Steam Linux Usage Slipped In July According To Valve's Current Data

[Guest]

Here's an idea for Valve regarding the survey:
- At first launch, ask the user if they want to opt in.
- Afterwards, on every launch, send survey data.

This may be much better than the current "ask-every-who-knows-how-long-months" approach.

For me, it's ask only when I launch Windows once a several weeks. And, never ask, when I'm running Linux.

[kneekoo]

I migrated to Linux, full-time, 10 years ago. I used the Steam client in quite a few years now and I was only asked once to participate in the survey. I don't know if the question pops-up randomly or they have some weird algorithm to choose the users they ask each month, but considering the purely-technical information they require for these stats, I'd rather stay permanently opted-in in their stats collection.

The only drawback I see in them allowing everyone to opt-in would be a huge increase of data points for their stats, but I think they can easily cope with the needed money for the extra hardware required to smoothly process all the data.

[d3coder]

This may be because of much faster growth of windows

[Termy]

In windoze-dual-boot times, i was only ever asked for the survey on windoze....but that regularly. On Linux, back then and the last 5 years of Linux-only i only got the survey one single time on Linux....
So, i'd be all in for a opt-in solution.
The mentioned numbers for <1TB and DualCores seem to support the weak significance of the data...

[s_j_newbury]

People are adopting the NVIDIA GeForce GTS 450 released 9 years ago faster than Linux (+0.19%)

Isn't that the model DXVK fakes?

[birdie]

Alternatively we can just admit it's next to zero and move one.

You cannot expect any discernible uplift in the market unless we have a desktop OS which we can call Linux. Right now we have dozens of incompatible distros which are not even compatible with themselves (e.g an application compiled in Ubuntu 19.04 will not run in Ubuntu 16.10).

[Xaero_Vincent]

I wonder what the value proposition is for Valve to continue investing in Linux gaming. I mean it's awesome and I'm a full-time Linux user (with the exception of a Windows VM), gaming on Linux has improved leaps and bounds but I'm not seeing a return on their investment with a sub 1% user base.

[Tuxee]

I was never asked to fill survey on Linux. But, I was asked 3 times to fill survey on Windows. And, I have used Windows only to try Windows-only games, and it was only 0.5% of the whole time I had spent playing on Steam. Of course, statistics show lower Linux usage, if survey opens only (mostly) on Windows platform.

It doesn't. I get it on my Ubuntu system every now and then. And - since there is no Windows system - never elsewhere. I even got it in rather short (about 2 months) succession recently on two of my machines. I still think that the collected data is way too spotty.

[birdie]

So on steam Intel CPU usage slighlty raises, meanwhile at the same period elsewhere AMD sells CPUs like crazy:

7nm Ryzen mops the floor with Intel: 75% of rev and 80% of sold CPUs @ mindfactory.de July 2019

Truth must be somewhere in between

What AMD CPUs market share has to do with Linux market share? Also, >95% of AMD fans have never heard of Linux and the ones who have tried it, quickly forgot about it.

[birdie]

In windoze-dual-boot times, i was only ever asked for the survey on windoze....but that regularly. On Linux, back then and the last 5 years of Linux-only i only got the survey one single time on Linux....
So, i'd be all in for a opt-in solution.
The mentioned numbers for <1TB and DualCores seem to support the weak significance of the data...

Ah, that old never-dying BS excuse. As if there aren't any Windows users who haven't been invited to Steam HW Survey.