The primary concern is toward the interests of the end-user.
Kind of, the primary concern is what the end user can do with the source code. This is the reason free software isn't particularly interesting to people, because end users are not programmers and thus are almost never interested in the source code. Free as in gratis is great for end users but freedom with respect to the source code isn't particularly useful to them.
It's a great model for advancing corporate interests, they can employ or contract developers to get their needs met and can even keep those changes inhouse if they wish. But there's no real model for advancing end user interests, that's just hobbyists and volunteers or some overlap with corporate needs. That's why free software really is lacking, sure end users could contract developers to implement a feature for them or spend time learning how to do it themselves but it's generally more cost effective to just pay the fee for a proprietary program that does it, cheaper for the end user and income for the developer.
10 might be a stepping stone where they offer both, but I have no doubt in mind SaaS windows will arrive with windows 10.
Not until having cellular data access in laptops becomes commonplace. Even now while they do have SaaS offerings they haven't replaced any software with a SaaS-only offering much less doing so with the operating system doing that would be fraught with problems and drive people straight to alternatives. What you suggest makes absolutely no sense at all.
It's been how long now and they still can't make a dent on the desktop. Same with macosx.
It's all about the programs it can run, no end user cares about the operating system, they care about what programs they can run on it. An operating system alone is not useful. This is why OSX does have at least some share of the desktop market, because there are quite a few popular professional-grade programs that run on it like Logic, Final Cut, Adobe Creative Suite, etc. so there are a group of professionals that can get their work done on a Mac. The choice of operating systems comes after the choice of programs for the task you want to accomplish. For example if you need Photoshop your options are Windows or OSX but not Linux or if you need Solidworks your only option is Windows, not OSX or Linux.
Linux as a desktop operating system in general has been perfectly fine for many years, but it doesn't have the application support that end users need which makes it pretty useless for most people.
So if systemd is so bad why not just fork one of the distros pre-systemd? Sure it's a big job to maintain a distro but apparently there's also a big group of people who don't like systemd too, if that's anything close to the majority then it will easily succeed.
A company tries to get their product to be more popular. Sounds like a good strategy. If it works, bully for them. If it doesn't, they'll try something else. Either people will use it or they won't.
I could see it potentially working on Android because it could be properly integrated but on iOS there's no way to make it a good user experience. I'm not a big fan of the voice command stuff on my phone on the rare occasion I do use it I just use Siri rather than unlocking my phone, finding the Google app and using their voice search.
Well it fits the definition of a Linux definition perfectly, I'm not particularly concerned if you want to say your iPod Touch runs a version of NextStep or care whether or not that's at all accurate.
Stop being an idiot, you know damn well the ability to avoid a problem based on a test is predicated on acting on the result of that test and not the test itself. You've been proven wrong and now you're trying weasel words to get out of it.
I can take such measures as well in C as in C++.
Demonstrate how do you do that both at runtime *and* compile time then.
Who's moving the goalposts again?
I already told you that in my last post, it's you. But you're still wrong on both counts, if you use the language features you can get type safety at the compiler and/or runtime levels.
Why did you ask that question if you were going to go on a big rant about something you profess to not understand rather than waiting for an answer anyway?
To clarify: Other high-priced watches have some differentiating factor, they aren't just a cheap watch with the same case mass-produced from a different material. This is putting a Timex in the same shaped case but mass-producing it out of a more expensive material.
...also please don't take personal offence to the comparison of Apple and Timex, I know some Apple fans (not saying you're one of them) would do that and then pretend they don't understand the point on that basis.
The high-end Apple Watch is $10k (which is definitely not that high in the watch-as-a-fashion-statement-world) and actually has a solid 18k Gold-Ceramic "alloy" housing.
It's definitely high when the company it comes from is the maker of common computers and smartphones though, particularly when the device itself is identical to the $349 version just in a different colored (let's be honest you're not noticing that it is actually a different material in any circumstance, which is exactly why no other version comes in a gold color unlike many of Apple's other products) case.
I don't know. Maybe it works fine as a Linux Machine.
I'm sure it would work just fine as a Linux machine, that's probably what most people will buy it for. ChromeOS being a Linux distribution means it should at least have good hardware support for other Linux distributions.
Think of it - it is a nice piece of hardware without Microsoft tax on it.
I'd say that's probably negligible, I mean this is hardly any cheaper than any comparable Windows laptop and it's pretty silly to think they wouldn't be amortizing some of the operating system development costs in the price of this system. On a Mac you're paying an OSX "tax" and on this you'd be paying a ChromeOS "tax".
Just doing a "test" is never going to save you no matter what that test is, that is obvious to everybody. I can c-style cast at runtime and have no idea whether it worked or not until I run it or - if im expecting that it can fail in some cases - I can do a dynamic_cast and test the result and not crash.
Again, not saying it's necessarily bad, but there's no point in claiming type safety there.
Rubbish, C++ offers type safety features where C does not. And this is an advantage that C++ has over C. I've already pointed out that you get static_cast which is checked at compile time and dynamic_cast that you can check at runtime so contrary to your assertion you are better off with C++ as it provides type checking features.
It appears you're now trying to move the goal posts and say "well it isn't idiot-proof", and of course nobody ever claimed it was.
MS can also be seen as embracing (everything you listed and then some) and extending (MSPL) open source, all it needs now is to find a legal route to extinguish and it's won.
You have a very odd definition of "won", I don't think any company would consider going through the process of developing products, getting people to use them, releasing them as open source and then somehow killing them off and then proclaiming "yes we won!"..."won" what?
I still can't see what all the fuss is about from a practical usage point of view.
There's the gimmicky:
Send a heartbeat
Send a sketch
Send a vibration
And there's the contrived examples around particular scenarios where "Hey this is so awesome because I can do it without going to the painstaking process of taking my phone out of my pocket".
There's just nothing outstanding about it, maybe that will change in future. But with the original iPhone the advantages were obvious in changing the way things worked and disrupting an existing industry, with this, not so much.
I think the concern is around the potential benefits of the Apple Watch outweighing its own failures and the advantages of traditional timepieces. But i think the overlap market would be pretty small, this is still only practically applicable to iPhone users, it can't even get a full days' charge out of the battery and it's not exactly "exclusive" or a brand with much cachet.
I have an iPhone for its utilitarian purpose, if I wanted to feel special in my choice of smartphone I wouldn't have chosen the most common and best-selling one - "Hey, check out my Ford F150!" - it's likely the Apple Watch will be much the same. Personally I don't see the usefulness of it (yet), though of course that may change over time, and they certainly have strayed from the genuine benefits to marketing silly gimmicks like sharing your heartbeat.
There are no nice pieces, no craftsmanship; you're paying $10,000 for a disposable gold watch made by Chinese peons instead of a quality timepiece made by a master craftsman. There's a huge difference.
You're right, if you're wearing an expensive watch it's generally hand-crafted and not mass-produced like these are. Most people in that market would prefer to spend that on something more bespoke, perhaps one with a platinum case with inset diamond made by a jeweller, which wouldn't be hard given that the internals of the $10,000 one are no different to the $349 one.
My point isn't that C or C++ have bad type safety (or good type safety for that matter) just that if type safety is a real concern for you, C++ isn't really going to leave you better off than C.
Of course it is, C++ has static_cast and dynamic_cast for example, C does not.
In either one, the compiler won't save you if you make a type related error.
Yes it will. Here I'm trying to cast a string to an int, are you actually telling me you think this compiles?
std::string str("hello");
int val = static_cast<int>(str);
Note that dynamic_cast adds a RUNtime test, the compiler still doesn't save you.
So? Who cares whether the compiler saves you or a runtime test saves you?
Yes, you SHOULD use that for casting, but you don't HAVE to.
Of course you don't have to, do you really need to be corralled and work in a padded environment to force you to do things correctly? If so then C++ is not the language for you, in fact programming probably isn't for you.
Slashdot Mirror
If you "lock yourself into" using something, you are necessarily less free than someone who doesn't.
When do you ever do that?
The primary concern is toward the interests of the end-user.
Kind of, the primary concern is what the end user can do with the source code. This is the reason free software isn't particularly interesting to people, because end users are not programmers and thus are almost never interested in the source code. Free as in gratis is great for end users but freedom with respect to the source code isn't particularly useful to them.
It's a great model for advancing corporate interests, they can employ or contract developers to get their needs met and can even keep those changes inhouse if they wish. But there's no real model for advancing end user interests, that's just hobbyists and volunteers or some overlap with corporate needs. That's why free software really is lacking, sure end users could contract developers to implement a feature for them or spend time learning how to do it themselves but it's generally more cost effective to just pay the fee for a proprietary program that does it, cheaper for the end user and income for the developer.
10 might be a stepping stone where they offer both, but I have no doubt in mind SaaS windows will arrive with windows 10.
Not until having cellular data access in laptops becomes commonplace. Even now while they do have SaaS offerings they haven't replaced any software with a SaaS-only offering much less doing so with the operating system doing that would be fraught with problems and drive people straight to alternatives. What you suggest makes absolutely no sense at all.
It's been how long now and they still can't make a dent on the desktop. Same with macosx.
It's all about the programs it can run, no end user cares about the operating system, they care about what programs they can run on it. An operating system alone is not useful. This is why OSX does have at least some share of the desktop market, because there are quite a few popular professional-grade programs that run on it like Logic, Final Cut, Adobe Creative Suite, etc. so there are a group of professionals that can get their work done on a Mac. The choice of operating systems comes after the choice of programs for the task you want to accomplish. For example if you need Photoshop your options are Windows or OSX but not Linux or if you need Solidworks your only option is Windows, not OSX or Linux.
Linux as a desktop operating system in general has been perfectly fine for many years, but it doesn't have the application support that end users need which makes it pretty useless for most people.
So if systemd is so bad why not just fork one of the distros pre-systemd? Sure it's a big job to maintain a distro but apparently there's also a big group of people who don't like systemd too, if that's anything close to the majority then it will easily succeed.
A company tries to get their product to be more popular. Sounds like a good strategy. If it works, bully for them. If it doesn't, they'll try something else. Either people will use it or they won't.
I could see it potentially working on Android because it could be properly integrated but on iOS there's no way to make it a good user experience. I'm not a big fan of the voice command stuff on my phone on the rare occasion I do use it I just use Siri rather than unlocking my phone, finding the Google app and using their voice search.
That just describes the formation of the X.org foundation and the release of X11, not a fork.
What was the fork?
It doesn't really look like Microsoft was listening.
Did you provide feedback?
Your project gets forked and you lose your users. It's happened many times.
Like when? And what negative effect has that had?
Well it fits the definition of a Linux definition perfectly, I'm not particularly concerned if you want to say your iPod Touch runs a version of NextStep or care whether or not that's at all accurate.
You claimed the runtime saves you
Stop being an idiot, you know damn well the ability to avoid a problem based on a test is predicated on acting on the result of that test and not the test itself. You've been proven wrong and now you're trying weasel words to get out of it.
I can take such measures as well in C as in C++.
Demonstrate how do you do that both at runtime *and* compile time then.
Who's moving the goalposts again?
I already told you that in my last post, it's you. But you're still wrong on both counts, if you use the language features you can get type safety at the compiler and/or runtime levels.
What in THE hell are you talking about???
Why did you ask that question if you were going to go on a big rant about something you profess to not understand rather than waiting for an answer anyway?
To clarify: Other high-priced watches have some differentiating factor, they aren't just a cheap watch with the same case mass-produced from a different material. This is putting a Timex in the same shaped case but mass-producing it out of a more expensive material.
...also please don't take personal offence to the comparison of Apple and Timex, I know some Apple fans (not saying you're one of them) would do that and then pretend they don't understand the point on that basis.
If you use XCode then why would you want Linux?
The high-end Apple Watch is $10k (which is definitely not that high in the watch-as-a-fashion-statement-world) and actually has a solid 18k Gold-Ceramic "alloy" housing.
It's definitely high when the company it comes from is the maker of common computers and smartphones though, particularly when the device itself is identical to the $349 version just in a different colored (let's be honest you're not noticing that it is actually a different material in any circumstance, which is exactly why no other version comes in a gold color unlike many of Apple's other products) case.
I don't know. Maybe it works fine as a Linux Machine.
I'm sure it would work just fine as a Linux machine, that's probably what most people will buy it for. ChromeOS being a Linux distribution means it should at least have good hardware support for other Linux distributions.
Think of it - it is a nice piece of hardware without Microsoft tax on it.
I'd say that's probably negligible, I mean this is hardly any cheaper than any comparable Windows laptop and it's pretty silly to think they wouldn't be amortizing some of the operating system development costs in the price of this system. On a Mac you're paying an OSX "tax" and on this you'd be paying a ChromeOS "tax".
The runtime test doesn't necessarily save you.
Just doing a "test" is never going to save you no matter what that test is, that is obvious to everybody. I can c-style cast at runtime and have no idea whether it worked or not until I run it or - if im expecting that it can fail in some cases - I can do a dynamic_cast and test the result and not crash.
Again, not saying it's necessarily bad, but there's no point in claiming type safety there.
Rubbish, C++ offers type safety features where C does not. And this is an advantage that C++ has over C. I've already pointed out that you get static_cast which is checked at compile time and dynamic_cast that you can check at runtime so contrary to your assertion you are better off with C++ as it provides type checking features.
It appears you're now trying to move the goal posts and say "well it isn't idiot-proof", and of course nobody ever claimed it was.
MS can also be seen as embracing (everything you listed and then some) and extending (MSPL) open source, all it needs now is to find a legal route to extinguish and it's won.
You have a very odd definition of "won", I don't think any company would consider going through the process of developing products, getting people to use them, releasing them as open source and then somehow killing them off and then proclaiming "yes we won!"..."won" what?
I still can't see what all the fuss is about from a practical usage point of view.
There's the gimmicky:
And there's the contrived examples around particular scenarios where "Hey this is so awesome because I can do it without going to the painstaking process of taking my phone out of my pocket".
There's just nothing outstanding about it, maybe that will change in future. But with the original iPhone the advantages were obvious in changing the way things worked and disrupting an existing industry, with this, not so much.
I think the concern is around the potential benefits of the Apple Watch outweighing its own failures and the advantages of traditional timepieces. But i think the overlap market would be pretty small, this is still only practically applicable to iPhone users, it can't even get a full days' charge out of the battery and it's not exactly "exclusive" or a brand with much cachet.
I have an iPhone for its utilitarian purpose, if I wanted to feel special in my choice of smartphone I wouldn't have chosen the most common and best-selling one - "Hey, check out my Ford F150!" - it's likely the Apple Watch will be much the same. Personally I don't see the usefulness of it (yet), though of course that may change over time, and they certainly have strayed from the genuine benefits to marketing silly gimmicks like sharing your heartbeat.
There are no nice pieces, no craftsmanship; you're paying $10,000 for a disposable gold watch made by Chinese peons instead of a quality timepiece made by a master craftsman. There's a huge difference.
You're right, if you're wearing an expensive watch it's generally hand-crafted and not mass-produced like these are. Most people in that market would prefer to spend that on something more bespoke, perhaps one with a platinum case with inset diamond made by a jeweller, which wouldn't be hard given that the internals of the $10,000 one are no different to the $349 one.
My point isn't that C or C++ have bad type safety (or good type safety for that matter) just that if type safety is a real concern for you, C++ isn't really going to leave you better off than C.
Of course it is, C++ has static_cast and dynamic_cast for example, C does not.
In either one, the compiler won't save you if you make a type related error.
Yes it will. Here I'm trying to cast a string to an int, are you actually telling me you think this compiles?
std::string str("hello");
int val = static_cast<int>(str);
Note that dynamic_cast adds a RUNtime test, the compiler still doesn't save you.
So? Who cares whether the compiler saves you or a runtime test saves you?
WiFi is not the bottleneck, so why would I tie myself to a wire?
What is the real world speed of WiFi in general?
Where are the stats that show when magsafe became obsolete and no longer required?
Yes, you SHOULD use that for casting, but you don't HAVE to.
Of course you don't have to, do you really need to be corralled and work in a padded environment to force you to do things correctly? If so then C++ is not the language for you, in fact programming probably isn't for you.