rkido

Minor correction on the article: Kunrei-shiki doesn't have a 1-to-1 mapping between kana and letters. You can see that directly in the table they show. Only Nihon-shiki does. Kunrei-shiki is the worst of both worlds.

That's why Nihon-shiki is and will likely always remain the best system for serious students of Japanese - if you even need it at all. But then again, if you're just starting out, the best place to start is usually to immediately start learning kana.

While individual game developers choose whether to support macOS, the broader phenomenon of developers dropping macOS support is Apple's fault. They either have no interest in being a serious gaming platform, or else they've made catastrophic missteps in their quest to stay relevant for gamers.

It's not a cult, 40 is just a sweet spot between performance, battery life, and comfort for a large number of demanding 3D games. If you often play in docked mode, the battery life doesn't mattery so you can probably achieve 50 or 60 in some modern games.

The bigger the display, the easier it is to visually detect a difference in refresh rates

Why would users need to re-purchase everything just to adopt a different store launcher as their primary platform? GOG Galaxy for instance interfaces with most existing stores and friend networks. As long as you're buying games on PC, you're not really locked into much of anything.

That said, buying certain games can lock you into using Windows. So there's that. But it's far better than buying games on modern consoles, which force you to use some special-snowflake hardware just to access the game you bought.

It's impossible to completely prevent all classes of lock-in, but it can be minimized. I would argue that my personal lock-in is minimized, across the axes I care about, by buying everything from Steam and GOG. No hard Windows dependency, no silly hardware dependency, and I'll almost certainly still have access to my games in 40 years.

That Steam was a first-mover is basically irrelevant. PC gamers are a particularly demanding audience and they will abandon your product in a heartbeat if a vastly superior one comes along. For instance, TeamSpeak and to a lesser extent Ventrilo and Mumble more or less dominated voice chat for many years. Then Discord came along and those older products became irrelevant practically overnight.

Valve uses its 30% cut extremely effectively by continuing to invest in the happiness and convenience of its customers and, just as importantly, its developers. For instance the profits of smaller developers dramatically increased after the new discovery/search features landed, enabling exactly the kind of people who would love your game to find it easily. Contrast that with every other store where it's basically impossible to find anything but the most mainstream top sellers.

Maybe Epic could afford to actually make the Epic Games Store a useful product if they took a higher cut of sales?

The cost of translating input calls from Linux to Windows APIs through Wine/Proton is so much less than the cost of rendering a single frame that it's effectively 0. The only real overhead Wine/Proton might be adding is in the rendering pipeline.

As Steve explained in the video, the Steam Deck's input-to-photon latency is lower than the Aya Neo's simply because it has a much more powerful GPU.

The only real way to test if there's any extra input-to-photon latency would be to compare the exact same game on a Steam Deck running SteamOS to a Steam Deck running Windows.

What features does Windows have that Linux doesn't?

The only one I can think of off the top of my head is HDR support, but that wouldn't affect these benchmark results.

That's just because no one actually uses HaikuOS or RedoxOS (no offense to these wonderful projects of course). If these came even remotely close to the popularity of Linux, you'd have remixes, spin-offs, and eventually full-blown distributions.

You can try to prevent it with project structure, but even something really unified like FreeBSD ended up having derivatives like GhostBSD as well as full-blown divergent forks like DragonFly BSD that are not necessarily compatible anymore with upstream.

The Link hardware just wasn't necessary anymore once the software became good enough to run on any device as an app. I have Stream Link on my TV as an app now.

This is news to me. What is an example of a portable gaming PC that has flopped? I'm honestly just not aware of any major company trying to make one.

As for Valve hardware, the only clear flop was Steam Machines -- which weren't Valve hardware at all.

I think it's odd to focus purely on the mass market success of Valve's actual hardware (Link, Controller, Index) since these can all be considered experiments by a company that can afford to fail, and since these have all been widely praised as the best-in-category. Clearly they know how to make great hardware; they just haven't done anything mass market yet. Who knows, maybe the disappointment with Nintendo will fuel sales for the Steam Deck.

It's a 7" screen, so this resolution is comparable to an Apple Retina display in terms of ppi.

Playing PC games on a cell phone is painful. You can't even see half the game because your thumbs always obscure the screen to operate the touch controls.

The mere fact that NVIDIA Shield Portable didn't take off doesn't tell us much. It had poor design sensibility (i.e. it looked like ass, like some kind of portable Xbox designed in the early 2000s), and it ran Android, which lacks a compelling games library.

Steam Deck at least stands a chance, although I'm worried about the weight and the battery life.

> However an optional is really just a list constrained to size 0 or 1.

What?