Hmmm . . . seems like things are really switching around. But what's the time frame here and how solid is the expectation? When will we really be able to say AMD cards and drivers work, work well, maybe work better/more reliably than NVidia?
Because let's face it, for a long time now whether we like it or not NVidia cards and the NVidia proprietary blob has been the practical approach to Linux gaming. And I'm a bit vague on it, but I have the distinct feeling that there have been a couple of occasions when AMD prospects, like for good open source drivers, have looked good but not really followed through. Things look good again right now, probably better and more solid than on such previous occasions, but I am not knowledgeable enough to tell how close to a sure thing that is.

View this comment - View article - View full comments

That environment is amazing. So much detail, so much interactable stuff. Plus it's refreshing that, how to put it . . . a lot of game environments are sort of, I dunno, symbolic (and often sarcastic/satirical). They don't try to make a real place, they just give you the key stuff so you have an impression of a place, with a couple of over-the-top humorous details so you know "This is a Dilbert-esque office setting" or "This is a somewhat gritty dystopia" or whatever. And that's fine, I often enjoy that.
But it's interesting that it looks like for the most part here, they're working really hard to simulate, like, a real place, with all the things in it that an actual space station with offices people really worked in and so on might have; not streamlined and not parodied, just there. Makes a nice change.
There is something about the office clutter that seems oddly . . . clean, though. Edges so well-defined, or something. It's like all the effort to make it real has caused it to fall into a sort of artifact version of uncanny valley. Still, really cool.

View this comment - View article - View full comments

When you said "thanks to the game store itch" I thought you meant some kind of gamer condition where you get a yen to rummage around in game stores. :D

View this comment - View article - View full comments

Quoting: cRaZy-bisCuiT

Quoting: gojulI now hope Feral will release TW Rome 2, Shogun 2, Napoleon for Linux (and possibly Rome 1, but not essential)

Without cross-platform-support? Thanks, but NO thanks! They should fix the linux and the windows client alike to make it work.

Au contraire, should such things happen I would be saying thanks and thanks. Not everyone is into multiplayer. I'm not, for instance.

View this comment - View article - View full comments

Quoting: edddeduckferalThanks for the suggestions regarding MP, however any changes need to be merged into Windows (and Linux and Mac) easily and without effecting Windows (and Linux and Mac) performance and stability in any major way. It's not really the network layer but the maths used throughout the game engine which is then used to keep games in sync. We've tried many maths library replacements (which is how we got Linux/Mac working on many titles) however we're yet to find a solution that works on all three platforms, is easily implemented and has no performance or stability impact on all platforms.

We won't give up looking into solutions but this isn't easy, between all the games we've investigated these problems on we've easily spent months heading towards a full year of investigation on the problem. It's disarmingly complex and a single byte out of alignment and the entire house of cards fall down.

This post (linked to previously in this discussion) has a good overview of the problem.

http://steamcommunity.com/groups/maclinux/discussions/0/528398719787800607/ [External Link]

Note: I am ignorant. But it seems like a weirdly fragile way to do things even on a "single" platform such as Windows 7,8,8.1,10 . . . I know computers are supposed to be deterministic in how they work, but I've had too many occasions of performing the exact same action and getting different results to believe that holds reliably at current levels of complexity. Depending on things staying precisely the same (as opposed to say expecting they will occasionally diverge in some way and having mechanisms to re-synch) seems lacking in robustness.

Obviously this isn't you guys' fault, you just have to deal with what you find. I'm just wondering why this would be the way developers would do this stuff. Is it always like that, is that really the only way?

View this comment - View article - View full comments

Quoting: sarmadWell, we have to admit. Valve tried to make Linux a viable gaming platform but they failed miserably at execution and it seems none of the major players in the industry is taking SteamOS seriously. Very sad. Though they did give Linux gaming a great push.

All in all, looking at responses, I'd say it's clear we don't have to admit any such thing.
And really, although I have a critique of Valve's execution and marketing of the Steam Machines myself, this seems an exaggeration painted with a too-broad brush and not that relevant as a response to this particular event.

View this comment - View article - View full comments

Got it, thanks.

View this comment - View article - View full comments

Quoting: Guest

Quoting: cRaZy-bisCuiT

Quoting: pete910Just need some games now! {looks toward feral/aspry)

Please not only Windows ports - I'd like to see some well optimized native-cross-platform developed and optimized games as well. Of course a well optimized vulkan port will run better than OpenGL 4 - still it will take them a lot of time to get used to Vulkan. The first ports won't offer us such a great performance gain as many people expect - at least it's very unlikely.

Have to agree that something natively developed from the start is what we should be pushing for. Almost all the people who have so far touted the performance gains from Vulkan have had engines designed with DX11 in mind and tried previously to kludge OpenGL into it.
The only game I can think of that might be more representative of moving to Vulkan is Doom, which of course isn't available for GNU/Linux, but does come from a codebase more friendly towards OpenGL. Even then, they had access to special extensions made just for them, so it's not really something that can be considered a true comparison.

By this point, developers really should be using libraries that do away with 99.99% of platform specifics, and be cross-platform developing from the start. It makes sense even as a basic sanity check. Not that I think sanity prevails as much as it should, of course.

It's been mentioned hereabouts before, but the prospects would be better if Apple had gone with Vulkan instead of turning their noses up and rolling their own (Metal) as usual. At least Android is using Vulkan, which is huge although I'm not sure how much mobile development overlaps with the PC space.

View this comment - View article - View full comments

Looks like a pretty nice feature.

It is a bit hard for me to get just how it works though. I guess there's this one crucial moment that makes things unclear for me--when you're inputting the "gesture". So from the looks of it, you don't input a gesture by, like, recording it or something, you type in arrow keys to represent directional movement? So that brief moment when you appeared to type in an up arrow and maybe a down arrow into a dialog box, that was inputting the gesture? 'Cause I think that's really the key moment and it goes by in a flash so I was left scratching my head thinking "Did he do what I thought he did?" And that uncertainty distracts me so it's harder to twig to the next bit where you're specifying which input method the gesture applies to.
(Watching again I realize most of that stuff I said is totally wrong, so I definitely was confused)

Basically I think it's easier/quicker to do something yourself than to assimilate seeing someone else do it, at least if it's new to the viewer, so if you're showing computer stuff in a video it's probably useful to slow it down significantly relative to what feels comfortable in terms of just performing the action.

View this comment - View article - View full comments

Quoting: liamdawe

Quoting: kit89Doesn't SC stand for Steam Controller?

Well, the application itself controls the Steam Controller, so technically it really is a Steam Controller...Controller :D

So if you wrote an app that did things like apply packaged-up gesture schemes to SC, would that be a Steam Controller Controller Controller?

View this comment - View article - View full comments