Gabe Newell LinuxCon Keynote. Announcement Next Week.

Subject: Editorial, General Tech, Systems, Mobile, Shows and Expos | September 16, 2013 - 09:15 PM |
Tagged: Steam Box, LinuxCon, Gabe Newell

Valve Software, as demonstrated a couple of days ago, still believe in Linux as the future of gaming platforms. Gabe Newell discussed this situation at LinuxCon, this morning, which was streamed live over the internet (and I transcribed after the teaser break at the bottom of the article). Someone decided to rip the stream, not the best quality but good enough, and put it on Youtube. I found it and embed it below. Enjoy!

Gabe Newell highlights, from the seventh minute straight through to the end, why proprietary platforms look successful and how they (sooner-or-later) fail by their own design. Simply put, you can control what is on it. Software you do not like, or even their updates, can be stuck in certification or even excluded from the platform entirely. You can limit malicious software, at least to some extent, or even competing products.

Ultimately, however, you limit yourself by not feeding in to the competition of the crowd.

If you wanted to get your cartridge made you bought it, you know, FOB in Tokyo. If you had a competitive product, miraculously, your ROMs didn't show up until, you know, 3 months after the platform holder's product had entered market and stuff like that. And that was really where the dominant models for what was happening in gaming ((came from)).

But, not too surprisingly, open systems were advancing faster than the proprietary systems had. There used to be these completely de novo graphics solutions for gaming consoles and they've all been replaced by PC-derived hardware. The openness of the PC as a hardware standard meant that the rate of innovation was way faster. So even though, you would think, that the console guys would have a huge incentive to invest in it, they were unable to be competitive.

Microsoft attempts to exert control over their platform with modern Windows which is met by a year-over-year regression in PC sales; at the same time, PC gaming is the industry hotbed of innovation and it is booming as a result. In a time of declining sales in PC hardware, Steam saw a 76% growth (unclear but it sounds like revenue) from last year.

Valve really believes the industry will shift toward a model with little divide between creator and consumer. The community has been "an order of magnitude" more productive than the actual staff of Team Fortress 2.

Does Valve want to compete with that?

This will only happen with open platforms. Even the consoles, with systems sold under parts and labor costs to exert control, have learned to embrace the indie developer. The next gen consoles market indie developers, prior to launch, seemingly more than the industry behemoths and that includes their own titles. They open their platforms a little bit but it might still not be enough to hold off the slow and steady advance of PC gaming be it through Windows, Linux, or even web standards.

Speaking of which, Linux and web standards are oft criticized because they are fragmented. Gabe Newell, intentionally or unintentionally, claimed proprietary platforms are more fragmented. Open platforms have multiple bodies push and pull the blob but it all tends to flow in the same direction. Proprietary platforms have lean bodies with control over where they can go, just many of them. You have a dominant and a few competing platforms for each sector: phones and tablets, consoles, desktops, and so forth.

He noted each has a web browser and, because the web is an open standard, is the most unified experience across devices of multiple sectors. Open fragmentation is small compared to the gaps between proprietary silos across sectors. ((As a side note: Windows RT is also designed to be one platform for all platforms but, as we have been saying for a while, you would prefer an open alternative to all RT all the time... and, according to the second and third paragraphs of this editorial, it will probably suffer from all of the same problems inherent to proprietary platforms anyway.))

Everybody just sort of automatically assumes that the internet is going to work regardless of wherever they are. There may be pluses or minuses of their specific environment but nobody says, "Oh I'm in an airplane now, I'm going to use a completely different method of accessing data across a network". We think that should be more broadly true as well. That you don't think of touch input or game controllers or living rooms as being things which require a completely different way for users to interact or acquire assets or developers to program or deliver to those targets.

Obviously if that is the direction you are going in, Linux is the most obvious basis for that and none of the proprietary, closed platforms are going to be able to provide that form of grand unification between mobile, living room, and desktop.

Next week we're going to be rolling out more information about how we get there and what are the hardware opportunities that we see for bringing Linux into the living room and potentially pointing further down the road to how we can get it even more unified in mobile.

Well, we will certainly be looking forward to next week.

Personally, for almost two years I found it weird how Google, Valve, and Apple (if the longstanding rumors were true) were each pushing for wearable computing, Steam Box/Apple TV/Google TV, and content distribution at the same time. I would not be surprised, in the slightest, for Valve to add media functionality to Steam and Big Picture and secure a spot in the iTunes and Play Store market.

As for how wearables fit in? I could never quite figure that out but it always felt suspicious.

Read on for our transcript of the keynote speech. Bare with us, it is a little bit rough.

Transcript, Gabe Newell at LinuxCon 2013, 9/16/2013

Gabe Newell: Hi, I'm Gabe. It feels a little bit funny coming here and telling you guys that Linux and Open Source are the future of gaming. It's sort of like going to Rome and teaching Catholicism to the Pope. Bare with me.

Valve was founded in 1996. We have a mix of single player and multiplayer games like Half Life, CounterStrike, and so on. We shipped our game on a bunch of different system. The PC, most obviously, as well as Mac, Xbox, Playstation, and other consoles. We've also created a set of tools for software developers and for customers called Steam.

Now a lot of our decisions and our thinking about the future are base on what we see as structural changes that are a consequence of the ongoing improvements in performance and price performance in computing and networking. I think that people always over-estimate the effect of these things in the short run and under-estimate them in the long wrong. One of the key decisions we made when we were first starting the company was the realization that as we continue to reap of the benefits of those kind of performance improvements that the relative value of line of business functions would change. That sales and marketing and distribution would be cheaper and easier and that meant the design and development of products were going to become more valuable.

And a lot of times you're going to see corporate structures that struggle to keep up with that pace of change. But it's not just in how corporations should be run that those changes end up being propagated, it has implications throughout my industry, or throughout the gaming industry. You'll see it in the emergence of digital distribution platforms like our own Steam. But you'll also see it in ways that are kind of surprising. For example, we reached the point a couple of years ago where the marginal costs of having another player were less than the marginal community benefits of having someone be in that audience or participating in that community. And that gave rise to what it first sounded a little perverse, the free to play style of games.

We also see other phenomenon occur. You'll see things like Twitch.tv come along or eSports where, you know, in order to have a model that encompasses the value people get out of it you have to think about the novel ways that content creators and experiencers are going to be generated. And, you know, at the last tournament we held we had over a million people watching it simultaneously. That is between a Swedish Team and a Russian-Ukrainian team.

Of course, right now, Valve is fairly successful so it would be nice if the world just stopped right here because this is a point in time where we're doing well, but the process is going to continue to go on. In the same way as we've seen the center of gravity of gaming moving from console hardware platforms into developers and services, we're going to see the process continue and it is the users who are going to be the focus of optimal strategies

((connection lost))

the conclusion that we reached is that games are going to essentially be nodes in a connected economy where the vast majority of digital goods and services are going to be user created rather than created by companies.

So let's take a step back and talk about Valve's own experiences with Linux. In 1999 we shipped our first game server for Linux and since then it has pretty much been the majority of game servers that has been out there. That means, around the world, there is actually probably over a million game servers out there, right now, running ((Linux)).

We have not only use Linux in terms of deploying a service footprint, we use it internally. All of our source code, all of our models, all of our animations, all of the assets are all being run on top of Linux. We love the performance and reliability and robustness of it.

One thing that tends to, sort of, shock people especially in the media worlds is just how big games have become. When we do a new update for something like DOTA2, near as we can tell we're generating about 2% of the worldwide mobile and land-based IP traffic. That tends to startle people who don't realize what a large sea change has been going on.

Now our experiences, we think, are fairly typical to people in the games industry. So if you go to Blizzard, or Bungie, or any of the other game companies you're going to find that in terms of building out a service provision or in terms of running their own operatings they're all super comfortable and have a lot of experience running Linux.

On the other hand, the user experience of being a gamer trying to play on top of Linux has been pretty painful. Just, you know, any metric that you use if you're just looking at how many players there are, how much time they spend playing, how much revenue they're generating for game developers, it's very small. Typically under 1%. That's way too small to ever get any attention in terms of making sure you're compatible with device drivers or distros or thinking about how to improve that experience.

However, what we become convinced of is that Linux really is the future of gaming. What I'm going to do for the remainder of this presentation is talk about why we think that Linux is critical and what Valve is going to do to help make that happen.

Several years ago we got very concerned about directions that the PC industry was going. We thought... how to be polite about this... so we thought that there was some bad thinking and that there were these new platforms that were starting to emerge and they had this nice characteristic that you could control access to those platforms. If you didn't like competing with Google you just didn't let them ship on your device. Or, you could determine how often they update. You could have a lot of control over things like pricing and other characteristics and that was a very seductive opportunity which I think led to some poor decision making by some of the key actors in the PC space. I think that the way they should have responded is very much to key off of the strength of the openness of the systems we had and sort of double-down on it rather than going the other way.

So, jumping forward, what are we looking at now? We're looking at steady year-over-year unit declines in PC sales. And, the people in the field sort of deer in the headlights. They were like, "We didn't have a model where this was occurring, we thought that people would keep buying more and more PCs regardless of what we did and what sort of restrictions we imposed on them. You see all these articles, PCs in decline, I think we'll see either significant or restructurings or market exists by top 5 PC players. It's looking pretty grim.

On the other hand, PC gaming seems to have been immune to this downturn. In the past, you know back to the Nintendo Entertainment System, gaming was really led by these proprietary hardware and software standards that also... if you wanted to get your cartridge made you bought it, you know, FOB in Tokyo. If you had a competitive product, miraculously, your ROMs didn't show up until, you know, 3 months after the platform holder's product had entered market and stuff like that. And that was really where the dominant models for what was happening in gaming ((came from)).

But, not too surprisingly, open systems were advancing faster than the proprietary systems had. There used to be these completely de novo graphics solutions for gaming consoles and they've all been replaced by PC-derived hardware. The openness of the PC as a hardware standard meant that the rate of innovation was way faster. So even though, you would think, that the console guys would have a huge incentive to invest in it, they were unable to be competitive.

There also used to be, almost nobody knows this, there used to be proprietary networks for gaming and the internet was like, "Who needs that?". Obviously the internet has completely replace any of those homegrown solutions for multiplayer gaming.

PC Gaming is where innovation is occuring. And you'll see that in lots of ways. It's not on the consoles. It's not on any of the closed systems where any of the innovation is happening. It's happening to the extent at which openness is embraced by the underlying platform. So digital distribution came out, like Steam, social gaming occurred on the PC. MMOs. Trading. Workshops. All of the most interesting topics in gaming are coming out of the open environment of the PC. So, in spite of seeing year-over-year units declines in PC sales on the gaming side we're seeing huge increases. The data I'm most familiar with is ours, but we're going up 76% year-over-year at the same time as PC sales are getting double-digit declines. And I think that's true for all of our partners on the PC space to a varying degree. That the innovation and the openness of the PC as a gaming platform have enabled us to be somewhat immunized against some of the broader structural decline of the PC.

We also think that the rate of change is increasing, that we're not really going to be slowing down. So that systems which are innovation friendly, which is equivalent to openness, are going to have greater and greater competitive advantages to closed or tightly regulated systems. We actually think that we're going to see a significant sort-of democratization in the content creation process. The users will be, you know, the distinction between a content creator and a content consumer will get blurrier and blurrier.

Just to be really concrete about that. Valve, we're kind-of a cocky company. We like to think that we can compete with any company in the world, right? If we sat down and said let's have a competition with Bungie, or Epic, or Blizzard, we're friends with all of these guys btw, we'd say, "Oh we can take them, we're good at what we do". But the one we wouldn't want to compete with is our own users. They have already outstripped us, spectacularly, you can't compete with them once you give them the tools that allow them to participate in the creation of the experiences that they find are valuable. And it's not by a little bit, it's an order of magnitude more productive, already, and we're only a couple of years into thinking about how to do that. You don't want to compete with your customers because they're going to be way better. You also don't want to compete with Reddit or 4Chan, but the point is the connected groups of users are going to be way more successful if they're properly enabled and supported than any of the individual game developers are going to be.

But there's this huge tension between if that's the direction that gaming is going or if that's the direction content creation is going these other systems actually put a tremendous number of roadblocks in the way of doing that. It takes Valve several months to get through certification process for a single update. It took us 6 months to get one update through the Apple Store to ship an iPad update. We have a lot of resources and have a huge commercial motivation and no individual user is going to, if they're the center of gravity for content production, is going to have the wherewithal or the stubbornness to get through that. That's just one example of the many ways that the closed systems appear to us to be antithetical to our user-centric model of content production going forward.

So in general we've seem tremendous evolution and innovation in the open space. We've seen that we think that that process is going to continue. And we think that we should be knocking down as many barriers and reducing friction, not going in the other direction.

So, that's great, what have we been doing? Several years ago we thought, "Wow, this is where we think games are going". There's a logical consequence to that which is, "We need to be working pretty hard to make sure Linux is a good solution for gamers and for game developers. It was kind-of distressing because there is just a lot of work we need to do in order to help address those issues. It wasn't just that there was the work, you have to figure out how to stage it out. If you try to do everything all at once, you'll do a bunch of things pretty badly. So, not only did we see ourselves needing to do a lot of work but we also had to be thinking about what was the process of how to roll that out, with partners and with customers.

So the first step was to get a game running on Linux and we used that, like developers do, as kind-of a sweater thread of issues. So we're going through and it's like, "Oh my God it's incredible slow" and then we talk with NVIDIA and work with them to fix their drivers and then we get it to the point that it's a lot faster than the Windows version. And, it's like, "Okay that's great". And then you find the user experience is impossible, like how do you do updates? Well it's a super long and it's like "well compile yourself". And that's not a good experience for your user. So going through the exercise of getting one game up you have to do a lot of different things. Then, the value of that, you're sort of optimistic going into it that if you solve a set of problems for one game you're going to solve it for a bunch of others and that definitely appears to have been true. If you get the kind of game that we produce all, you know, create a good user experience around that we've definitely solved problems for the Call of Duty team or Unreal Tournament, or whatever. Games aren't that much different the key is to get it all the way through to customers.

Then, this February, this year we shipped a Steam client for Linux. And, as much as anything, that was a signal to our development partners that we really were serious about this Linux thing that we were talking with them about. Now, today, we have 198 (at least when I got on the plane we got 198 games) that were running on Linux.

Being a... helping make Linux successful also involves a bunch of other things as you all know. We have engineers who work on SDL which is an open source project started by another Valve employee, Sam Latinga, and we're a contributing member of the Khronos Group.

One thing I don't think we talk about is we're helping developing a Linux debugging, that's in addition to the work we've been doing with the LLVM people. So when we go out and talk to developers and say, "Okay, if you can pick one thing for Valve to work on on the tools side to help make Linux a better development target what would it be?" and they were always coming back to us and telling us that we should build a debugger. So we're working with another company on building a Linux debugger.

These are useful things to do to help move things forward on the developer's side.

Now I'm going to take a step back, first, to talk about the living room. You know there's this notion that... there's this kind-of struggle that I think you guys go through on the cloud side that you recognize sort of the implications for how you can abstract certain problems away from both developers and users where it doesn't really matter where you are you just know that it's going to work. Those are exactly the kinds of problems that we feel it is important to attack on the gaming side. Right now you're sort of in this bizarre situation where as soon as you sit on your couch you're supposed to have lost connection with all of your other computing platforms. It's like, "Oh, just buy all your games all over again". The input methods incompatible and, yes you can have music, but you need to buy it from us rather than fsomebody else.

So we thought that that was the incorrect way and really through design through thinking hard about how to create appropriate abstractions for both users and developers you could build something which spanned through the desktop and the 10-foot living room experience. It's one of those things where people are like, "Oh, people don't want PCs in their living room" and basically there's no amount of arguing you can do with that person to convince them otherwise or they say, "Yes, this is trivial"

((connection loss))

... the basis for the future of gaming. We need to build something which showed that yes, in fact, you could take everything that you like about your PC and get it to work in your living room, that's called Big Picture.

Our next step, having done these other pieces, is now on the hardware side. We think, especially, there's sets of issues to making sure that whatever computing platform you have works well in a living room environment. There's thermal issues, and sound issues, but also a lot of input issues. Our next step in this is to release some work we've done on the hardware side.

You know, even more broadly in terms of the grand unification, we don't really think that the fragmentation around the physical location or the input devices of computation is either necessary or desirable for software developers or for consumers. Everybody just sort of automatically assumes that the internet is going to work regardless of wherever they are. There may be pluses or minuses of their specific environment but nobody says, "Oh I'm in an airplane now, I'm going to use a completely different method of accessing data across a network". We think that should be more broadly true as well. That you don't think of touch input or game controllers or living rooms as being things which require a completely different way for users to interact or acquire assets or developers to program or deliver to those targets.

Obviously if that is the direction you are going in, Linux is the most obvious basis for that and none of the proprietary, closed platforms are going to be able to provide that form of grand unification between mobile, living room, and desktop.

Next week we're going to be rolling out more information about how we get there and what are the hardware opportunities that we see for bringing Linux into the living room and potentially pointing further down the road to how we can get it even more unified in mobile.

So thank you very much, I hope this has been interesting.

Source: LinuxCon
September 16, 2013 | 10:50 PM - Posted by derz

I have to give kudos to Valve for their strong support to an open platform that is not paying off at the moment, but is still viewed as the future of gaming.
The games I have played on Linux at present work great without bugs. I could not say that a few years ago, so progress is being made.

September 17, 2013 | 09:51 AM - Posted by YTech

When I was first introduced to Linux platforms, I felt that graphics were better supported than WinOS and iOS. Also it had some really neat features that I couldn't access on other platforms. During that time, I configured a Dual-Boot just for that.

Now, seeing that Valve is going that path, I might become more inclined to return.

However, the challenge is many software/hardware, that I use most often, are only supported in WinOS.

As for extra revenues for Valve, a scenario could be; that extra PC for the children will have Linux and they can play all their Valve games on it, and keep WinOS or iOS for business usages.

**And who usually have more time to develop? --Children**

September 17, 2013 | 03:16 PM - Posted by pdjblum

He has done so much for the pc gaming community. Despite his achievements, unlike most successful ceo's in the tech world, he is still humble and customer driven. Maybe his efforts in the linux space will be what brings linux for pc's into the mainstream. I use win 7 on one rig and centos on the other. I like them both.

September 17, 2013 | 09:35 PM - Posted by Anonymous (not verified)

anddddddddddddd this is going no where.

September 18, 2013 | 02:08 PM - Posted by Anonymous (not verified)

This is great for Valve titles but I think it will be harder to get others to develop just for Linux on top of the Windows based PC platform which already is established and huge.

I applaud Gabe for trying, but not sure how this is going to go.

My problem is I have soooo many Windows based games on my PC from steam already and having to boot back and forth Linux/windows to get to the game I want sounds kind of sucky.

Post new comment

The content of this field is kept private and will not be shown publicly.
  • Lines and paragraphs break automatically.
  • Allowed HTML tags: <a> <em> <strong> <cite> <code> <ul> <ol> <li> <dl> <dt> <dd> <blockquote><p><br>
  • Web page addresses and e-mail addresses turn into links automatically.

More information about formatting options

By submitting this form, you accept the Mollom privacy policy.