I doubt this’ll be well received, but I actually don’t think Silksong should be used to set price expectations. Hollow Knight made a shocking amount of money, massive sales were guaranteed, and the tiny dev team has enough money to pretty much vibe and make cool stuff forever.

Please don’t compare other indie game prices to this, when those games can’t guarantee their financial security, or massive sales number to turn a profit regardless of price.

Also, unrelated, but reading through the Bloomberg interview, and knowing what they charged for HK, 20$ is actually exactly what I assumed Silksong would cost well before it was announced, the shock for that kinda caught me off guard.

Ugh, it’s incredible. Hollow Knight has been one of my favourite games for years, and thus far my expectations for Silksong have actually been exceeded, not just met. Utterly loving it.

Seemed flawless besides the external controller issues. It’s already steam deck verified, after all, so hopefully this is a quick fix for them.

But yeah, until that patch comes, I’ll be forcing proton :/

Yes! I had that issue as well. I managed to sidestep that one by screwing with gamescope, scopebuddy, steam input, switching the controller to 2.4GHz, and restarting the game several times.

Once I unlocked Sprint everything fell apart again, and that’s when I did some googling and realized I was running the native Linux version.

Yeah, proton is a bit of an odd dual edged sword like that. Obviously the dream would be the Linux market share getting large enough that it’s a no-brainer to focus on that version and make it as excellent as possible, and proton is essential for that, but at least for now, proton is so good that it makes it hard to justify a native version.

If you can’t maintain a high standard of excellence for your Linux port, savvy players will just use your Windows version through proton anyway, because it’s already a high quality port. Easy to understand why many studios forego a native Linux version altogether.

This was also the solution to me for a weirder problem, running on Bazzite with an 8BitDo Ultimate 2, I was sprinting randomly, especially when cresting ledges, and the dash button was inconsistent.

Extremely frustrating, the game feels significantly better with sprinting working as intended via Proton (I used GE-latest, but I assume it works with most proton versions). Would be nice to see the native version fixed, but proton is perfectly fine for now, and “external controller on Linux” is likely a lower priority bugfix.

The way I imagine it in my head is like a text autocomplete trying to carry on a story about a person talking to a brilliant AI.

If something is real, of course the hypothetical author would try to get those details correct, so as not to break the illusion for educated readers. But if something is fake (or the LLM just doesn’t know about it), well of course the all knowing fictional AI it’s emulating would know about it. This is a fictional story, whatever your character is asking about it is probably just part of the setting. It wouldn’t make sense for the all knowing AI in this story to just not know.

Obviously, OpenAI or whoever would try to prompt their LLMs to believe they’re not in a fictional setting, but the LLMs are trained on as much fiction as non-fiction, and fiction doesn’t usually break to tell you it’s fiction, but often does the opposite. And even in non fiction there aren’t many examples of people saying they don’t know things. I wouldn’t write a book review just to say I haven’t heard of the book. Not to mention the non-fiction examples of people confidently being wrong or flat out lying.

Simply based on the nature of human writing, I frankly wouldn’t ever expect LLMs to be immune to writing fiction. I expect that it’s fundamental to the technology, and “hallucinations” (a metaphor that gives far too much credit, IMO) and jailbreaks won’t ever be fully stamped out.

Unfortunately, I don’t think this would work.

The answer to where you should plug in is directly into your GPU, as streaming the data from your external GPU to your iGPU will cause data throughput issues as it has to constantly stream data back and forth through the PCIE bus. Even in simple games at low resolutions where that wouldn’t be an issue, you’d still be introducing more input lag. That’s why connecting your display to your motherboard is usually considered a rookie mistake.

But obviously, if you’re outputting from your external GPU, that silicon is still being used while rendering on the iGPU, which I believe would erase any potential power savings.

I think the better solution if you really want to maximize power savings, would be to use a conservative power setting on your main GPU, and do things like limiting your framerate/selecting lower resolutions to reduce your power draw in applications where you don’t need the extra grunt. Modern GPUs should be pretty good at minimizing idle power draw.

The problem isn’t the tech itself. Getting a pretty darn clean 4k output from 1080p or 1440p, at a small static frametime cost is amazing.

The problem is that the tech has been abused as permission to slack on optimization, or used in contexts where there just isn’t enough data for a clean picture, like in upscaling to 1080p or less. Used properly, on a well optimized title, this stuff is an incredible proposition for the end user, and I’m excited to see it keep improving.

I’m down for uh… one tiny part of this. I certainly think we could do to make games smaller, I’m sick of massive open worlds and colossal play times, which seem like an astounding amount of developer time to make swathes of stuff that ends up so soulless that I don’t want to play it.

More focus on fundamentals, shorter, more meaningful campaigns with well executed gameplay and ideas would be wonderful, because we’re rapidly finding the limits of every studio on earth trying to make the “forever” game. Players only have so much time.

The best recent example I have is Mario Kart World. It’s a marvellous game, wall and rail grinding are amazing, the tracks are some of the best in the franchise, it’s fantastic. But you can tell a massive amount of effort and years went into the open world, which uh… actively makes the game worse? Free roam is fun for an hour or so, but I have no idea why I’d want to do it with friends, and the game shoves its 200+ “intermission” tracks down your throat constantly. Time trials are the best mode in the game, because it’s the only real way to consistently play the excellent tracks enough to actually unpack and learn the shortcuts and tricks that are afforded by the game’s deep new mechanics. I feel bad that the team wasted so much time on something the community begs for better ways to avoid.

Oh, does it? I was literally thinking to myself that Teardown was an interesting example of destruction, and wondering how they did their lighting. RT makes perfect sense, that must be one of the earliest examples of actually doing something you really couldn’t without RT (at least without lighting it well).

But yes, agreed that recent performance trends are frustrating, smearing DLSS and frame gen to cover for terrible performance. Feels like we’re in a painful tween period with a lot of awkward stuff going on, and also deadlines/crunch/corporate meddling etc causing games to come out half-baked. Hopefully this stuff does reach maturity soon and we can have some of this cool new stuff without so many other compromises.

The big benefit of raytracing now, imo (which most games aren’t doing), is that it frees games up to introduce dynamic destruction again. We used to have all kinds of destructible walls and bits and bobs around, with flat lighting, but baked lighting has really limited what devs can do, because if you break something you need a solution to handle all the ways the lighting changes, and for the majority of games they just make everything stiff and unbreakable.

Raytracing is that solution. Plug and play, the lighting just works when you blow stuff up. DOOM: TDA is the best example of this currently (although still not a direct part of gameplay), with a bunch of destructible stuff everywhere, and that actually blows up with a physics sim rather than a canned animation. All the little boards have perfect ambient occlusion and shadows, because raytracing just does that.

It’s really fun, if minor, and one of the things I actually look forward to more games doing with raytracing. IMO that’s why raytracing has whelmed most people, because we’re used to near-flawless baked lighting, and haven’t really noticed the compromises that texture baking has pushed on us.

Honestly, the delays have increased my hype more than decreased it. I’m not one to obsess over a release, I’ve played other things and enjoyed them in the interim, so I really have no resentment for the long dev cycle.

Lately my habits have been to try to avoid games for a couple months to let them get polished up anyway (I recently regretted picking up DOOM TDA at launch after they reworked combat across the whole game, and that would’ve been a better first playthrough experience). Team Cherry is a team I know can use time well like that, in fact, HK did get broad balance overhauls before I discovered it. They also added an astounding amount of well integrated post-launch content, so I’m excited to see just how much they’ve managed to create and polish Silksong with all this time, and will feel comfortable playing at or close to launch now due to these delays.

Exactly what I’ve done. Set my settings to hide NSFW, blocked most of the “soft” communities like hot girls and moe anime girls and whatever else (blocking the lemmynsfw.com instance is a great place to start), and I use All frequently. That’s how I’ve found all the communities I’ve subscribed to, but frankly, my /all feed is small enough that I usually see all my subscribed communities anyway.

Hard to blame them. Proton is dang impressive, and if it works, it works.