Quoting: liamdawePersonally I don't see it as an attack, and it frustrates me when people see criticism as an attack.

I agree, but Linux users are pretty hyper-sensitive to criticism of the platform too.

View this comment - View article - View full comments

From what I've heard the music makes up for the graphics. ^_^

View this comment - View article - View full comments

Quoting: WorMzyIt's still frustrating to see that 16.8% of intel GPU users are still confused about their drivers. I wonder if these select few are optimus users.

If you are downloading the drivers from the official Intel site (01.org) it's not hard to assume they are proprietary drivers. To put it another way, it's as proprietary as the Java JRE is.

And plenty of people do this too, especially those with older distros.

View this comment - View article - View full comments

It's not wierd at all to never see a survey one Linux. More people refuse the survey on Windows (and Wine, especially Wine). This means they need to survey more people to get the number of responses they need for statistical significance (which means you see more surveys on Windows and Wine than on Linux native). Refusing the survey does nothing to change the result, they will just ask someone else on that platform.

This is standard practice with opt-out surveys. It's much easier to manage with a system like Steam as the client identifies itself by OS on connection. (Side note: if you have ever been telephone surveyed they often start with demographic questions for this reason.)

View this comment - View article - View full comments

Also, if you take a look at the numbers in more detail you can see that Linux isn't growing in the same areas that Windows is. Easiest way to pull something from the survey data is to look at languages. People generally don't change the language steam is set to, and language correlates somewhat with country.

Linux (two fastest growing):
French 2.14% +0.16%
Spanish 1.88% +0.22%

Windows (two fastest growing):
Thai 2.53% +0.99%
French 3.60% +0.18%

Mac (two fastest growing):
Spanish 2.91% +0.17%
Dutch 0.83% +0.13%

I've no idea what is happening in Thailand, but that's a change equal to 3 million new or returning users there.

View this comment - View article - View full comments

Quoting: linuxgamerThey should introduce a survey day eg. the 15th of every month to make the survey representative. It also wouldn't be a problem at all to just send 1 additional byte to identify the OS every time users log into Steam. Maybe this is against Valve's user data policy.

Your first point: they do this already. The surveys go out at the same time.

Your second point: they don't need to. They already know how many people are loading steam what each OS. This information is going to be accurate because they have a 100% check of it (client identify at login). The point of the survey is to maybe find out what specific OS you are running (Ubuntu 15.04 64-bit, etc), but more likely what hardware you have on-board. You already know you see platform specific sales and notices, of course they know.

0.88% (+/- error) is the real number, accept it. By the way if you convert the percentages to the numbers of users the Linux user base is growing (roughly 30k, a month). The total steam user base is growing two orders of magnitude faster though (3-4million per month).

View this comment - View article - View full comments

Quoting: oldrocker99YES. I know LT calls GNU/Linux "Linux," but RMS, the creator of GNU, which is the OS we're actually using, asks (and reasonably politely) that we call it GNU/Linux, inasmuch as he correctly points out that in a running distro, about 5% of the whole system is the kernel. I have test-driven this rant on Ubuntu Formus, and got pretty promptly shot down, but I still feel that RMS' contributions are worthy of recognition.

And what's wrong with the Tux logo? Note my avatar...

RE: the GNU thing - Do you jump on every troll? Actually, it wasn't even a troll. It was a flippant joke. It didn't need a response, clearly I was aware people feel strongly about it. I was suggesting that the people upset with the removal of the Tux icon were the same people upset over the lack of a GNU prefix. Why should GNU or even Linux get specific acknowledgement over and above other pieces of enabling software or technologies used? People playing games on SteamOS don't need to be reminded about the GNU project anymore or less than they need to know about Autodesk Gameware, SteamWorks, Chromium or SDL.

Also, as I said the Tux logo doesn't work in two-colours. Even if it did, it's got too much fine detail to work at less than 32 pixel height. It's a mess at that size, in fact it's probably fairly unrecognisable in two-colour 20 pixel high form [External Link]. It's a designer's nightmare.

View this comment - View article - View full comments

Valve had it wrong from the start, they should always have placed a GNU icon before the Linux icon.

(But seriously, the Tux icon looks rubbish. It doesn't really work in two colour. The head-alone logo looks even worse. SteamOS looks gooooood.)

View this comment - View article - View full comments

Quoting: HonorEDnlKThe Unity increase could be to the greater number of responses submitted.

But in my case, after 3 years of using Unity I switched to KDE Plasma 5 as my main desktop environment. I consider it to be the most efficient and good looking desktop I've used so far, and having been using it since the beta I've noticed how the bugs are already going away with each update. It also seems they plan on switching from X to Wayland this year.

Likely, yeah. From memory of the various posts along the same lines on the subreddit most people use Ubuntu (and consequently Unity). GoL has a slightly more experimental crowd it seems.

View this comment - View article - View full comments

Quoting: CybaCowboyBut those statistics are only based off those that receive the survey request and choose to complete the survey, right?

Not a very accurate way of documenting the actual hardware statistics, considering you're going to have a number of users refuse to participate and no way of ensuring a balanced collection of data... Valve Software should factor-in the number of Steam clients running on a given platform, either in the background or in-game.

There's no reason why Valve Software couldn't have such a data-collection method as an "opt-in" feature and it would surely have to be more accurate than the way they currently do things...

There is nothing to say that they don't correct for response (ie. opt-out) rates. They almost certainly do. In fact the evidence is pretty clear, users are more likely to see a survey in Windows (or Wine) than Linux native. That's down to differing response rates on those platforms. It's not that you don't get survey requests on Linux it's that the quotas required for statistically significant data is reach much earlier when response rates are higher.

The Steam client identifies by platform (obviously) and distro, so throwing up a survey request on startup to those platforms without a full quota is pretty straight-forward. Hell, even browsing to store.steampowered.com shows platform specific titles almost exclusively.

View this comment - View article - View full comments