Quoting: soulsourceA Mozilla developer posted about Snap (and Flatpak) just recently, and I think it's a very interesting read.

Especially since it was Mozilla that asked Canonical to replace the repository version of Firefox with the snap.

View this comment - View article - View full comments

Quoting: Purple Library GuyWell, sure, but the existing Linux bit of that 132 million is a titch over 1%, so like 1.5 million, max. A couple of million units would more than double that. So I would have expected if sales were high, the variations in the sampling would not normally be enough to make Linux numbers actually fall.

No. The 132 million is the number of users that were on Steam in a particular month, not the total number of users. The total number of users is higher, but how much higher only Valve knows, and to estimate how much higher you'd need a model of which proportion of users would be online in a particular month, which we also don't have.

We can estimate (assuming that all Deck users are active, because they've just bought a new shiny, and that one hardware sample = one user, which isn't realistic but getting better than that approximation would be super hard) that the total active Steam Linux user base is five times larger than the number of Decks sold, since SteamOS is only 20% of the distros sampled by the survey, and the total active Steam user base is 80 times bigger than that. You get a similar Steam Deck is 1/400th of the Steam user base estimate if you look at the GPU stats: the Deck's GPU is 0.27% of the GPUs surveyed.

View this comment - View article - View full comments

Quoting: Purple Library GuyYeah . . . but I was really expecting, given the apparent success of the Steam Deck, that it would kind of swamp all that variability.

The Steam user base is huge, with 132 million monthly active users. There are 20 million people on Steam right now. A couple of million units is just a drop in the ocean. If the Steam Deck sold as many as the best selling console ever (PS2, 155 million) to brand new, never on Steam before, customers, the Steam user base would be less than twice as big.

View this comment - View article - View full comments

Quoting: Eike-0.11% last month. Is a million Steam Decks or two not enough to compensate for Steam's growth?!?

Linux Steam is growing marginally faster than Steam as a whole (check Liam's trend line), but the data is just a sample rather than a continuous measurement of the whole so the data is noisy. In particular, samples with a higher proportion of respondents in China (where Linux usage is essentially non-existent) than respondents in Europe or the USA (where Linux usage is in the low single digits) will show the Linux share dropping, even if Linux is growing overall.

View this comment - View article - View full comments

Quoting: JarmerInteresting update for sure! I remember adding this to my wishlist years ago - honestly had forgotten all about it. I wonder how it compares to Space Haven.

I've got both of them on my wishlist - I don't buy Early Access games. I guess I'll see which one crosses the finish line first.

View this comment - View article - View full comments

Quoting: stormtuxinstalling and playing a game in steam from Linux will raise the sales statistics the developers/publishers have access to? Are we sure "activated games" are handled the same way as "bought games"?

Yes. The platform that gets the most playtime during the refund period (two weeks) is the platform that counts for the sales statistics, unless there isn't any which means that it's the platform it was purchased on, unless that isn't available information (because it's Android or just a key) in which case it's left undefined.

will Valve get a percent of the cost of the keys in the bundle? You know, bandwidth, infrastructure etc.

No. Keys are given to developers for free, and Valve just eats the infrastructure costs. The only restriction (other than total number of keys) is that "you plan to give a comparable offer to Steam customers within a reasonable amount of time."

View this comment - View article - View full comments

Quoting: MohandevirStill, I wonder what design choice they would have made in using a 1280x800 8" no bezel screen... Meaning, keep the exact display size surface and just remove the actual bezels. Could the Steam Deck get smaller? Would it cause airflow issues? Does these no-bezel screens generate more heat than the one they chose?

If they'd known for sure what the demand would be, so they could order in sufficient quantity to still hit their price point with a better screen, I think they'd have kept everything else the same without the bezels. Same resolution, since the APU has solid performance at 1280×800 but would be too weak for too much of the catalogue at a higher resolution, and the same size, since they had to work hard at the cooling and fitting everything in. Just with physically more of the front of the device showing an image.

View this comment - View article - View full comments

Quoting: MohandevirMy point is more pixels means more strain on the APU. Unless the Steam Deck 2 comes with a much more powerful APU, which may mean shorter battery autonomy, that's a tradeoff I would avoid.

For sure, it's definitely a balance that need to be struck. APU and battery tech improvements might cover it by the time the next gen is ready, and if they don't just stick with the lower res.

As for the bezel thing it might be caused by the internal components that require too much space?

Nah, it's because they were trying to hit a price point, the same as it not having great colour reproduction. The bezel is the (quite chunky in this case) part of the screen that doesn't display anything. Having pixels that go all, or almost all, the way to the edge is an engineering challenge, so displays where they've managed to do so are more expensive.

View this comment - View article - View full comments

Quoting: MohandevirI don't see the point in the resolution race.

Game devs are more likely to test their UI at 1080p than 720p (unless they're targeting the Deck specifically, in which case the actual resolution doesn't matter, since they're testing for whatever it is). The bezels are too big so, if you get rid of those, you either get a smaller screen, bigger pixels, or more pixels. Of those (assuming there's adequate performance), more pixels is best. That's why I suggested it, anyway; OEMs just like putting big numbers on spec sheets.

View this comment - View article - View full comments

Ah. They've today put up a request for feedback [External Link], so I think I'll go through the survey with my little one and highlight that deficiency in a PC game. Maybe they will spare the dev resources to fix it after all?

View this comment - View article - View full comments