Then there was the Blackberry Playbook. It ran an abortion of an OS. QNX based, but not really BB10. It also didn't support native email... on a Blackberry device.
This remark keeps getting repeated again, and again, and again. While a valid complaint, IMHO, its *not* the reason the PlayBook failed. As a developer who was actually paying attention when the PlayBook was being launched, the *real* problem was the lack of a decent SDK. At launch, the only public SDK was a roll-your-own-UI hodge-podge with Adobe Air. Nearly a year later, they added an NDK that offered little more in terms of UI than an OpenGL canvas. At this point, the PlayBook could do a decent job at running games. But it couldn't do much else app-wise.
They never had a decent first-party app SDK/framework until BB10, and it was never back-ported to the PlayBook.
In retrospect, that might have actually worked. Except for that bit where I don't think Microsoft wants to build platforms that don't run Windows, and I don't think another platform vendor would want to use Windows as a starting point.
When Microsoft tried re-inventing their mobile platform (WP7 through WM10), no one really cared and the platform was allowed to compete on its own merits. Of course because no one cared, it had to be propped up by Microsoft's large bank account and a plethora of dirt-cheap devices. This worked for a time, and temporarily allowed it to grow to a small-but-reasonable number of users.
When BlackBerry tried re-inventing their mobile platform (BB10), the hate was so overwhelmingly strong that the platform was never allowed to compete on its own merits. They constantly had crap thrown in their face, and most people even refused to believe that they had something new. Even though I'd argue they made a better platform, they didn't have Microsoft's large bank account or plethora of dirt-cheap devices. So once they fizzled out, it was with a far smaller user base.
Everyone loves to make suggestions like this, conveniently forgetting that the incident occurred on the left-hand side in a spot without enough room to pull over. So what exactly do you all suggest? An automatic "pull hard right across 4 lanes of traffic in event of confusion" feature?
I suppose a massive multi-car/multi-directional T-boning mess is better than one car crashing into a barrier, right?
US-101 is a major highway. You do not have bike lanes on a major highway.
On the left side of the road, along the stretch in question, there is nowhere to pull over. For much of it, the concrete barrier comes right up to the lane. In a few spots, there's a gap of a few feet.
Airplanes are actually a good example. Most of what we currently know as commercial airplanes were pretty much solidified by the late 1960's. Most changes since then feel very much like incremental optimizations.
I used to live in a part of Orlando, FL, where most of the highways were toll roads. If you lived there, you bought the transponder and forgot about it. Basically, it turned into an automatic "cost of living in this area" tax. (They even had full-speed lanes where locals could skip having to stop/slow at every damn toll booth just to pay up.)
But if you didn't live there, or lived on the other side of town (where you didn't need those highways 75% of the time), then it was a constant burden and/or annoyance. (I'll also bet that the cost of these tolls really added up for the poor, encouraging them to take much longer to get anywhere by getting lost on side-streets.)
In theory, charging for actual road use isn't a terrible idea. In practice, unless you can make the actual "toll collection" painless for people who aren't local residents of that region, it kinda is a bad idea.
Given that he uses the phrase "Arduino-class microcontroller" as the token alternative, I feel like he has no real grasp of what a microcontroller really is. "Arduino" is just a hobby-grade software stack, typically (but not exclusively) run on top of certain Atmel microcontrollers (along with certain reference designs) to make them more accessible.
A few years ago, it seemed like every week there was another SBC whose whole claim to fame was that it ran Linux, and they had a Slashdot post. This kinda culminated in the Raspberry Pi. I felt like many of these products exist to help introduce pure software people (like many here) to the world of embedded programming.
Meanwhile, Arduino is kinda doing the same thing from the other end. It runs on a more typical microcontroller, but it provides a simplified development environment to help introduce hardware people (okay, or anyone afraid of big scary platform SDKs in C or assembler) to embedded development.
And once you add in the thickness of all these "add-on components", you might as well have just made the damn phone thicker and more robust... and doubled the battery capacity in the process.
At this point, I don't think Apple is the only one guilty of this. However, they are the trend-setter everyone else is following.
Unfortunately, in the modern smartphone world, this is far easier said than done. Many products depend on back-end infrastructure integration that cannot be so easily opened up (e.g. push services, API tokens, etc.). Also, many products depend on expected behaviors of other apps using the product, and open implementations can easily violate these expectations (or be way behind in supporting any protocol changes).
Unfortunately, the only way anything new gets built is by tearing down older stuff. There really isn't any unused space for brand new residential housing development.
(Of course we could build denser/upwards, but I think everyone is aware of the push-back problem that always gets out here.)
And the sad part is a lot of the area actually does not even look upscale or expensive. If you look at a photo of a neighborhood with a $2M house in it for the Mountain View / Sunnyvale area, and stripped off the address, you'd probably assume it was a $200k (or less) place in a boring part of the country.
You mean the distro I ditched years ago, because every day around 4pm it would randomly decide to completely hog my system resources? Or the distro I ditched because its package manager took so long to initialize that it was faster to download and compile the friggin source code to something? Or the distro that used to be a big KDE holdout, until they decided to also throw in the towel and join the Gnome camp?
"Universality" was *the* reason I originally forced myself to learn how to use vi. If I tried to use any random UNIX machine of any era, it was the only editor I could consistently depend on as always being there. (assuming I could run a full-screen editor, of course)
Emacs? Vim? Nano? And we're talking about desktop distributions? Hands down, none of those compare to the quality of Sublime Text as a text editor.
Sublime Text is a commercial product, which will always put it at a huge disadvantage in overall popularity in the Linux world.
There are a lot of people who primarily use Linux because it (and everything on it) is free, with all other factors secondary. (Though they'll tend to deny it if you say it to their face.)
That being said, the two most used applications (that I paid real money for) on my Linux desktop are Sublime Text and Beyond Compare. I'm also one of those rare folks who would gladly pay real money for a Linux distro if it got me some extra polish and the benefits of being able to license bits (rather than exclude them in the name of F/OSS purism).
I only wish that KDE wasn't universally shunned by most of the major Linux distro players, and thus relegated to second-class status.
That being said, a big part of why I use Fedora over Ubuntu on my home desktop, is that Fedora does a much better job with their KDE build. (even if the top-down attitudes aren't that different)
The editor wars kinda annoy me, because it really isn't somewhere that we need absolutes. For me, the "best editor" really depends on what it is that I'm editing. The best editor for configuration files may not be the best editor for source code. The best editor for a bunch of related files may not be the best editor for a single one-off file or scratch editing of a text snippet. Likewise, the best editor in a GUI might not be the same as the best editor in a terminal window.
Except he's caused all sorts of "officials" to take "notice" and "offense" to something that no one really cared about before. I wonder if this stunt will have the side-effect of suddenly making business much more difficult for everyone who already makes and/or owns similar devices that were previously fine.
the result is that people need to find a quiet place from time to time to not be disturb so that they can concentrate on a specific task.
This reminds me of a time when I needed to take a private phone call, which you obviously cannot do at your desk in an open-plan office. So I walked around the entire building, and every single conference room was occupied by "one person sitting in front of a laptop." I think I eventually found some corner by a stairwell.
The way Gnome 3+ is obsessed with removing features for fear of scaring off some "mythical idiot user", it might as well be TWM with a prettier theme. Well actually, TWM probably lets you minimize an app to an icon, so it has more functionality.
Seriously, if you need a bunch of config hacks and tweak tools to take your environment from "stock" to "basic usability," you should rethink your design priorities.
NOTE: I fully understand that Tesla's 'autopilot' feature isn't a full-on self-driving car.
I really wish more people would understand this. Pretty much everyone I talk to, who only casually observes Tesla headlines, assumes that it is full-on self-driving. I'm sick and tired of having to constantly explain to them that it isn't.
Let's Encrypt has short-lived certificates, which are kinda useless and annoying when you have a device that is *not* a general-purpose computer capable of running their scripts.
Am I really going to do a manual process on every cable modem, WAP, router, printer, switch, AP, IoT device, etc, every 3 months?
The "local network devices" problem is a real problem, and its never given proper attention in these HTTPS proclamations.
I "solved" it for myself by setting up a local CA to make certs for my stuff. Unfortunately, getting the cert for that CA into all my browsers is annoying, and can introduce its own share of issues.
That's the new TiVo, it's gone downhill after acquisition. But the original TiVo was a great product, it invented the entire concept of DVRs, and all the competitors had lousy products with bad UIs in comparison.
Unfortunately, due to the way modern digital cable works, these lousy products simply worked with their respective cable TV systems "out of the box." Meanwhile, TiVo users had to struggle through frustrating CableCARD nightmares while being constantly encouraged to "just give up and get the Scientific Atlanta POS-box like all our other customers."
I feel like this has probably hurt TiVo, over time, more than almost anything else. Its also probably why they actually have no serious competitors in the "DVR not provided by the cable company" market.
Considering how its "easy" to get an LCD TV (or whatever Samsung's "quantum dot" actually us) in this size class, its thus far been impossible to find OLED beyond 77" (/w insane price).
As such, I'm very much looking forward to seeing a display like this go into mass production. (Heck, I'd be equally happy if they did it in only 4k.)
Right now, I'm still using an old Samsung 61" back-projection DLP set (a late model, so it actually looks good) and have wanted to upgrade to something bigger for a while (moved and its now in a bigger room). I've been waiting for OLED to scale up and mature, so this is definitely a good sign.
Slashdot Mirror
Then there was the Blackberry Playbook. It ran an abortion of an OS. QNX based, but not really BB10. It also didn't support native email... on a Blackberry device.
This remark keeps getting repeated again, and again, and again. While a valid complaint, IMHO, its *not* the reason the PlayBook failed. As a developer who was actually paying attention when the PlayBook was being launched, the *real* problem was the lack of a decent SDK. At launch, the only public SDK was a roll-your-own-UI hodge-podge with Adobe Air. Nearly a year later, they added an NDK that offered little more in terms of UI than an OpenGL canvas. At this point, the PlayBook could do a decent job at running games. But it couldn't do much else app-wise.
They never had a decent first-party app SDK/framework until BB10, and it was never back-ported to the PlayBook.
In retrospect, that might have actually worked. Except for that bit where I don't think Microsoft wants to build platforms that don't run Windows, and I don't think another platform vendor would want to use Windows as a starting point.
When Microsoft tried re-inventing their mobile platform (WP7 through WM10), no one really cared and the platform was allowed to compete on its own merits. Of course because no one cared, it had to be propped up by Microsoft's large bank account and a plethora of dirt-cheap devices. This worked for a time, and temporarily allowed it to grow to a small-but-reasonable number of users.
When BlackBerry tried re-inventing their mobile platform (BB10), the hate was so overwhelmingly strong that the platform was never allowed to compete on its own merits. They constantly had crap thrown in their face, and most people even refused to believe that they had something new. Even though I'd argue they made a better platform, they didn't have Microsoft's large bank account or plethora of dirt-cheap devices. So once they fizzled out, it was with a far smaller user base.
Everyone loves to make suggestions like this, conveniently forgetting that the incident occurred on the left-hand side in a spot without enough room to pull over. So what exactly do you all suggest? An automatic "pull hard right across 4 lanes of traffic in event of confusion" feature?
I suppose a massive multi-car/multi-directional T-boning mess is better than one car crashing into a barrier, right?
US-101 is a major highway. You do not have bike lanes on a major highway.
On the left side of the road, along the stretch in question, there is nowhere to pull over. For much of it, the concrete barrier comes right up to the lane. In a few spots, there's a gap of a few feet.
I think you've got it wrong...
"You can get this great discount on your health insurance if you get this implant!"
(Meanwhile, we're doubling the base rate.)
Airplanes are actually a good example. Most of what we currently know as commercial airplanes were pretty much solidified by the late 1960's. Most changes since then feel very much like incremental optimizations.
I used to live in a part of Orlando, FL, where most of the highways were toll roads.
If you lived there, you bought the transponder and forgot about it. Basically, it turned into an automatic "cost of living in this area" tax. (They even had full-speed lanes where locals could skip having to stop/slow at every damn toll booth just to pay up.)
But if you didn't live there, or lived on the other side of town (where you didn't need those highways 75% of the time), then it was a constant burden and/or annoyance. (I'll also bet that the cost of these tolls really added up for the poor, encouraging them to take much longer to get anywhere by getting lost on side-streets.)
In theory, charging for actual road use isn't a terrible idea.
In practice, unless you can make the actual "toll collection" painless for people who aren't local residents of that region, it kinda is a bad idea.
Given that he uses the phrase "Arduino-class microcontroller" as the token alternative, I feel like he has no real grasp of what a microcontroller really is. "Arduino" is just a hobby-grade software stack, typically (but not exclusively) run on top of certain Atmel microcontrollers (along with certain reference designs) to make them more accessible.
A few years ago, it seemed like every week there was another SBC whose whole claim to fame was that it ran Linux, and they had a Slashdot post. This kinda culminated in the Raspberry Pi. I felt like many of these products exist to help introduce pure software people (like many here) to the world of embedded programming.
Meanwhile, Arduino is kinda doing the same thing from the other end. It runs on a more typical microcontroller, but it provides a simplified development environment to help introduce hardware people (okay, or anyone afraid of big scary platform SDKs in C or assembler) to embedded development.
And once you add in the thickness of all these "add-on components", you might as well have just made the damn phone thicker and more robust... and doubled the battery capacity in the process.
At this point, I don't think Apple is the only one guilty of this. However, they are the trend-setter everyone else is following.
Helmets!
https://www.youtube.com/watch?v=6bV7pM_HS70
Unfortunately, in the modern smartphone world, this is far easier said than done. Many products depend on back-end infrastructure integration that cannot be so easily opened up (e.g. push services, API tokens, etc.). Also, many products depend on expected behaviors of other apps using the product, and open implementations can easily violate these expectations (or be way behind in supporting any protocol changes).
Unfortunately, the only way anything new gets built is by tearing down older stuff. There really isn't any unused space for brand new residential housing development.
(Of course we could build denser/upwards, but I think everyone is aware of the push-back problem that always gets out here.)
And the sad part is a lot of the area actually does not even look upscale or expensive. If you look at a photo of a neighborhood with a $2M house in it for the Mountain View / Sunnyvale area, and stripped off the address, you'd probably assume it was a $200k (or less) place in a boring part of the country.
You mean the distro I ditched years ago, because every day around 4pm it would randomly decide to completely hog my system resources? Or the distro I ditched because its package manager took so long to initialize that it was faster to download and compile the friggin source code to something? Or the distro that used to be a big KDE holdout, until they decided to also throw in the towel and join the Gnome camp?
"Universality" was *the* reason I originally forced myself to learn how to use vi.
If I tried to use any random UNIX machine of any era, it was the only editor I could consistently depend on as always being there. (assuming I could run a full-screen editor, of course)
Emacs? Vim? Nano? And we're talking about desktop distributions? Hands down, none of those compare to the quality of Sublime Text as a text editor.
Sublime Text is a commercial product, which will always put it at a huge disadvantage in overall popularity in the Linux world.
There are a lot of people who primarily use Linux because it (and everything on it) is free, with all other factors secondary. (Though they'll tend to deny it if you say it to their face.)
That being said, the two most used applications (that I paid real money for) on my Linux desktop are Sublime Text and Beyond Compare.
I'm also one of those rare folks who would gladly pay real money for a Linux distro if it got me some extra polish and the benefits of being able to license bits (rather than exclude them in the name of F/OSS purism).
I only wish that KDE wasn't universally shunned by most of the major Linux distro players, and thus relegated to second-class status.
That being said, a big part of why I use Fedora over Ubuntu on my home desktop, is that Fedora does a much better job with their KDE build. (even if the top-down attitudes aren't that different)
The editor wars kinda annoy me, because it really isn't somewhere that we need absolutes.
For me, the "best editor" really depends on what it is that I'm editing. The best editor for configuration files may not be the best editor for source code. The best editor for a bunch of related files may not be the best editor for a single one-off file or scratch editing of a text snippet. Likewise, the best editor in a GUI might not be the same as the best editor in a terminal window.
Except he's caused all sorts of "officials" to take "notice" and "offense" to something that no one really cared about before. I wonder if this stunt will have the side-effect of suddenly making business much more difficult for everyone who already makes and/or owns similar devices that were previously fine.
the result is that people need to find a quiet place from time to time to not be disturb so that they can concentrate on a specific task.
This reminds me of a time when I needed to take a private phone call, which you obviously cannot do at your desk in an open-plan office. So I walked around the entire building, and every single conference room was occupied by "one person sitting in front of a laptop." I think I eventually found some corner by a stairwell.
The way Gnome 3+ is obsessed with removing features for fear of scaring off some "mythical idiot user", it might as well be TWM with a prettier theme. Well actually, TWM probably lets you minimize an app to an icon, so it has more functionality.
Seriously, if you need a bunch of config hacks and tweak tools to take your environment from "stock" to "basic usability," you should rethink your design priorities.
NOTE: I fully understand that Tesla's 'autopilot' feature isn't a full-on self-driving car.
I really wish more people would understand this. Pretty much everyone I talk to, who only casually observes Tesla headlines, assumes that it is full-on self-driving. I'm sick and tired of having to constantly explain to them that it isn't.
Let's Encrypt has short-lived certificates, which are kinda useless and annoying when you have a device that is *not* a general-purpose computer capable of running their scripts.
Am I really going to do a manual process on every cable modem, WAP, router, printer, switch, AP, IoT device, etc, every 3 months?
The "local network devices" problem is a real problem, and its never given proper attention in these HTTPS proclamations.
I "solved" it for myself by setting up a local CA to make certs for my stuff. Unfortunately, getting the cert for that CA into all my browsers is annoying, and can introduce its own share of issues.
That's the new TiVo, it's gone downhill after acquisition. But the original TiVo was a great product, it invented the entire concept of DVRs, and all the competitors had lousy products with bad UIs in comparison.
Unfortunately, due to the way modern digital cable works, these lousy products simply worked with their respective cable TV systems "out of the box." Meanwhile, TiVo users had to struggle through frustrating CableCARD nightmares while being constantly encouraged to "just give up and get the Scientific Atlanta POS-box like all our other customers."
I feel like this has probably hurt TiVo, over time, more than almost anything else. Its also probably why they actually have no serious competitors in the "DVR not provided by the cable company" market.
Considering how its "easy" to get an LCD TV (or whatever Samsung's "quantum dot" actually us) in this size class, its thus far been impossible to find OLED beyond 77" (/w insane price).
As such, I'm very much looking forward to seeing a display like this go into mass production. (Heck, I'd be equally happy if they did it in only 4k.)
Right now, I'm still using an old Samsung 61" back-projection DLP set (a late model, so it actually looks good) and have wanted to upgrade to something bigger for a while (moved and its now in a bigger room). I've been waiting for OLED to scale up and mature, so this is definitely a good sign.