The magic is how you approach solving problems. One of the problems with Common LISP is that it lets you write imperative like code. The old joke about a FORTRAN programmer being able to write FORTRAN in any language comes to mind. That's why I prefer to teach with Haskell because it takes away so much of how you commonly write code that it forces the mental paradigm shift.
Passing blocks of code by itself is not a paradigm shift. Genuinely understanding that all code is data, that you can think of data as a function that acts on functions through duality is the paradigm shift. As the abstractions build the complexity falls away. And from there you really do see the silver bullet.
Don't you think the Mac introduced a new kind of black-box mentality to the PC world?
No. I think the early Mac was a failure at doing this. It certainly was the intent but Mac sold poorly for many years and ended up in niches liked education and desktop publishing. If you look at the early business PC users they frequently mostly ran one application: WordStar / WordPerfect, Lotus 1-2-3, Harvard Graphics.... These applications had OS features (like a mini shell) so you never had to exit. Microsoft would in the 1990s push the office suite but again that hadn't happened yet. A Word Perfect user had a very black box experience. A Lotus 1-2-3 user had a very black box experience. It wasn't the Mac that created this attitude it was the success of office productivity applications in attracting non hobbyists to the platform and that mostly happened on the MS-DOS PC machines.
What changed was during the later 1980s was the attraction of non-hobbyists who saw the machine as a tool. Then came the mainstreaming to corporations as the interface of choice and suddenly the overwhelming majority of users were non hobbyists.
at least at the human level of spreadsheets and word processing.
Spreadsheets and word processing were paradigms that came from Personal Computing. A word processor was an advanced electric typewriter. Document authoring systems on mainframes and minis were having multiple authors compute documents with complex rulesets. They were vastly more sophisticated than the one off documents that word processing handled. Word Processing as it got more sophisticated killed off the electric typewriters and then killed off all but the most niche document authoring systems. That process took decades and hadn't happened by the early 1980s.
In the case of spreadsheets again there was something of a paradigm shift. Accounting and financial systems on mainframes and minis were quite complex. Spreadsheets were designed around allowing a single person to do the easiest 20% of that kind of work easily. As time when on a generation was created that understood spreadsheets and didn't understand the more complex applications and thus spreadsheets are often to this day an interface more complex backends. But that was not the situation in the early 1980s.
No question PCs disrupted big boxes. But that happened in the 1990s. It can be very hard to distinguish between the 1990s disruption and the invention of PCs in the late 1970s and growth in the 1980s but you have to do that when talking about the history. What was in place when.
I think you can find that experience in embedded systems. It may be hard to understand an entire computer. It isn't hard to understand something like a ram controller chip or a controller on a DVD drive. The 8086 was a professional chip designed for professional CP/M application assembly language programmers to be able to write more advanced applications. Clearly after that the chips could get more complicated because compilers were the target not assembly.
There are really kits for kids. Mindstorms for example do what you are talking about.
There are far more people today doing low level work. Storage administrators do low level storage I/O professionally. People who write graphical device drivers. Game tookit designers.... You just have to work in hardware not software. Applications don't use low level interfaces but that doesn't mean controller cards, chipsets, and to some extent OS components don't.
The 8 bit understandable systems were toys then. You were just younger and were OK with toys.
Most real work in corporations was being done on mainframes and minis using terminals for user input. It was during the 1990s that PCs took over more and more of the computing space in corporations first at the input level and fringe applications. Then Windows Server, Solaris... replaced 1 tier computing mini / mainframe with 2 tier client server models. Later the web came in and took over more of the space. You didn't understand real computers in the early 1980s, you understood fun toys. Nothing wrong with that but lets be realistic. 12 year old shanen was not running manufacturing scheduling systems on his VIC-20.
True. I felt terribly robbed when I was a child that analogue computers using vacuum tubes and only running a single small program weren't available anymore and I was stuck with digital computers that could run arbitrary code. Amazing I survived this deprivation.
I worked on both IRIX and SunOS. Those machines often ran $20k. There pretty much were no kids involved, those were often reserved for graduate students and even college kids used worse systems. They also were used professionally.
As for Javascript being a step back from C. They aren't remotely related to one another. JavaScript is an interpreted applications language that operates safely cross platform on large computer networks. In the time of IRIX or SunOS interpreted languages were things like BASIC, ksh, csh. And I think that most people would agree that applications in JavaScript are quite a bit more fun than typical shell scripts. The analogy to C would be platforms like C#.net which are vastly more sophisticated and powerful than the C of the early 1990s.
The Atari, Commodore 64, Apple 2... were hobbyist machines generally designed to interest children in computers. Windows 10 machines are generally work machines generally designed for adults who have some objective not related to entertainment that requires computer assistance to accomplish it. The systems you are describing are the ones that killed the DIY kits from the 1970s, they were part of migration away from hobbyist culture in that they allowed kids (and middle class families) and not adult hobbyists to have a personal computer.
That being said though there are plenty of hobbyist / enthusiast aspects of computing. There are fun simulators for just about everything. There are DIY kits that are far more educational (and stepped) than the ones from the 1970s since the parts are now so cheap. There are tons of educational programming languages. It may feel a little more empty because computers are a profession not a hobby for millions of people. But objectively hobbyists have it much better today than in the golden age.
Britannica was disrupted by Wikipedia. There was a window of time where Wikipedia had a quantity of articles which was comparable through to the point that while Wikipedia's quantity was larger than Britannica didn't just dwarf it to the extent that Britannica was simply not in the same league. During that period of time Wikipedia was experiencing exponential growth. Britannica CD (later DVD) version was available but not free (or they had an online version). The argument was mainly about whether one should pay for the expert editing vs. the open model of Wikipedia. By the time the press started covering the two wikipedia was probably the size of all specialized encyclopedia's combined (i.e. math encyclopedia, medical encyclopedia...) Wikipedia's growth has slowed considerably from those days, its linear growth not exponential anymore. Still wikipedia adds a new encyclopedia Britannica worth of content every 2 months.
Britannica was a wonderful product. Had Britannica early on combined many of the specialized reference works (like the OED, specialized encyclopedias..., S&Ps investing data...) into a single site and made it accessible at a reasonable cost along with a news, there likely never would have been a wikipedia. Today such a thing might be possible because everyone now agrees the internet is going to chew up the margins that used to exist. A good quality site for academic content still doesn't exist at a cost affordable to non academics. But even that is a maybe. The open model was incredibly successful.
I liked Wikipedia in the inclusionist days. It would probably have 20-100x the number of articles it has today were it not for the deletionism. The articles themsleves would be far more complete. Deletionism has been incredibly expensive. OTOH the deletionists were somewhat successful in raising the bar for accuracy. I don't think it was worth the cost but unlike say 4 years ago the changes are quite evident.
Linux is pretty darn successful: 1st place dominant in server / cloud, dominant player in embedded about $10 / unit, dominant in mobile. dominant in supercomputing. There are areas Linux doesn't do well in, like its original target of desktop, but its far from a failure by any reasonable standard.
SCO failed because of inexpensive server big box Unix, Linux and Windows NT. There were better OSes for X86. There are better Unixes at around the $7k price point. And by the mid 1990s there was even a better Unix for x86. And with the invention of the 486 the advantages of the x86 / i860 dual motherboards became far too niche. Essentially SCO never found a large enough market to sustain the level of development needed to compete. An example of how
I think the way you generally handle something like this is through void pointers. Part of the data structure becomes a void function pointer. That's perfectly legal in C. You just have to be careful that when you set out to run a step in the evaluation you have all your types in line.
How (and why) would you convert your Lisp code into C
Well that's pretty easy. There are tons of simple Lisp compilers. Heck building a LISP used to be an exercise for many years included (full source) with Microsoft's C compiler.
If you want to argue Apple did the marketing better, and correctly judged that being seen as the innovator by rushing a crappier product to market to beat the competition by a few months was worth the tradeoffs from a business point of view, you'll get little argument from me.
OK now we have established they were first with that crucial triple. I think you are still wrong on both months and "crappier". But I wanted to argue the dates first.
1) Everyone uses glass. Presumably if plastic were better and easy to manufacture (in 2007) we'd be seeing phones with it now.
2) As far as high speed web rendering. Remember the iphone supported wifi. It also was sold exclusively with AT&T's unlimited data 3G plan. The whole point of the device was to drive up the demand for data usage.
3) Finally on high speed animations I think you would be hard pressed to find a phone remotely less laggy. This has remained true more or less to this day using a comparable interface. Vertical integration has particularly paid off here as Apple has often been able to use objectively slower hardware combined with customized OS and applications to achieve much less lag. Nokia who you have pointed to many times had horrific lag problems.
As for Nokia being the competition. In the United States Nokia wasn't the competition. Nokia USA was dismal. Nokia did not focus on the USA market. Even when the USA became interesting Nokia wasn't able to integrate Nokia USA (the sales division) into their corporate decision making many years later. As for Maemo, Maemo used a resistive touch screen and required a stylus. It did not involve the critical triple and couldn't. Nokia internally had engineers who saw the advantage of the triple but couldn't get the changes into Gnome fast enough. A failure that both the Gnome community and Nokia reacted strongly too by restructuring. Could an alternative universe Nokia have won, absolutely. But in this one they dithered didn't make critical choices when there still were two sides lost their lead, then fell behind then died. I was working with Nokia USA during those last years when they couldn't either execute or not on exploiting the gap that Apple created in enterprise phones. In the end the hardware guys ignored Elop advice on ecosystems and they focused instead on a few hardware features.
Nokia is a perfect example of how good Apple is in developing a total package that is often unappreciated by technical people. Hardwarewise even when Nokia was hemorrhaging share to Apple they were quite often from a hardware perspective better phones. They were however vastly inferior phones from a software (OS in particular) and then from ecosystem perspective. The company was directionless.
Nokia also disproves your marketing theory. Nokia lost enterprise to Apple at a time when Apple was doing anti-marketing in enterprise phones. Apple wanted RIM or Nokia/Microsoft to take the enterprise market, they refused to make the concessions that enterprise customers wanted, and still the iphone's total experience was so much superior that even when anti-marketing Apple ended up winning with the move to BYOD.
As an aside on dates, the N900 came out 2 1/2 years after the iPhone 1 and still had a keyboard and resistive touchscreen. As an aside one of the things on ecosystems that caught Nokia off guard was the power of a closed ecosystem. They had never seriously considered what a closed ecosystem done well would look like when they focused so heavily on an open ecosystem
Finally on the claim that Apple's choices were obvious let me just point you to another of your comments, "And a quality hard keyboard is still the only sane input method for people who are serious about things. A lack of a keyboard was a money saving measure, that's all." You yourself a decade after Apple's approach still don't appreciate how important lack of a keyboard is. RIM was the primary vendor who was devastated by failing to appre
Your claim was that they didn't do anything innovative. The question is not how well they implemented the triple (though I dispute with you that at the time they didn't have a very good implementation of all 3) but rather that they presented this triple. They saw the potential of using these things together to build a new UI. Nokia has slipped with Maemo and was focusing on the MeeGo project which would still take years. MeeGo potentially could have been like that but too many factions wanted Symbian compatibility and thus at the time no evidence they were moving towards what would become the standard for smartphones.
As for Android at the time it was working on a blackberry clone with small screens and a keyboard. So no at the time it also was not ahead and wouldn't be even similar for almost 2 years.
In 2007 Apple released the first phone with: a) capacitive touchscreen as the primary or sole means of input b) animation based interaction c) high speed web rendering
There was no other product with that triple. That triple is now the standard. Obviously since they were the only one that was unique and since it has become more or less the standard it was quite innovative.
I think you should watch the introductory video to get an idea of how much Job's ideas contrasted with the competition at the time: https://www.youtube.com/watch?...
Big features that are missing: * Text size adjustment that consistently works (a lot of the population has trouble reading on their phone). * Better integration with cars for safety (cell phones still kill and injure a tremendous number of people when mixed with automobiles). * Voice commands that work * Ability to switch to gloved mode for colder climates * Better ability to track battery and data usage on an app or task basis
Being beaten on a single feature and being beaten across the board are two very different things. Apple was much closer to having an overall inferior problem about 5 years ago when high end Android was way ahead from a hardware perspective and Apple was competing on usability and richness of applications. With Android having moved down market now the major feature on which Apple lags is lower cost models. Used are the only lower cost Apple models and these lag in many respects.
What positions are Trump supporters accused of holding that there is not substantial demographic evidence showing that at least millions of his supporters do hold? In short what are you talking about?
Buy a few million of them a year. The old pros didn't sell very well, they were often seen as too big and too noisy. They didn't mesh with Apple's brand image. The trash can sold well for many months.
Apple seems to be killing the desktop pros off the killed their server line. Apple liked a narrow product line. To get diversification you need volume sales. The numbers aren't there. The numbers certainly aren't there for a physically large machine.
Cell phone used to use peak and off peak for minutes. Things like $10 / mo for 20 peak minutes and 500 off peak minutes. Clearly designed to encourage off peak usage. The ratios started to change as cell phones began to be used socially and they couldn't do anything that extreme. Data is similar. Sure there are peaks but it is gently rolling hills not the sort of sharp drop offs that make a variable pricing scheme make sense.
These are public companies their spend for the networks are public documents. The cost of spending to create 3G and then LTE was many billions for each of them every year. They have huge debts from it and your job is to indirectly pay down the bond holders. It is not pennies on the dollar to provide when you count the capital cost. That's why most of the cell phone providers went broke.
Slashdot Mirror
The magic is how you approach solving problems. One of the problems with Common LISP is that it lets you write imperative like code. The old joke about a FORTRAN programmer being able to write FORTRAN in any language comes to mind. That's why I prefer to teach with Haskell because it takes away so much of how you commonly write code that it forces the mental paradigm shift.
Passing blocks of code by itself is not a paradigm shift. Genuinely understanding that all code is data, that you can think of data as a function that acts on functions through duality is the paradigm shift. As the abstractions build the complexity falls away. And from there you really do see the silver bullet.
Don't you think the Mac introduced a new kind of black-box mentality to the PC world?
No. I think the early Mac was a failure at doing this. It certainly was the intent but Mac sold poorly for many years and ended up in niches liked education and desktop publishing. If you look at the early business PC users they frequently mostly ran one application: WordStar / WordPerfect, Lotus 1-2-3, Harvard Graphics.... These applications had OS features (like a mini shell) so you never had to exit. Microsoft would in the 1990s push the office suite but again that hadn't happened yet. A Word Perfect user had a very black box experience. A Lotus 1-2-3 user had a very black box experience. It wasn't the Mac that created this attitude it was the success of office productivity applications in attracting non hobbyists to the platform and that mostly happened on the MS-DOS PC machines.
What changed was during the later 1980s was the attraction of non-hobbyists who saw the machine as a tool. Then came the mainstreaming to corporations as the interface of choice and suddenly the overwhelming majority of users were non hobbyists.
at least at the human level of spreadsheets and word processing.
Spreadsheets and word processing were paradigms that came from Personal Computing. A word processor was an advanced electric typewriter. Document authoring systems on mainframes and minis were having multiple authors compute documents with complex rulesets. They were vastly more sophisticated than the one off documents that word processing handled. Word Processing as it got more sophisticated killed off the electric typewriters and then killed off all but the most niche document authoring systems. That process took decades and hadn't happened by the early 1980s.
In the case of spreadsheets again there was something of a paradigm shift. Accounting and financial systems on mainframes and minis were quite complex. Spreadsheets were designed around allowing a single person to do the easiest 20% of that kind of work easily. As time when on a generation was created that understood spreadsheets and didn't understand the more complex applications and thus spreadsheets are often to this day an interface more complex backends. But that was not the situation in the early 1980s.
No question PCs disrupted big boxes. But that happened in the 1990s. It can be very hard to distinguish between the 1990s disruption and the invention of PCs in the late 1970s and growth in the 1980s but you have to do that when talking about the history. What was in place when.
I think you can find that experience in embedded systems. It may be hard to understand an entire computer. It isn't hard to understand something like a ram controller chip or a controller on a DVD drive. The 8086 was a professional chip designed for professional CP/M application assembly language programmers to be able to write more advanced applications. Clearly after that the chips could get more complicated because compilers were the target not assembly.
There are really kits for kids. Mindstorms for example do what you are talking about.
There are far more people today doing low level work. Storage administrators do low level storage I/O professionally. People who write graphical device drivers. Game tookit designers.... You just have to work in hardware not software. Applications don't use low level interfaces but that doesn't mean controller cards, chipsets, and to some extent OS components don't.
The 8 bit understandable systems were toys then. You were just younger and were OK with toys.
Most real work in corporations was being done on mainframes and minis using terminals for user input. It was during the 1990s that PCs took over more and more of the computing space in corporations first at the input level and fringe applications. Then Windows Server, Solaris... replaced 1 tier computing mini / mainframe with 2 tier client server models. Later the web came in and took over more of the space. You didn't understand real computers in the early 1980s, you understood fun toys. Nothing wrong with that but lets be realistic. 12 year old shanen was not running manufacturing scheduling systems on his VIC-20.
True. I felt terribly robbed when I was a child that analogue computers using vacuum tubes and only running a single small program weren't available anymore and I was stuck with digital computers that could run arbitrary code. Amazing I survived this deprivation.
I worked on both IRIX and SunOS. Those machines often ran $20k. There pretty much were no kids involved, those were often reserved for graduate students and even college kids used worse systems. They also were used professionally.
As for Javascript being a step back from C. They aren't remotely related to one another. JavaScript is an interpreted applications language that operates safely cross platform on large computer networks. In the time of IRIX or SunOS interpreted languages were things like BASIC, ksh, csh. And I think that most people would agree that applications in JavaScript are quite a bit more fun than typical shell scripts. The analogy to C would be platforms like C#.net which are vastly more sophisticated and powerful than the C of the early 1990s.
The Atari, Commodore 64, Apple 2... were hobbyist machines generally designed to interest children in computers. Windows 10 machines are generally work machines generally designed for adults who have some objective not related to entertainment that requires computer assistance to accomplish it. The systems you are describing are the ones that killed the DIY kits from the 1970s, they were part of migration away from hobbyist culture in that they allowed kids (and middle class families) and not adult hobbyists to have a personal computer.
That being said though there are plenty of hobbyist / enthusiast aspects of computing. There are fun simulators for just about everything. There are DIY kits that are far more educational (and stepped) than the ones from the 1970s since the parts are now so cheap. There are tons of educational programming languages. It may feel a little more empty because computers are a profession not a hobby for millions of people. But objectively hobbyists have it much better today than in the golden age.
Britannica was disrupted by Wikipedia. There was a window of time where Wikipedia had a quantity of articles which was comparable through to the point that while Wikipedia's quantity was larger than Britannica didn't just dwarf it to the extent that Britannica was simply not in the same league. During that period of time Wikipedia was experiencing exponential growth. Britannica CD (later DVD) version was available but not free (or they had an online version). The argument was mainly about whether one should pay for the expert editing vs. the open model of Wikipedia. By the time the press started covering the two wikipedia was probably the size of all specialized encyclopedia's combined (i.e. math encyclopedia, medical encyclopedia...) Wikipedia's growth has slowed considerably from those days, its linear growth not exponential anymore. Still wikipedia adds a new encyclopedia Britannica worth of content every 2 months.
Britannica was a wonderful product. Had Britannica early on combined many of the specialized reference works (like the OED, specialized encyclopedias..., S&Ps investing data...) into a single site and made it accessible at a reasonable cost along with a news, there likely never would have been a wikipedia. Today such a thing might be possible because everyone now agrees the internet is going to chew up the margins that used to exist. A good quality site for academic content still doesn't exist at a cost affordable to non academics. But even that is a maybe. The open model was incredibly successful.
I liked Wikipedia in the inclusionist days. It would probably have 20-100x the number of articles it has today were it not for the deletionism. The articles themsleves would be far more complete. Deletionism has been incredibly expensive. OTOH the deletionists were somewhat successful in raising the bar for accuracy. I don't think it was worth the cost but unlike say 4 years ago the changes are quite evident.
Linux is pretty darn successful: 1st place dominant in server / cloud, dominant player in embedded about $10 / unit, dominant in mobile. dominant in supercomputing. There are areas Linux doesn't do well in, like its original target of desktop, but its far from a failure by any reasonable standard.
SCO failed because of inexpensive server big box Unix, Linux and Windows NT. There were better OSes for X86. There are better Unixes at around the $7k price point. And by the mid 1990s there was even a better Unix for x86. And with the invention of the 486 the advantages of the x86 / i860 dual motherboards became far too niche. Essentially SCO never found a large enough market to sustain the level of development needed to compete. An example of how
I think the way you generally handle something like this is through void pointers. Part of the data structure becomes a void function pointer. That's perfectly legal in C. You just have to be careful that when you set out to run a step in the evaluation you have all your types in line.
How (and why) would you convert your Lisp code into C
Well that's pretty easy. There are tons of simple Lisp compilers. Heck building a LISP used to be an exercise for many years included (full source) with Microsoft's C compiler.
A few examples:
http://www.buildyourownlisp.co...
http://howtowriteaprogram.blog...
http://sbcl.sourceforge.net/
http://clisp.sourceforge.net/
A little off topic but a Lisp in Java by a master of the programming craft which is worth looking at: http://norvig.com/jscheme.html
And finally also off topic but the original LISP in assembly: http://www.softwarepreservatio...
In short the way you convert LISP to C is writing a LISP interpreter or compiler.
OK now we have established they were first with that crucial triple. I think you are still wrong on both months and "crappier". But I wanted to argue the dates first.
1) Everyone uses glass. Presumably if plastic were better and easy to manufacture (in 2007) we'd be seeing phones with it now.
2) As far as high speed web rendering. Remember the iphone supported wifi. It also was sold exclusively with AT&T's unlimited data 3G plan. The whole point of the device was to drive up the demand for data usage.
3) Finally on high speed animations I think you would be hard pressed to find a phone remotely less laggy. This has remained true more or less to this day using a comparable interface. Vertical integration has particularly paid off here as Apple has often been able to use objectively slower hardware combined with customized OS and applications to achieve much less lag. Nokia who you have pointed to many times had horrific lag problems.
As for Nokia being the competition. In the United States Nokia wasn't the competition. Nokia USA was dismal. Nokia did not focus on the USA market. Even when the USA became interesting Nokia wasn't able to integrate Nokia USA (the sales division) into their corporate decision making many years later. As for Maemo, Maemo used a resistive touch screen and required a stylus. It did not involve the critical triple and couldn't. Nokia internally had engineers who saw the advantage of the triple but couldn't get the changes into Gnome fast enough. A failure that both the Gnome community and Nokia reacted strongly too by restructuring. Could an alternative universe Nokia have won, absolutely. But in this one they dithered didn't make critical choices when there still were two sides lost their lead, then fell behind then died. I was working with Nokia USA during those last years when they couldn't either execute or not on exploiting the gap that Apple created in enterprise phones. In the end the hardware guys ignored Elop advice on ecosystems and they focused instead on a few hardware features.
Nokia is a perfect example of how good Apple is in developing a total package that is often unappreciated by technical people. Hardwarewise even when Nokia was hemorrhaging share to Apple they were quite often from a hardware perspective better phones. They were however vastly inferior phones from a software (OS in particular) and then from ecosystem perspective. The company was directionless.
Nokia also disproves your marketing theory. Nokia lost enterprise to Apple at a time when Apple was doing anti-marketing in enterprise phones. Apple wanted RIM or Nokia/Microsoft to take the enterprise market, they refused to make the concessions that enterprise customers wanted, and still the iphone's total experience was so much superior that even when anti-marketing Apple ended up winning with the move to BYOD.
As an aside on dates, the N900 came out 2 1/2 years after the iPhone 1 and still had a keyboard and resistive touchscreen. As an aside one of the things on ecosystems that caught Nokia off guard was the power of a closed ecosystem. They had never seriously considered what a closed ecosystem done well would look like when they focused so heavily on an open ecosystem
Finally on the claim that Apple's choices were obvious let me just point you to another of your comments, "And a quality hard keyboard is still the only sane input method for people who are serious about things. A lack of a keyboard was a money saving measure, that's all." You yourself a decade after Apple's approach still don't appreciate how important lack of a keyboard is. RIM was the primary vendor who was devastated by failing to appre
Your claim was that they didn't do anything innovative. The question is not how well they implemented the triple (though I dispute with you that at the time they didn't have a very good implementation of all 3) but rather that they presented this triple. They saw the potential of using these things together to build a new UI. Nokia has slipped with Maemo and was focusing on the MeeGo project which would still take years. MeeGo potentially could have been like that but too many factions wanted Symbian compatibility and thus at the time no evidence they were moving towards what would become the standard for smartphones.
As for Android at the time it was working on a blackberry clone with small screens and a keyboard. So no at the time it also was not ahead and wouldn't be even similar for almost 2 years.
Sort of you have a math error it should be -50% and that number actually makes sense.
In your scenario: industry profits = $1b - $3b = $-2b.
Apple profit $1b = -50% of $-2b.
or flipping the negative
industry losses = $2b.
Apple's loss = $-1b = -50% of $2b.
A negative profit is a loss and visa versa.
Actually they most certainly do represent physical things. For example the equations for potential energy of a particle.
That being said finance is only semi physical and negative numbers do exist.
In 2007 Apple released the first phone with:
a) capacitive touchscreen as the primary or sole means of input
b) animation based interaction
c) high speed web rendering
There was no other product with that triple. That triple is now the standard. Obviously since they were the only one that was unique and since it has become more or less the standard it was quite innovative.
I think you should watch the introductory video to get an idea of how much Job's ideas contrasted with the competition at the time: https://www.youtube.com/watch?...
Big features that are missing:
* Text size adjustment that consistently works (a lot of the population has trouble reading on their phone).
* Better integration with cars for safety (cell phones still kill and injure a tremendous number of people when mixed with automobiles).
* Voice commands that work
* Ability to switch to gloved mode for colder climates
* Better ability to track battery and data usage on an app or task basis
Being beaten on a single feature and being beaten across the board are two very different things. Apple was much closer to having an overall inferior problem about 5 years ago when high end Android was way ahead from a hardware perspective and Apple was competing on usability and richness of applications. With Android having moved down market now the major feature on which Apple lags is lower cost models. Used are the only lower cost Apple models and these lag in many respects.
What positions are Trump supporters accused of holding that there is not substantial demographic evidence showing that at least millions of his supporters do hold? In short what are you talking about?
Buy a few million of them a year. The old pros didn't sell very well, they were often seen as too big and too noisy. They didn't mesh with Apple's brand image. The trash can sold well for many months.
Apple seems to be killing the desktop pros off the killed their server line. Apple liked a narrow product line. To get diversification you need volume sales. The numbers aren't there. The numbers certainly aren't there for a physically large machine.
Cell phone used to use peak and off peak for minutes. Things like $10 / mo for 20 peak minutes and 500 off peak minutes. Clearly designed to encourage off peak usage. The ratios started to change as cell phones began to be used socially and they couldn't do anything that extreme. Data is similar. Sure there are peaks but it is gently rolling hills not the sort of sharp drop offs that make a variable pricing scheme make sense.
These are public companies their spend for the networks are public documents. The cost of spending to create 3G and then LTE was many billions for each of them every year. They have huge debts from it and your job is to indirectly pay down the bond holders. It is not pennies on the dollar to provide when you count the capital cost. That's why most of the cell phone providers went broke.