Quoting: ShabbyXAre we no longer too few to ignore? :)

I blame AI.
All those AI training servers run Linux and NIVIDIA is raking in big bucks, because of them.
A better driver might just what keeps in front of the competition.

View this comment - View article - View full comments

I think its really nice of world of goo that they were willing to actively keep Linux support when they already rake in the big Epic bucks.

I'm not a fan of Epic, but I think its really brave to go up against 3 industry titans at the same time.
It's also impressive that they're currently winning 2 of those 3 battles.

Also now it's not an upgrade or update. Now it's an enhancement.
I'll be pedantic and try to differentiate.
An update tackles problems with an by external events changed priority.
An enhancement adds previously planned/announced improvements.
An upgrade adds improvements so large that the product moves up a price category.

View this comment - View article - View full comments

Quoting: Ali_JohnI have M1 Max with 32 gb and LoL on MacOS does not have Vanguard. Even with the Metal option on(which gives great FPS improvement over OpenGL), the OS is so garbage that if you move the camera game goes from 144 Frames to 100ish and keeps jumping up and down. I have to restart the system just so that OS does not get weird with it's resource management system. MacOS being supported while Linux not is more like a side-effect that we are playing the windows version. If someone can find a way to run MacOS build on Linux we might be fine.

The relevant project for that is theoretically speaking darling, but if I hear you correctly it seems to run on Rosetta, which means that making a port could be achieved with Wine and a good Arm emulator(or an ARM device).

Further research shows: LOL runs on emulated the emulation layer of Rosetta and probably also on the wine part obtaining the relevant x86 binaries could be enough.

View this comment - View article - View full comments

Quoting: Philadelphus

Quoting: LoudTechieOoh I know the answer.
You can just include a piece of data you also include in the drm of the game in the nft and sell it from a developer approved wallet/signing key(means the same in this case).
This way the blockchain itself is the proof as long the first wallet it came from(which by design is tracable in most blockchains) is the approved wallet, the data is the same as the hardcoded value and the current owner has the key of the last wallet we know it's legit. We still have to check if multiple players have the key of the currently holding wallet, but that can simply be achieved that the wallet contains sum too attractive to steal or by embedding some kind of hardware id hash in the transaction.

The blockchain is that monitoring database you talk about. It's inneficient, because for it to be trustworthy thousands of copies of it have to exist. It's big, because it tracks much more than only your ownership. Making changes is intentionally expensive, because otherwise everybody would be doing it, but it's possible.

This is the one thing blockchain can actually do. Assign ownership to a single instance of arbitrary data it can still be copied, but the copy can be recognized as such.

Interesting, thank you! This hypothetical DRM would have to be Internet-connected so it could check the state of the blockchain every time you try to play the game, right? So you can't copy it to an air-gapped machine while you own it, then keep playing after you've sold it?

Oh, totally and to avoid tampering root permissions still have relevance.
Another win for this hypothetical blockchain based drm is that it diminishes one of the many many problems with drm a little though. It could allow for a game that is playable not so long as it parent company supports it, but so long as it parent company supports it + the blockchain exists.

Also you better hope partial downloading is actually supported for your blockchain otherwise you will be paying those chain checks in golden internet premiums.

View this comment - View article - View full comments

[quote=Philadelphus]

Quoting: tarmo888But OK, maybe instead of items in games we could buy and sell games themselves. It certainly sounds like an appealing idea. But how would it actually work? As a permissionless database, everything in it is public, and anything traded on it can be seen by anyone. You could trade some sort of "token" that you now "own" a particular game, but…what does that mean? You still need to download it from somewhere server somewhere in such a way that everyone else on the blockchain can't just download it as well (so you can't have the token include any sort of "key"), which means we're back to some sort of centralized database monitoring who "owns" each game to be able to dole them out (and of course, the games would require some sort of DRM to make sure you couldn't just keep playing them after you've sold them!). And if you have such a centralized database already, you don't need to add a blockchain onto it, you could just have that database (or multiple centralized databases with a hand-off mechanism).

I agree that Valve would like you to stay in their ecosystem for economic reasons, but if Valve and Epic Games suddenly both decided one day that they'd like people to be able to sell games between their platforms they could do it without any blockchain technology.

Ooh I know the answer.
You can just include a piece of data you also include in the drm of the game in the nft and sell it from a developer approved wallet/signing key(means the same in this case).
This way the blockchain itself is the proof as long the first wallet it came from(which by design is tracable in most blockchains) is the approved wallet, the data is the same as the hardcoded value and the current owner has the key of the last wallet we know it's legit. We still have to check if multiple players have the key of the currently holding wallet, but that can simply be achieved that the wallet contains sum too attractive to steal or by embedding some kind of hardware id hash in the transaction.

The blockchain is that monitoring database you talk about. It's inneficient, because for it to be trustworthy thousands of copies of it have to exist. It's big, because it tracks much more than only your ownership. Making changes is intentionally expensive, because otherwise everybody would be doing it, but it's possible.

This is the one thing blockchain can actually do. Assign ownership to a single instance of arbitrary data it can still be copied, but the copy can be recognized as such.

The reason why this will never work. Is, because it requires active work from the people who benefit from a lack of a second hand market and because nobody trusts anything blockchain related for good reason.

View this comment - View article - View full comments

Quoting: reaperx7The ignorance of their developers is staggering. Saying nothing can be done on Linux to mitigate this and that is hogwash drivel. They're just bad developers with no talent.

They say "We can't control what goes on in a Linux environment". Well, you developers can't even control what goes on in a Windows environment either, so pot meet kettle. Cheater are NOT hacking the binaries anymore. Those days are long gone. Hackers are modifying data packets being sent to servers which falls outside what Vanguard and other anticheats do. All anticheats handle is processor and memory address tagging. They lock the process to a CPU core and then lock the memory address used. However, that have NEVER targeted the data packets sent to the server. Why? Because Vanguard is a client side anticheat, not a server side anticheat.

By contrast, it's easier to defeat cheaters in Minecraft than it is on League of Legends. Why? Because Minecraft uses, with PaperMC and similar servers, serverside anticheats which scan for inbound data packets and block functions that would be on the blacklist of allowed executions. It's 1000x harder to defeat Minecraft anticheats because once they update, trying to rebreak the anticheat to re-enable a hacked client function, can be months out or longer.

These developers are pathetic and making excuses only peoves how inept they are. They could easily link against AppArmor to prevent execution and memory address tampering and implement a server-side anticheat and Linux would be 100% playable and they'd see less hackers.

According to their post they've already moved the entire game server side and the client is just a requester and the only thing they're still fighting is scripting.
Scripting can't be defeated with moving the game server side, because it can use standard input controls, just inhuman fast.
One could implement server side pattern recognition, but that is a really expensive way of doing anti-cheat, because you're basically reverse engineering all the known bots.
This's what Minecraft does and can afford, because major parts of the gameplay are really hard to script in the basis(it's creativity reliant game with way too much content that randomly spawns).
LOL is a lot more scriptable, because scripts can be essentially:
if(warning)
{
dodge();
}
else
{
act();
}

Which according to me is a mistake and can be easily fixed by adding more hard to scrip elements in the game(some don't even have to be noticeable by the player, such as not sending coordinate information to a client that doesn't do business logic anyway, lets see these scripters implement unnoticable image and sound recognition.)

Quoting: MarrondThat could be a valid argument however there is one little caveat they like to ignore - Vanguard already is implemented in Valorant and IT STILL HAS CHEATERS... it's as disingenuous as Epic saying no EAC support for Linux in Fortnite (not that I care about that game but it's about the principle) as if EAC was preventing any cheating on Windows to begin with. It's nuts.

That's addressed in their post linked at the bottom of the article they claim that their metrics show that it has a lot less cheaters.

What I find most shocking about this's not that Riot games doesn't care about Linux(duh), but that they're planning to demand TPM2.0, because they found Microsoft's Windows 11 requirements not draconian enough and that they brick keyboards if they can be used for cheating(yes not crashing the game, bricking the keyboard)

View this comment - View article - View full comments

This is too good to be true.
After the release of the Steam Deck I've seen nobody blunt their teeth on either Linux or Steam Deck support. I've seen developers aiming to become Steam Deck verified, but support, no.
This would be the first in nearly a decade to try it.
They will fail.
I don't know what will make it hard, but although they might get it working they won't be able to consistently support it.
Also they're way too optimistic with this roadmap.
38 features in a year for one game.
Yeah some overlap, but nah won't happen.

View this comment - View article - View full comments

:woot:
I blame it on the Q2 2023 revenue drop. New paths being explored.

View this comment - View article - View full comments

Quoting: ShabbyX

Quoting: LoudTechie

Quoting: ShabbyXNever underestimate how awesome Faith is! I'm excited to see what major issue she'll tackle next after nvk is done and left to others to maintain.

Based on her latest post on the matter [External Link] Maxwell support and performance.

I mean, that'd be great but at her speed it'll take, what, 1 month? :P

I'm thinking more long term, like overhaul the kernel graphics subsystem so all the problems are magically gone, that sort of scale.

You're completely right.
I tried to include the right link, but ehm sorry.
This is what I meant [External Link]

View this comment - View article - View full comments

proprietary standards+open source clash.
Frustating to experience. Fun to watch.

View this comment - View article - View full comments