I have one, but most of my games don't work with it (or work poorly).
For instance, Broforce just doesn't work (but Xbox360 pad does work).

Also, TF2 still displays the keyboard shortcuts although I have the SteamController plugged in. It isn't really easy to learn games this way.

Do you have the same problems?

View this comment - View article - View full comments

Quoting: liamdaweAdded in an option to hide your PC info (no link to it, doesn't show on profiles), that way you can still be included, but be 100% anonymous. From a request in here.

Thanks!

Just a small note: I was thinking of a more global option that would prevent our entire profile to be publicly available. Like Steam profiles: https://steamcommunity.com/id/MightyCreak. But that is already very good :D

View this comment - View article - View full comments

Quoting: liamdaweOkay point well made. I will include an option for people to hide their PC info from their profiles.

I wasn't expecting it to be done that quickly! Thanks ;)

Quoting: ripperI don't think it's much useful to make this distinction and it will only confuse people. Catalyst is now dead, it will phase out from distros very soon. And AMDGPU-PRO is still closed, it doesn't really matter that some parts are open, the bundle is closed overall. You are running a closed blob, and whether it is a kernel driver or a userspace driver is just an unimportant detail. So having just "opensource" and "closed" covers all cases and is simple.

I thought I was the only one thinking that so I didn't bother commenting, but I agree, it's either closed or opened. If you only take the open part of the hybrid driver, the driver just doesn't work, so overall, it's closed.

Quoting: liamdawethis is why it's in BETA so I can make lots of changes

That's the spirit! ;)
Keep up the good work, thanks!

View this comment - View article - View full comments

Quoting: liamdaweYes, profiles can be looked at of course, but I am talking purely in regards to this page right now. The information behind the stats page does not store user_id's, usernames or anything. Only the overall counter for each bit.

I think what @no_information_here is saying is that if you go to my profile page in a private window, you'll see all my information, including the ones in your anonymous stats ;)

From there is it quite easy to scrape the information from all the users using a simple `curl` script. Basically, all our information are publicly available and that is a pretty important concern.

It would be nice if it was possible to set different levels of privacy ("public", "admins only" or, maybe later, "GoL friends" ;)). Making it available to all GoL members is almost as useless as publicly available for someone concerned with his anonymity, since a malicious user can still create an account and scrape all he wants.

(BTW you've got a problem with `https` links for user's website ;))

View this comment - View article - View full comments

Thanks Liam! That's a really cool tool that you've made here! In the long run, you might show a little reminder when a user visits the website and that his data haven't changed for 6 months (with two buttons "Let's update this" and "I still have the same old rig" ;))

Hope more people will fill their data!

View this comment - View article - View full comments

Quoting: CAPTNCAPSBut I am on Linux for just about one year now, so I have no experience with AMD drivers, but telling from the comments here, it sounds like they're not that good... but hey, can't be worse than nouveau, right?
And as my current graphics card gets old, I wanted to switch soon anyway.

Personally, I switched to AMD years ago because they started to do open source drivers. Having the drivers directly in the kernels is a real blessing and just that is enough for me to take an AMD card again next time. I won't lie, performances aren't as good as with an NVIDIA card with proprietary drivers, but they have improved a lot in just one year and now I can even play Dota on my (very) old HD 6870 (in 2560x1440). And with the catching up of OpenGL versions going to an end, I guess mesa devs will focus on more performance and stability (and Wayland!).

View this comment - View article - View full comments

Quoting: AnxiousInfusionEven if we have a healthy marketplace with competition between AMD/Nvidia or between Intel/AMD for CPU it's still only two. A duopoly! And don't point to VIA or Matrox, there are only two companies competing in either CPU and GPU which produce respectable, high performance parts worthy of desktop + gaming. Patents and licensing deals have removed freedom and choice from the x86 ecosystem and we're all worse off because of it.

We'll agree that it is impossible to arrive as a credible competitor, out of the blue. Also, high performance CPU and GPU aren't as profitable markets than in the past decade. So, in that regards, I think ARM is a serious competitor since it owns everything but the desktop.

And I actually won't be surprised to see high end desktop ARM CPU in the years to come.

And to go a little further, I'm not sure graphics cards is a safe horse on which to bet today. I think graphics cards will follow the same fate as the sound cards and the APU will be the norm. I'm fairly sure both AMD and NVIDIA know that very well.

So I do think there is an interesting competition out there. It's just that it's not the one we're expecting. We would like to see a giant competitor arriving and splitting the market, but that rarely happens and even less on a market that is decreasing.

View this comment - View article - View full comments

I switched from NVIDIA a long time ago. If performance isn't your primary criteria (it happens, I'm the proof of it ;)), I strongly encourage to buy AMD. Just for the pleasure of seeing your machine booting up with a pretty good GPU acceleration without having anything to install, or even configure. But maybe it's something that only rings bells to the old Linux veterans like me ;)

As for Intel, I switched to AMD a year ago and I'm fairly happy with it. But here, the main reason was simply the price. Intel's good CPUs are ludicrously expensive, so I looked at AMD's offer and there I find that the quality/price ratio is better.

I'm actually looking for a new AMD GPU that might fit in my case. Since Polaris GPUs seem to dissipate less heat, will the card be a bit smaller too?

View this comment - View article - View full comments

Quoting: NyamiouThe number of Linux users doesn't matter, what matters is how much money they have in their pockets and how much money they are willing to spend. There is a lot of Steam users that only play one game, and often it's a free to play game, those guys don't matter much. The reasons we get so many games right now is that in terms of money spend on new games (especially outside of the blockbusters) we are more close to 5% than 1% and porting to Linux have become less expensive and easier.

Quoting: dmantioneSteamDB has the answers: The game with the most concurrent users, by far, on Steam is Dota 2, which is indeed free to play. So yes, free games can be a factor in ruïning statistics for which platform is most interresting. On the other hand, we have no reason to assume that Linux gamers do not play free games, in fact, we are used to free software.

I agree that the number of Linux users doesn't matter and that the top 10 games has an influence of the real number of Linux users, but the trend is still interesting to see. It's a bit like distrowatch: we know it's not the real numbers, but the protocol is the same since so much time, that you can extract trends from the curves.

View this comment - View article - View full comments

Interesting statistics!

I've got a question about it... I don't know how the peak of concurrent users number was taken, but if it was taken the same day the Steam statistics went out, we might have a problem here.

I'm pretty sure that if we display the curves of the peaks of concurrent users overtime, we would not see a straight line, but a bumpy one, depending on the day in the week. Meaning that if the statistics came out on a Sunday, the 1% of it would be quite different than during some day in the middle of the week.

So if it's not the case, it might be better to do an average of the peaks around the day where the statistics came out.

View this comment - View article - View full comments