC show is Unicorn, but there's a lot of references in it you won't get unless you've seen some of the older shows. Another option is the latest Gundam series, The Witch from Mercury. It's set in a new universe unconnected from UC so there's no prior knowledge necessary.


Witch from Mercury isn't the latest show, FYI. The latest is Gundam GQuuuuuuX (yes, really, that's the name). But I wouldn't recommend it as a first entry. Despite being set in its own universe, it's supposed to be a sort of alternate version of UC and assumes viewers are already familiar with the original.


Contrarianism leads to feelings of intellectual superiority, but that doesn't get you anything if everyone else doesn't also know you're intellectually superior


How do you guarantee your accelerationism produces the right results after the collapse? If the same systems of regulation and power are still in place then it would produce the same result afterwards


Typically, Docker and Kubernetes run container images based on Linux and run them on top of Linux servers. I can say from my experience using Kubernetes to run Docker containers that having separate Linux experience is very helpful


Good, move slow and avoid breaking things


If your config language doesn't have its own control flow then it's going to get a meta layer added on with control flow. Like, what do you do if this service has a different hostname in staging versus prod? Or connects to a different DB, or whatever. Either have one template file and give it a values file when you go to deploy it, or have two fully-written out files... which becomes annoying when you have to make sure you keep them in sync across two or more environments, often leading to a hacked on and poorly implemented templating system anyway (eg, Helm)


If your program requires executing another program to prepare its options, then something has gone wrong in design or architecture.

Configuration is a type of API - and should be designed with the same care and goals including human legibility, orthogonality, “simple things should be simple to do”, surprise avoidance, stability, minimalism, etc.

Instead often what we get is little more than a sort of serialized form of the internal state of the system.


If configuration is an API, what's the harm in scripting against it?


Yeah too many times the Helm chart is barely less complex than writing all the manifests yourself because all the manifest options are still in the chart


Honestly from the domain I figured this was going to be a Xorg/Wayland/XWayland issue


Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: