Open source has won the battle but is losing the war
steveburge.com
steveburge.com
Web is mostly a distraction for proponents of FOSS. Web developers consume the foundation of FOSS and, without moving past web development, cannot contribute to that foundation. I don't honestly look to domination of web software components as a source of open source growth or an indicator of open source health. Web developers are users and FOSS is powered by developers.
Put more stock in Valve and Crytek's long-term influence on developers using Linux to game on. They subsequently can learn to program Linux, and that's way more influential than 99% of the garbage out there about web being some egalitarian magic kingdom where open source and individuals can thrive, so there. Gaming is one of the last great bastions of totally proprietary software platforms where users with the natural propensity to do so do not eventually become developers.
Honestly I don't see what your first paragraph adds to the discussion besides being an argument of authority of your own opinion being more valid than of anyone younger than you.
About the rest, I can not see how having an end-to-end full open source ecosystem is a worthwhile goal. Open source developers have shown a great propensity towards politics and petty conflicts of personality.
I believe the basic stack of our computing lives should be open source, to have a level playing field where all competitors and collaborators can gain users based on merit. But software applications should be able to be closed source if the users wants them.
Gaming can stay closed source if that creates better games, as a game is almost never a library or an API that can influence other software. And specially: because experience and practicality have shown that closed source games are better.
I want to enjoy the best software possible. In the case of games, I want the best games possible. I don't care if they are closed source. In fact, it's better if they are not, as this will help prevent multiplayer cheating.
I want Valve doing Linux games because that will motivate someone in the future to create some kernel innovation, but I want Valve doing Linux games also because I want good games in my favourite open platform, instead of the crap tux games showcased as better in some way just because they are open source.
I mean, look at the HN frontpage on any given day. A good half of the posts will be about some new open source module or framework that someone has put hundreds of hours into. Closed frameworks are regularly scoffed at and treated with great suspicion. In the Node community, it is a common practice to open-source modules without giving it a second thought, just part of a day's work.
I'm not sure I really understand your point, could you explain?
Imagine a world where every developer learns to program using open source tools and has participated in one open source community prior to college or on-the-job training. Does any individual developer choose to use closed source and become embedded in a bio-dome where their capability to survive is permanently bound to the decisions of the bubble around them? Does any manager of developers expect to be able to direct developers to do this? Does any executive expect to retain even part of their personnel by telling them to willfully extract themselves from the rich ecosystem in which such demands cannot be made of them?
Let's take OpenGL vs Direct X for example. MS says, "Learn Direct X or else your games are doomed to only use 50% of the display's luminosity range!" This is obviously crap, and MS will make tons of moves specifically to implement their software as a yoke upon developers captured by their tool-chain treadmill. You like Active X. You really do, and because you've learned to use it, now you will have a little sign on your website that tells people that IE is required to use your government service. You will specify it in job ads. You will recommend it as a developer, a manager, an executive, a South Korean politician. Your investment will chain you, your users, perhaps your entire country to tribute payments.
Oh, the company can make money. So can the developers, and there is no union necessary because there are no scabs, no hard manufacturing assets, or any other reason other than entrenched, slowly eroding fiefdoms that can convince developers to choose the bio-dome over the rain forest. Corporate, closed-source software development is in a quite precarious arrangement compared to manufacturing and amorphous labor. This is true, even withstanding the still-entrenched gorillas like Apple, MS, Oracle, Google, etc etc etc.
The point of this is that if everyone develops with FOSS first, get ready to see major hemorrhaging of anachronistic bio-domes until their glass is shattered and the ecosystem thrives straight through them. This is bigger than web. Imagine how many wasted resources will be recovered and how productive new companies can be.
This is a quite attractive future depending on who you are. Web can't take us there. Web is not the thing that breaks down all walls. Mozilla can try and try and try. Web is always going to be waiting on standards while other programmers just do, no call for comments required.
The other thing I'm getting at is that web is not a primary indicator of the growth of FOSS because not everyone who uses web has access to the tools that can be used to create web. Web is more Gutenberg Bible than printing press. You need access to the press to become more than just a consumer of the fruit.
In the sense that drawing nascent developers, before they even think of programming, into the FOSS community, forever freeing them from the illusion that rowing the monopolist corporate bireme is a necessary part of life, what Valve and Crytek are doing by improving their strategic relationship with FOSS might as well be the invention of artillery in a land of castles. Linux as a gaming platform fuels a new wave of Linux users with a strong propensity to find joy in programming, the users most likely to become developers. What's going on in gaming on Linux is what leads humanity to a FOSSier world. Valve is one of the rare companies that is in position to push this change and benefit if this change becomes reality.
I have a 3.5 year old phone running Cyanogen Mod 10.1 (Android 4.2.2) and all the sources for it. How is it "faux-open"? It passes the ultimate test: it builds.
It's faux-open because almost every single Android device needs proprietary kernel blobs & firmware to operate almost any non-trivial hardware feature on the device. In practice this pins you to specific kernel versions/APIs, and you're running code you can't change or even audit.
I also have a double down that by Android 6.0, open source Android will effectively be reduced to a boot loader that boots the closed source Google+ Play blob that contains all the system services.
There isn't one Android anymore. There's the base pure open source Android, aka AOSP. There's Android with Google services, which supports things like the Play Store, in-app purchases, location services, etc. There's Amazon's Android fork which has some of their own APIs inserted for in-app purchases and location, but Play Store, Google Chrome, Firefox, etc are unavailable for it. Now there's Nokia's new Android fork which will have its own set of APIs for location, in-app purchases, etc.
You can indeed say that most mobile users are on a proprietary OS and most desktop users are on a proprietary browser.
https://f-droid.org/repository/browse/?fdfilter=HN&fdid=com....
And HN Droid which is "unlicensed."
https://f-droid.org/repository/browse/?fdfilter=HN&fdid=com....
On a less provocative note, I don't understand the argument about web apps. There's tons of FOSS web application frameworks and widget toolkits, and the web is slowly conquering the desktop, for better or for worse. Has it not dominated?
Yes in term of lob/100 page "fairly large" form apps, No in term of native apps that actually do something.
When people show me a Photoshop,Indesign,Final Cut or a 3ds Max (fully featured,not just 3 edit tools hanging around) built in HTML,that work OFFLINE,that can open gigabytes of datas, with great performances on low end hardware, i will agree ( Which my old macbook does with no problem with these native apps).
Let's just take the many text editors built in HTML/JS , try to open a 10/15 mb text file(an apache log file for instance),execute a search/replace with it and see how it performs.
HTML everything is a promise, an idea that can be only fulfilled if the browser engines(not just the javascript engines,the DOM is not javascript) are as fast as native os(and hardware becomes faster too). That's not the case today. So people can say javascript 'is fast' all they want,that's not the bottleneck anymore.
The future of computing is on apps coded against one of two proprietary platforms: Apple's or Google's.
That's an overly narrow view that seems to equate all of "computing" with "smart phones / tablets". I don't think that's a very accurate view. A lot of what happens on computers has nothing to with either.
It's not like Google isn't providing closed source api's too, though.
I hate apple, and I hate that I'm giving them my money, but just the latency issue alone is enough for me to give up on Android.
Help me Mozilla, you are my only hope.
On my Nexus 4 (without carrier apps) I can't remove Chrome, Drive, Currents, Email, Gmail, Keep, Hangouts, Maps, Photos, Play Games, Play Movies and TV, Play Music, Play Newstand, News and Weather ... the list goes on and on. That adds up to a ton of space and clutters my menu. I want some of them, but I don't want some of the others, either way I'm stuck with them, which means I by default always need to get the more expensive phone with more storage because more and more things come with the phone, that's even _before_ a carrier touches it.
And both the UI performance and battery performance woes on Android phones has been well documented. I've had every flagship Nexus/Google phone since the G1 and battery life and performance are not getting better overall.
Every time they make an improvement in Android that improves battery life they just ship a thinner phone with a faster CPU. My G1, Nexus One, Nexus S, Nexus 4, always have lasted about a day. I don't own a Nexus 5 yet but have friends that do and they do the same thing I do - a house/car full of chargers and a battery pack for travel.
Don't get me wrong I still like my Android phone, I just don't love it.
UI performance woes? Coming from an iPhone 3GS, it's flawless to me. As for battery, it fits my bill and it's the price to pay for the huge screen.
My Galaxy S4 is rooted and flashed to semi-vanilla Android with no extra application bundled. I charge my phone every 3 days. Using my preference of keyboard software makes typing much faster and easier and my keyboard is more efficient than keyboard on iOS 7.
Regarding your latency issue, this is probably the hardware fault. It's not a "known" issue with Android.
However what you were saying might make sense if the last time you tried Android device was 4+ years ago, in this case, Android has really improved greatly lately.
Also, I easily get about 24+ hours of battery life on my Nexus 5 with 4-5 hours of screen use. If you look on the Nexus 5 subreddit, some people have difficult getting more than 3-4 hours of screen time while others (like me) have no problem getting 4-5 hours.
The iPhone/iOS just seems to have a better-optimized LTE stack. If you look at Anand's data, the iPhone drops less than 5% from WiFi to LTE, while the Nexus 5 drops 23%. iOS seems to also be much more aggressive about powering down the radio to avoid burning a bunch of battery life in bad signal conditions.
Did you pay a premium price for your low-battery-life, high-latency phone? Or was it at the cheap end of the spectrum?
Mozilla would be more of the same right now, as Javascript is also not native.
My best bet is Ubuntu Phone, whenever it's ready.
Open Source Software is there because somebody is writing it, software isn't going to write itself. And mostly it's alive not (at least "not only") because somebody is financing it, but because there are people that both can and want to write it. And, oh, I should've said "make", not "write", because making it is so much more than just writing it. Especially when we talk about general purpose client-side apps author cares about so much.
And here we have problems with both "can" and "want". Generally, people that are actually able to write software prefer slightly different stuff from your "typical end user". Actually I don't know a single browser that I would still like and it's pretty obvious that your "typical user" won't be so much delighted by my beloved CLI apps.
But what really makes me sad is that "can" problem. There might be enough of those who can do some simple coding in community, but it still takes large company to hire somebody who will cooperate with programmers to make beautiful design for their app or some folks that can do some serious math/programming. FOSS apps still generally suck in terms of UX and visual design (at least for that typical end-user). There's no FOSS that could compete against Photoshop or Cubase or be even close to quality of Google's text-to-speech services.
Of course there's no one to blame except ourselves, but if you want to see the problem — here it is. It's about us improving our skills to be able to build something outstanding, to build it for free and make it open source.
In smaller countries like Scandinavia centralization works VERY well. That is due to their homogeneous nature. In US, we may require something decentralized and more resilient model to have the Free/Open Source software movement. Something where people have incentive to work on good software without wanting to get bought out by the next google or facebook.
I just meant in other aspects of government, everything is very centralized in Scandinavia, and it seems to be work very well there. However, that should not be translated to US, where we have much larger and diverse population.
I hope that make sense?
(If you are playing in the high end of the market and can bring something other than technology to the fight, quality, user experience, ease of use for example then proprietary can keep a foothold).
So the question then becomes can the other layers become commoditized to the point that Open source wins out in those ?
If people want to choose to view the tech world through a filter that projects everything into some open source vs. proprietary software conflict, fine. There are many such filters though.
Someone could have a similar filter that only cares about vertical vs. horizontal integration. It might even be instructive! Microsoft was on top of their game in the '80s and '90s because they had the right business model (horizontal integration- sell the OS and have everyone else make the hardware it runs on and the apps that run on top of it). Other companies have more recently had more success with a more vertically integrated approach, where the hardware and the software are explicitly designed with each other in mind to maximize things like performance, power management, size, shape, and UX of the resultant products.
As things like the mobile industry continue to evolve, we may be moving back into an era where horizontal integration becomes dominant again. And maybe that will lead to another era where it goes back to being dominated by vertical. Who knows. But at no point did vertical or horizontal ever "win" (although plenty of pundits said it did in the '90s), and there is no endgame where one does. It's just a pendulum that swings back and forth. If it ever stops, that means that the entire industry has ground to a halt.
The exact same things can be said about open source vs. proprietary, including the fact that the two are not mutually exclusive. Plenty of things have open source as well as proprietary components, and plenty of companies have some mix of vertical and horizontal integration. I'm not saying these perceived dualities in our industry are comparable, as they describe very different things, only that describing it as a war with an ultimate winner or loser is futile for the exact same reasons.
People should absolutely love open source and be inspired to be proponents of it wherever possible, but I always take great exception to philosophies like RMS', which seem less about emphasizing a particular set of virtues (freedom, openness, sharing, etc.) and much more about demonizing something else. RMS' life goal is not to promote free software, it's to destroy proprietary software. Do what you do because you love something, not because you hate the opposite of that thing. That always leads down very unfruitful paths, including casting everything in binary terms and thinking about things like imaginary battles and wars instead of simply doing great things in the manner that most inspires you and encourages others.
It's simple. I don't know about a "fundamental, axiomatic" proof, but the situation is that proprietary software is obscurantist in its nature and is in direct conflict with information dissemination and the openness that we herald our computing and the Internet as being so important for.
Our society is destined to be dominated by digital technology and computers. All facets of our lives will become digitized and as a result, the ones who control our computing control the way we live. Proprietary software is tantamount to slavery, where the users are helpless and unable to exert control over the technology that governs them.
Don't get me wrong. I love open-source. I've been using and contributing to it for 15 years. But to me, success is having a high-quality product, a vibrant ecosystem surrounding it and an active, engaged community using and contributing to it.
As long as all those things are true, who cares what everybody else does?
Truism. I believe in it.
I do laugh at its hacked together nature, inconsistent naming conventions and multiple functions that do almost the same thing. Still, at the end of the day, I'm able to crank out pages very fast, and the writers are able to easily customize and populate them with content.
I went through a WP-hating phase for a while too, but at this point, I've realized that it is THE tool for your standard, somewhat-dynamic, large to medium sized website.
Yes, Ruby on Rails has been widely and frequently publicized. Yes, there are some people who have a semi-religious devotion to Ruby on Rails, and are quite vocal about it. Yes, a number of books have been written about it. Yes, training is available for it.
But in reality, we really don't see it used all that much, at least compared to PHP, Java, and ASP.NET. The numbers given in that graphic may not be exact, but I wouldn't be so quick to rule them out, either.