Yeah, but we’ve been on this stressful ride before, and we know where it ends.
There were lots of attempts at a closed source proprietary Internet protocol. They have all resoundly failed, after looking unbeatable. Some folks still fondly remember the closed Internet protocols like OLE COM, ActiveX, Flash, Cold Fusion, and SilverLight, but few of us miss them. Okay, I do miss Flash games.
Good touchscreen phone operating systems were a “will this ever be matched?” trade secret at Blackberry and Apple. Now the vast majority of phones run open source Android.
Much earlier, most good-enough C compilers were expensive proprietary closed source products. Now I see very little being compiled on anything other than the free and open source GCC. Even most other programming languages and tools are now FOSS, as well. I can’t think of much for development that cracks the top 20 that isn’t FOSS. JetBrains IDEs stand out as a lone closed source hold-out.
Open standards always win, in the end.
The desktop computing default is honestly way overdue to switch to FOSS. That’s why it’s the year of the Linux desktop.
The Fediverse is here to stay, and is all that’ll be left in a couple decades. But in the meantime, it’s cozy!
Yep. It often takes quite awhile. And I honestly don’t mind supporting innovators who want to sell something closed but really good.
But as I get older, and watch the pattern over and over, I’m starting to appreciate skipping the cycle by directly adopting the open thing as early as I can.
yeah, the general rule of thumb seems to be that if it’s universal, it needs to be open. The farther niche it goes, the less open it has to be, on principle of utility. Open standards are only good people it’s so easy for them to get accepted. That’s why closed standards often just don’t go very far.
it’s a rather weird concept, but it makes sense. If you want to standardize, let’s say, threaded hardware across the continental US that you would inherently need to do away with any closed standards, assuming you want it to actually work, and along with that, whatever you settle on, needs to be open.
You could theoretically do this with closed source, but the problem here is that there will be someone that comes along and does it with open source, and if it’s better, you’re fucked. And if it’s equal, and cheaper, you’re fucked. And if it’s marginally worse, but trivial to adopt, you’re fucked.
They have all resoundly failed, after looking unbeatable
“Failed” how? Failed in that they sucked to use, yeah. Failed as in people stopped using them? No. Failed as in their profits plummeted? No. Their users are gluttons for abuse and exploitation. And their users are nearly everyone I know. They don’t care.
Now the vast majority of phones run open source Android.
No they don’t. It’s only ~50% of the US market and ~60% of the global market. Also at the time it was a very competitive market and OEMs like Nokia and Motorola all had their day. That hasn’t been true since the original iPhone some 15 years ago.
Open standards always win, in the end.
LOL no they don’t? In fact they almost always don’t.
The desktop computing default is honestly way overdue to switch to FOSS.
It will never be, for the same reason it has a whopping 4% today. It’s complicated and difficult to use, and no one is spending money on marketing to convince people to even try it.
The Fediverse…is all that’ll be left in a couple decades.
That’s nothing but a wild utopian delusion…
Honestly, I wish it were true. I wish you were right. I wish people respected themselves to take the time and learn about all the corporations that are fucking up their lives, and to take the time and learn how to take the according action, but at best they don’t care, and more often then not they will actually go out of their way to defend the corporations that are fucking them. If they cared, the corporations wouldn’t be able to fuck them, and we would live in a completely different world.
Failed" how?
Failed as in people stopped using them? No.
Failed as in their profits plummeted? No.
What the actual fuck?
I gave several concrete examples whose usage was originally seen as unassailable, and is now easily measured as essentially zero.
Of the examples I listed, only Shockwave still has any publicly recorded examples of actual continued use, because there’s a virtual museum dedicated to preserving it’s memory.
That’s a fine definition of a failed technology.
You’re out of your element, Donny.
Edit: Your other points are essentially that those technologies aren’t at their dominant phase yet. I can agree about that.
If you still need convincing (your clearly do) about open standards, read the history of licensed screwdrivers. Closed standards either die off, or become open ones. There are no exceptions.
Windows and iOS are both notable because, in my expert opinion, both have already missed their window of time when they could have become successful open standards.
Their respective owners actually realize that, as well. IBM Mainframe also missed that window, and there’s history available to read. We are now seeing the same business patterns (as IBM Mainframe) with Windows and iOS.
Incidentally, IBM Mainframe actually doesn’t qualify for my failed technology list, because it’s still holding on. Windows probably has similar staying power to IBM Mainframe (hanging on in zombie death for decades). iOS isn’t lucky enough to live on huge expensive machines that are hard to move, though. It’s not going to be as lucky.
OLE COM, ActiveX, Flash, Cold Fusion, and SilverLight
I don’t know what any of these things are but I’m pretty sure they’re not popular social media platforms. If you don’t understand why that matters then you have a fundamental lack of understanding of the situation as it stands.
Sorry. Movie quote. The Big Labowski. Check it out. It’s fun. For context, the guy that says that line is a blowhard, not to be taken too seriously. (Like me!)
I don’t know what any of these things are but I’m pretty sure they’re not popular social media platforms. If you don’t understand why that matters then you have a fundamental lack of understanding of the situation as it stands.
I understand network effects. All of my examples had large network effects supporting them, in their time.
Seriously. Open standards win. It takes flipping forever sometimes. But they do. Check into the screwdriver thing. It’s a cool read. Or for something more recent, the histories of open and closed web browsers. I think you’ll find it encouraging.
Open standards win. It takes flipping forever sometimes. But they do.
Did XMPP win? Did RSS win? Did Linux win? These open standards have been around for decades and are still not widely adopted. At what point are they considered a failure?
Is ActivityPub winning? At best it has stalled after gaining a few defectors like myself from those who are unwilling to tolerate bullshit. But at this point I think it’s abundantly clear that there is no amount of abuse that the majority of users won’t tolerate on proprietary platforms.
When open standards win, it’s usually because the platform was built on them, like email or podcasts.
Again, I’m happy to concede that XMPP looks doomed today, like RSS did a couple decades ago.
Did RSS win?
RSS certainly hasn’t won, yet. But RSS is doing fine, behind the scenes. Most of the RSS the average person interacts with doesn’t look, to them, like RSS. There’s a lot of RSS still in wide use, today. Competing solutions are currently enshitifying (Google Search, Reddit, Facebook, Xitter), while RSS is still free and still just works.
That’s not an automatic win for RSS, until you consider that RSS has already outlived WebCrawler, Digg, MySpace and GeoCities, among others.
I’m calling it early in favor of RSS.
We’ve agreed that I am prone to do so, though.
Did Linux win?
Yes. Linux won. The vast majority of computation today runs on Linux.
Windows used to hold a serious percentage of web hosting. My best guess is it was around half. The current percentage is unknown, but generous estimates put it at 3%, at most. For some context, the Azure cloud (Microsoft’s web hosting that Office 365 runs on) is known to mostly run on Linux.
But to address the other part of your question:
Is Windows desktop going away?
Something mostly proprietary that costs money and is called Windows with be with us for a long time.
But the Windows kernel is counting it’s final days now, while most people haven’t noticed.
The Windows kernel is cool, but it’s a pure cost center and no longer offers anything that Linux doesn’t.
Game developers noticed, this year. I personally, held onto Windows desktop for decades, solely for gaming. I suspect the shift this year will turn out to be a key moment in the spin down of the Windows kernel.
A desktop OS has a ton of moving pieces. We’re currently seeing the natural trend for those pieces to take advantage of existing open solutions.
I predict that we will see more and more of that, until the switching cost reaches the current low cost of switching web browsers.
Except it doesn’t “just work” because most platforms have abandoned it.
The vast majority of computation today runs on Linux.
Arguably not where it matters.
the Windows kernel is counting it’s final days
it’s a pure cost center and no longer offers anything that Linux doesn’t.
You’re delusional. Show me evidence that “the Windows kernel” is going away. Linux is still a giant pain in the ass to use unless you have lots of experience with command line.
Game developers noticed, this year.
Some of them? Sure. Big ones? Can’t even be bothered to tick a box in the Steam client to make them available. Actively adding features that break Linux compatibility and openly acknowledging that, at best, they don’t care.
until the switching cost reaches the current low cost of switching web browsers.
My dude have you not seen marketshare for Chrome and Safari?
Yeah, but we’ve been on this stressful ride before, and we know where it ends.
There were lots of attempts at a closed source proprietary Internet protocol. They have all resoundly failed, after looking unbeatable. Some folks still fondly remember the closed Internet protocols like OLE COM, ActiveX, Flash, Cold Fusion, and SilverLight, but few of us miss them. Okay, I do miss Flash games.
Good touchscreen phone operating systems were a “will this ever be matched?” trade secret at Blackberry and Apple. Now the vast majority of phones run open source Android.
Much earlier, most good-enough C compilers were expensive proprietary closed source products. Now I see very little being compiled on anything other than the free and open source GCC. Even most other programming languages and tools are now FOSS, as well. I can’t think of much for development that cracks the top 20 that isn’t FOSS. JetBrains IDEs stand out as a lone closed source hold-out.
Open standards always win, in the end.
The desktop computing default is honestly way overdue to switch to FOSS. That’s why it’s the year of the Linux desktop.
The Fediverse is here to stay, and is all that’ll be left in a couple decades. But in the meantime, it’s cozy!
to be fair, this was almost certainly a reaction to the iphone. Still open, so there’s that.
Seems like the cycle is either:
Yep. It often takes quite awhile. And I honestly don’t mind supporting innovators who want to sell something closed but really good.
But as I get older, and watch the pattern over and over, I’m starting to appreciate skipping the cycle by directly adopting the open thing as early as I can.
yeah, the general rule of thumb seems to be that if it’s universal, it needs to be open. The farther niche it goes, the less open it has to be, on principle of utility. Open standards are only good people it’s so easy for them to get accepted. That’s why closed standards often just don’t go very far.
That’s a great point! I kind of skipped over, that. Good add, thanks.
it’s a rather weird concept, but it makes sense. If you want to standardize, let’s say, threaded hardware across the continental US that you would inherently need to do away with any closed standards, assuming you want it to actually work, and along with that, whatever you settle on, needs to be open.
You could theoretically do this with closed source, but the problem here is that there will be someone that comes along and does it with open source, and if it’s better, you’re fucked. And if it’s equal, and cheaper, you’re fucked. And if it’s marginally worse, but trivial to adopt, you’re fucked.
“Failed” how? Failed in that they sucked to use, yeah. Failed as in people stopped using them? No. Failed as in their profits plummeted? No. Their users are gluttons for abuse and exploitation. And their users are nearly everyone I know. They don’t care.
No they don’t. It’s only ~50% of the US market and ~60% of the global market. Also at the time it was a very competitive market and OEMs like Nokia and Motorola all had their day. That hasn’t been true since the original iPhone some 15 years ago.
LOL no they don’t? In fact they almost always don’t.
It will never be, for the same reason it has a whopping 4% today. It’s complicated and difficult to use, and no one is spending money on marketing to convince people to even try it.
That’s nothing but a wild utopian delusion…
Honestly, I wish it were true. I wish you were right. I wish people respected themselves to take the time and learn about all the corporations that are fucking up their lives, and to take the time and learn how to take the according action, but at best they don’t care, and more often then not they will actually go out of their way to defend the corporations that are fucking them. If they cared, the corporations wouldn’t be able to fuck them, and we would live in a completely different world.
What the actual fuck?
I gave several concrete examples whose usage was originally seen as unassailable, and is now easily measured as essentially zero.
Of the examples I listed, only Shockwave still has any publicly recorded examples of actual continued use, because there’s a virtual museum dedicated to preserving it’s memory.
That’s a fine definition of a failed technology.
You’re out of your element, Donny.
Edit: Your other points are essentially that those technologies aren’t at their dominant phase yet. I can agree about that.
If you still need convincing (your clearly do) about open standards, read the history of licensed screwdrivers. Closed standards either die off, or become open ones. There are no exceptions.
Windows and iOS are both notable because, in my expert opinion, both have already missed their window of time when they could have become successful open standards.
Their respective owners actually realize that, as well. IBM Mainframe also missed that window, and there’s history available to read. We are now seeing the same business patterns (as IBM Mainframe) with Windows and iOS.
Incidentally, IBM Mainframe actually doesn’t qualify for my failed technology list, because it’s still holding on. Windows probably has similar staying power to IBM Mainframe (hanging on in zombie death for decades). iOS isn’t lucky enough to live on huge expensive machines that are hard to move, though. It’s not going to be as lucky.
My guy, you are completely overreacting.
I don’t know what any of these things are but I’m pretty sure they’re not popular social media platforms. If you don’t understand why that matters then you have a fundamental lack of understanding of the situation as it stands.
…who? are you talking to?
Sorry. Movie quote. The Big Labowski. Check it out. It’s fun. For context, the guy that says that line is a blowhard, not to be taken too seriously. (Like me!)
I understand network effects. All of my examples had large network effects supporting them, in their time.
Seriously. Open standards win. It takes flipping forever sometimes. But they do. Check into the screwdriver thing. It’s a cool read. Or for something more recent, the histories of open and closed web browsers. I think you’ll find it encouraging.
Did XMPP win? Did RSS win? Did Linux win? These open standards have been around for decades and are still not widely adopted. At what point are they considered a failure?
Is ActivityPub winning? At best it has stalled after gaining a few defectors like myself from those who are unwilling to tolerate bullshit. But at this point I think it’s abundantly clear that there is no amount of abuse that the majority of users won’t tolerate on proprietary platforms.
When open standards win, it’s usually because the platform was built on them, like email or podcasts.
That’s the perception I’m trying to counter with the web technology examples I gave above.
I was there building the web, on proprietary products, and I believed that, myself.
I’m delighted to report that I was wrong.
It took decades, but the far less visible corner of the web running on open technologies is now the only portion we currently still have.
With a big delightful exception for Shockwave Flash, and the folks valiantly keeping it alive to preserve it’s part in gaming history.
That remains to be seen. I’ll gladly accept XMPP as a point in the “against” column, as it has a long way to go, if it succeeds.
Google succeeded handily at their last round of embrace, extend, extinguish, against XMPP, by dropping support from Google Chat.
It’s worth noting that the question isn’t really whether XMPP replaces WhatsApp, it’s whether it can unseat SMS.
SMS is seriously entrenched. I don’t know it’s state of openess. My understanding is it’s mostly run/owned by a few large proprietary players.
https://en.m.wikipedia.org/wiki/SMS
Again, I’m happy to concede that XMPP looks doomed today, like RSS did a couple decades ago.
RSS certainly hasn’t won, yet. But RSS is doing fine, behind the scenes. Most of the RSS the average person interacts with doesn’t look, to them, like RSS. There’s a lot of RSS still in wide use, today. Competing solutions are currently enshitifying (Google Search, Reddit, Facebook, Xitter), while RSS is still free and still just works.
That’s not an automatic win for RSS, until you consider that RSS has already outlived WebCrawler, Digg, MySpace and GeoCities, among others.
I’m calling it early in favor of RSS.
We’ve agreed that I am prone to do so, though.
Yes. Linux won. The vast majority of computation today runs on Linux.
Windows used to hold a serious percentage of web hosting. My best guess is it was around half. The current percentage is unknown, but generous estimates put it at 3%, at most. For some context, the Azure cloud (Microsoft’s web hosting that Office 365 runs on) is known to mostly run on Linux.
But to address the other part of your question:
Something mostly proprietary that costs money and is called Windows with be with us for a long time.
But the Windows kernel is counting it’s final days now, while most people haven’t noticed.
The Windows kernel is cool, but it’s a pure cost center and no longer offers anything that Linux doesn’t.
Game developers noticed, this year. I personally, held onto Windows desktop for decades, solely for gaming. I suspect the shift this year will turn out to be a key moment in the spin down of the Windows kernel.
A desktop OS has a ton of moving pieces. We’re currently seeing the natural trend for those pieces to take advantage of existing open solutions.
I predict that we will see more and more of that, until the switching cost reaches the current low cost of switching web browsers.
Except it doesn’t “just work” because most platforms have abandoned it.
Arguably not where it matters.
You’re delusional. Show me evidence that “the Windows kernel” is going away. Linux is still a giant pain in the ass to use unless you have lots of experience with command line.
Some of them? Sure. Big ones? Can’t even be bothered to tick a box in the Steam client to make them available. Actively adding features that break Linux compatibility and openly acknowledging that, at best, they don’t care.
My dude have you not seen marketshare for Chrome and Safari?
We have established that. It doesn’t mean I’m wrong.
That’s what my other examples are evidence for. I’m extrapolating a trend I’ve seen many times before. I could easily be wrong.
I’m not stressing over it though, because I’m happy being a delusional old person.
Yes. And it supports my point. Here’s the source code to Chrome: https://github.com/chromium/chromium
Here’s the license: https://github.com/chromium/chromium/blob/main/LICENSE
And here’s the source code to Safari: https://github.com/WebKit/WebKit
There’s an advertising campaign by the current big players that everyone may as well accept their bullshit, because everyone else does.
I’m telling you, from experience, that putting up with that (current) bullshit is temporary.
They’ll innovative new bullshit, of course. That’s how the pattern goes.