There are no two things more opposite to each other than OOP and functional programming. Functional programming is about FUNCTIONS, which are those that return a value that is completely determined by their arguments, not depending on, or modifying, any state. OOP is all about calling METHODS, which have values that are determined by both the arguments and the state of the object they're called on, and often are meant to modify that state.
"A Discipline of Programming" is more specifically, about how to write programs that are correct by construction, than it is about how to prove that already written programs are correct.
IMHO what I understand of 30 years of history of the software industry, and my experiences of formal methods and programming language theory in the research world on the one hand, and real-world professional software development on the other hand, validate everything that's in this book.
Systematically buying everything you read new and hardcover is pretty uncommon anyways, except maybe for people who have amounts of disposable income that make the convenience factor the only that counts anyway.
Reading paperback, pocket-book and other discounted editions, buying used books on peer-to-peer marketplaces, borrowing from a library, from a friend, your spouse, of from your parent's bookshelf, all these are not taken into account. If we divide that value of each book that I've read in my life, by the total number of people who've read that particular physical book too, and take the average, we're most definitely in the sub-$1 area, even counting all of the technical and textbook stuff.
I'd agree there's definitely the potential for a "$99-style" psychological effect of labeling a software product "1.0". However instead of outright lying about how many versions of this product have actually already been field-tested, I think it is more elegant to completely remove the version number from the product information and marketing, instead just leaving at most a build number in the Help/About dialog.
Maybe there is proportionally more PROFESSIONAL photographers that use Macs for image processing, but Photosynth is clearly a crowdsourcing application and its target audience is everybody, since now everybody has a digital camera and everybody processes their amateur images on their home computer. So we're back at the general PC/Mac split in home computers.
Actually Intel has one big reason to wish for as many x86 vendors as possible : in case of a collapse of AMD, it would take as many other competitors as possible to keep Intel out of trouble from the anti-trust authorities.
Word uses essentially the same algorithm to manage paragraph flow as TeX does, just more sophisticated in the areas of inserting non-text objects such as images and much more sophisticated for mixing languages of different scripts.
The reason TeX output looks "better" to some is mostly that it looks different. TeX was created exclusively for dealing with European languages, and it was created for creating printed books, with only the typographical conventions of printed books, before desktop publishing really existed. By default it looks different and "more bookish", kind of quaint, people interpret that as being "better".
Don't get me wrong I love TeX and its legacy is gigantic (again, Knuth's algorithms were reused in modern word processor). It's just not the best tool anymore.
XML is good because it promotes keeping orthogonal, two very different problems : structuring data and encoding it. Tying these two problems creates complications down the road, as entangling independent areas of information processing and algorithm sciences usually does.
Frankly, the issue is a bit old. Who cares ? These keys do a few useful things under Windows, and I, for one, have grown used to count on them for shortcuts that wouldn't exist without some third party, custom configured hotkey app.
RISC vs CISC was the architecture flamewar of the late 1980s. Welcome to the 21th century, you'll like it here. It's a world when, since the late 90s, the ISA (instruction set architecture), is so abstracted away from the actual micro-architecture of microprocessor, as to make it completely pointless to make distinctions between the two. Modern processors are RISC, they are CISC, they are vector machines, they're everything you want them to be. Move on, the modern problems are now in multi-core architecture and their issues of memory coherence, cache sharing, memory bandwidth, interlocking mechanisms, uniform vs non-uniform, etc. The "pure RISC" standard bearers of yore have disappeared or have been expelled from the personnal computing sphere (remember Apple ditching PowerPC ? Alpha anyone ? Where are those shiny MIPS-based SGIs gone?). Even Intel couldn't impose a new ISA on its own (poor adoption of IA-64). The only RISC ISA that has any existence in the personnal computing arena, including mobile, is ARM, but precisely, they do only mobile. There's really no reason at all to build any device on which you plan to run generic OSes and rich computing experience on anything else than x86 or x86-64 machines.
I do believe that one of the few major adjustments that humanity will take under constraint in our particular lifetime is to shift from fishing in the ocean, to fish farming. It's nothing really, mankind has already done that with great success with most of its other food supplies : hunting was replaced by cattle and piking fruit and berries in the forrest was replaced by crops. That might imply some types of fish to get more rare, just as we haven't eaten nearly as much boar as we used to since we started breeding pigs.
That's ridiculous. Fossil fuels are not at all free : they need to be mined and refined. And I don't see how "more expensive" energies such as wind or solar power would be.
I resent people who are stating that 5, or 6, or 7 billions is too many and that the growth of world population should make us worry. I would like to point out that, compared to the era when world population was less than 1 billion, the average life expectancy, quality of life and, yes, access to ressources and opportunities has dramatically increased for our species. How far is the time when a single pandemic, natural disaster or mass migration would wipe out a third of a continent population and make whole civilization disappear from History ? Notwithstanding the current price fluctuations that call for natural adjustments in production and distribution systems, REAL hunger, the one where the basic intake of food necessary for survival simply isn't available within reach, has been reduced to cases relatively limited in scope and mostly due to geopolitical circumstances rather than natural resource limitations.
The video game market is one of the most expensive and toughest to crack into of all global markets. Only two new companies managed to make it from scratch in more than 10 years : Sony and Microsoft, each of them gambling huge amounts of money over many years. Apple certainly "could" theoretically make it, it has the talent and the cash, but as a business decision it would not make sense for a company that is mostly known for breaking changes and creating whole new markets. As for the "mobile" focus, doesn't make any difference : that field is crowded already, by Nintendo and Sony no less.
Avoiding jamming on mechanical typewriters is the same as having as much alternating between left hand and right hand as possible for successive letters, which makes typing faster and more comfortable. The story that qwerty slows typing down is one lame urban myth. One case of perfect match between engineering practicality and ergonomics, which is why it was so succesfull.
I didn't mean that the MIT didn't have brilliant academics and important contributions to their fields with scientific discoveries and seminal publications.
I was just pointing out that the Media Lab, which has hardly anything to do with all this, has this tradition of hyping klunky prototypes of improbable gadgets and making broad statements about what the future will be. I remember how for they had announced for years in the late 90s, the advent of computerized doorknobs with an IP address.
I notice also that your examples of contributions to computer technology are a bit weak. The Lego Mindstorms is a toy with a mixed success and not a noticeable contribution to the everyday use of robotics. Multics was a failure in itself and in the underlying concept (a centralized architecture providing commoditized computing power). X Window is a complete failure in its underlying concept of centralized applications relying on display/input "servers" with a very fine grained interaction protocol. The concept of centralized apps with remote display/input is only now being realized under a radically different architecure (browser/ajax/flash/silverlight).
Stallman was never capable of delivering the OS he had planned, something that a 20-year old undergrad student finally did in a dorm room in Finland.
SICP is a museum piece that people outside of MIT have read mostly out of curiosity and snobbery. I am myself of a formal semantics / FP background and a huge fan of Scheme but I cannot say that the peculiarity of teaching undergrad programming in scheme is a raging trend accross the campuses in the nation.
People at MIT are notoriously good at creating buzz around the concepts, demos, protypes and inventions that they come up with, especially at the Media Lab. Unfortunately, like everything that happens in academia, the signal to noise ratio is what it is and most of it has no future, sometimes for blatant reasons that one doesn't need to be a very sharp V.C. to figure out. Unfortunately that creates the impression that they are really a bunch of clowns that come up with useless stuff on a regular basis.
So much time and energy wasted trying to come up with fancy data processing metaphors that refer to "intuitive" concepts and situations... Like if educated adults were retarded children and needed to be spoonfed some special way of handling their environment so that they don't find it too challenging. This is mostly useless, the signal/noise in this kind of "breakthrough" research is historically one order of magnitude lower that what valuable R&D dollars are supposed to be spent for.
99.99% of what we take for granted today in data manipulation ergonomics is incremental improvements brought about by REAL experts in ergonomics who observe REAL people using REAL computers.
Describing SCART as a bad idea is very unfair. It's true you couldn't tell which signals were being monitored (unless a sophisticated TV would tell you), but consider this : thanks to SCART compliance, all European TVs on from the early-to-mid 80s were component RGB monitors. This was great for the consoles and home computers of the time. In the US at the same time, TVs only had RF inputs, and only later on the mediocre composite and S-video inputs, and only in the late 90s - early 2000s, and on higher end TVs saw component input generalized. And then not RGB component, rather that inferior differential component. So SCART has forced european TVs a twenty years headstart on the quality of analog input and changed the experience of everyone with a TV-based home computer in the 80s.
Also it was bi-directionnal : a composite signal could travel from the TV to the peripheral and be simultaneously fed back from the peripheral to the TV. This allowed over-the-air pay-TV with a de-scrambler box that was simply plugged in on one of the SCARTs.
That's why it would be good if many people would re-implement these libs directly from the specs. That would weed out incompatible files. That's one of the points of open architectures that content specifications not be tied to implementation.
I don't understand this obsession in software development, of always considering that if a piece of code exists somewhere that does what you want to do, and somehow you have the right to use it, then you must use it. Bitmap image libraries do not represent any expertise or rocket science that you won't find in a freshman textbook and that anybody with a bit of time on their hands cannot re-implement. Yes it's a pain in the ass and in many cases some people like hobbyist programmers who are trying to put together the ultimate "linux desktop" and such are happy to find them ready-to-use as free software. But Google ? Come on, they're supposed to have all these brilliant minds around. Everybody knows that PNG and JPEG libraries are major vulnerabilities. Pulling one off the shelf just to hack something together and show it in trade shows ? That is so lame. Come on Google and everybody else : invest a little bit and do "re-invent the wheel", just for the chance to do it alittle bit better this time.
There is a market for IP, and when there is a market, there is a way to assess a market price. Granted, it's a very complex and relatively not liquid market but so is the real estate market, yet every homeowner has her property assessed by the county. Financial engineering has developped very complex tools for valuating things like stock options granted to employees and such. That would not be that different.
The formal specification for, say, liveness of an interlocking system is a one-liner in a typical temporal logic notation, and you can apply it without significant modification to any number of different implementations, of any number of different applications, whatever their complexity. This is leverage : you put your trust in a very short piece of "code" (the formal spec for your property) and in the tool itself (which is the same kind of trust you put in your compiler), and in return you get trust on a huge complicated piece of software that you wrote. Then you break down your testing into many, independent property checks that all validate one aspect of one big piece of inter-mangled software. That's hugely powerful.
Slashdot Mirror
There are no two things more opposite to each other than OOP and functional programming. Functional programming is about FUNCTIONS, which are those that return a value that is completely determined by their arguments, not depending on, or modifying, any state. OOP is all about calling METHODS, which have values that are determined by both the arguments and the state of the object they're called on, and often are meant to modify that state.
"A Discipline of Programming" is more specifically, about how to write programs that are correct by construction, than it is about how to prove that already written programs are correct.
IMHO what I understand of 30 years of history of the software industry, and my experiences of formal methods and programming language theory in the research world on the one hand, and real-world professional software development on the other hand, validate everything that's in this book.
Systematically buying everything you read new and hardcover is pretty uncommon anyways, except maybe for people who have amounts of disposable income that make the convenience factor the only that counts anyway.
Reading paperback, pocket-book and other discounted editions, buying used books on peer-to-peer marketplaces, borrowing from a library, from a friend, your spouse, of from your parent's bookshelf, all these are not taken into account. If we divide that value of each book that I've read in my life, by the total number of people who've read that particular physical book too, and take the average, we're most definitely in the sub-$1 area, even counting all of the technical and textbook stuff.
I'd agree there's definitely the potential for a "$99-style" psychological effect of labeling a software product "1.0". However instead of outright lying about how many versions of this product have actually already been field-tested, I think it is more elegant to completely remove the version number from the product information and marketing, instead just leaving at most a build number in the Help/About dialog.
Basically yes, but we're not interested in contradictory systems anyway, so that leaves us with systems that are incomplete.
In short ad is somewhat condescending
Yeah, condescending. That's not something that an Apple ad or an Apple fanboy would ever be, is it ?
Maybe there is proportionally more PROFESSIONAL photographers that use Macs for image processing, but Photosynth is clearly a crowdsourcing application and its target audience is everybody, since now everybody has a digital camera and everybody processes their amateur images on their home computer. So we're back at the general PC/Mac split in home computers.
Actually Intel has one big reason to wish for as many x86 vendors as possible : in case of a collapse of AMD, it would take as many other competitors as possible to keep Intel out of trouble from the anti-trust authorities.
Word uses essentially the same algorithm to manage paragraph flow as TeX does, just more sophisticated in the areas of inserting non-text objects such as images and much more sophisticated for mixing languages of different scripts.
The reason TeX output looks "better" to some is mostly that it looks different. TeX was created exclusively for dealing with European languages, and it was created for creating printed books, with only the typographical conventions of printed books, before desktop publishing really existed. By default it looks different and "more bookish", kind of quaint, people interpret that as being "better".
Don't get me wrong I love TeX and its legacy is gigantic (again, Knuth's algorithms were reused in modern word processor). It's just not the best tool anymore.
XML is good because it promotes keeping orthogonal, two very different problems : structuring data and encoding it. Tying these two problems creates complications down the road, as entangling independent areas of information processing and algorithm sciences usually does.
Frankly, the issue is a bit old. Who cares ? These keys do a few useful things under Windows, and I, for one, have grown used to count on them for shortcuts that wouldn't exist without some third party, custom configured hotkey app.
RISC vs CISC was the architecture flamewar of the late 1980s. Welcome to the 21th century, you'll like it here. It's a world when, since the late 90s, the ISA (instruction set architecture), is so abstracted away from the actual micro-architecture of microprocessor, as to make it completely pointless to make distinctions between the two. Modern processors are RISC, they are CISC, they are vector machines, they're everything you want them to be. Move on, the modern problems are now in multi-core architecture and their issues of memory coherence, cache sharing, memory bandwidth, interlocking mechanisms, uniform vs non-uniform, etc. The "pure RISC" standard bearers of yore have disappeared or have been expelled from the personnal computing sphere (remember Apple ditching PowerPC ? Alpha anyone ? Where are those shiny MIPS-based SGIs gone?). Even Intel couldn't impose a new ISA on its own (poor adoption of IA-64). The only RISC ISA that has any existence in the personnal computing arena, including mobile, is ARM, but precisely, they do only mobile. There's really no reason at all to build any device on which you plan to run generic OSes and rich computing experience on anything else than x86 or x86-64 machines.
I do believe that one of the few major adjustments that humanity will take under constraint in our particular lifetime is to shift from fishing in the ocean, to fish farming. It's nothing really, mankind has already done that with great success with most of its other food supplies : hunting was replaced by cattle and piking fruit and berries in the forrest was replaced by crops. That might imply some types of fish to get more rare, just as we haven't eaten nearly as much boar as we used to since we started breeding pigs.
That's ridiculous. Fossil fuels are not at all free : they need to be mined and refined. And I don't see how "more expensive" energies such as wind or solar power would be.
I resent people who are stating that 5, or 6, or 7 billions is too many and that the growth of world population should make us worry. I would like to point out that, compared to the era when world population was less than 1 billion, the average life expectancy, quality of life and, yes, access to ressources and opportunities has dramatically increased for our species. How far is the time when a single pandemic, natural disaster or mass migration would wipe out a third of a continent population and make whole civilization disappear from History ? Notwithstanding the current price fluctuations that call for natural adjustments in production and distribution systems, REAL hunger, the one where the basic intake of food necessary for survival simply isn't available within reach, has been reduced to cases relatively limited in scope and mostly due to geopolitical circumstances rather than natural resource limitations.
The video game market is one of the most expensive and toughest to crack into of all global markets. Only two new companies managed to make it from scratch in more than 10 years : Sony and Microsoft, each of them gambling huge amounts of money over many years. Apple certainly "could" theoretically make it, it has the talent and the cash, but as a business decision it would not make sense for a company that is mostly known for breaking changes and creating whole new markets. As for the "mobile" focus, doesn't make any difference : that field is crowded already, by Nintendo and Sony no less.
Avoiding jamming on mechanical typewriters is the same as having as much alternating between left hand and right hand as possible for successive letters, which makes typing faster and more comfortable. The story that qwerty slows typing down is one lame urban myth. One case of perfect match between engineering practicality and ergonomics, which is why it was so succesfull.
I didn't mean that the MIT didn't have brilliant academics and important contributions to their fields with scientific discoveries and seminal publications.
I was just pointing out that the Media Lab, which has hardly anything to do with all this, has this tradition of hyping klunky prototypes of improbable gadgets and making broad statements about what the future will be. I remember how for they had announced for years in the late 90s, the advent of computerized doorknobs with an IP address.
I notice also that your examples of contributions to computer technology are a bit weak. The Lego Mindstorms is a toy with a mixed success and not a noticeable contribution to the everyday use of robotics. Multics was a failure in itself and in the underlying concept (a centralized architecture providing commoditized computing power). X Window is a complete failure in its underlying concept of centralized applications relying on display/input "servers" with a very fine grained interaction protocol. The concept of centralized apps with remote display/input is only now being realized under a radically different architecure (browser/ajax/flash/silverlight).
Stallman was never capable of delivering the OS he had planned, something that a 20-year old undergrad student finally did in a dorm room in Finland.
SICP is a museum piece that people outside of MIT have read mostly out of curiosity and snobbery. I am myself of a formal semantics / FP background and a huge fan of Scheme but I cannot say that the peculiarity of teaching undergrad programming in scheme is a raging trend accross the campuses in the nation.
People at MIT are notoriously good at creating buzz around the concepts, demos, protypes and inventions that they come up with, especially at the Media Lab. Unfortunately, like everything that happens in academia, the signal to noise ratio is what it is and most of it has no future, sometimes for blatant reasons that one doesn't need to be a very sharp V.C. to figure out. Unfortunately that creates the impression that they are really a bunch of clowns that come up with useless stuff on a regular basis.
So much time and energy wasted trying to come up with fancy data processing metaphors that refer to "intuitive" concepts and situations... Like if educated adults were retarded children and needed to be spoonfed some special way of handling their environment so that they don't find it too challenging. This is mostly useless, the signal/noise in this kind of "breakthrough" research is historically one order of magnitude lower that what valuable R&D dollars are supposed to be spent for.
99.99% of what we take for granted today in data manipulation ergonomics is incremental improvements brought about by REAL experts in ergonomics who observe REAL people using REAL computers.
Describing SCART as a bad idea is very unfair. It's true you couldn't tell which signals were being monitored (unless a sophisticated TV would tell you), but consider this : thanks to SCART compliance, all European TVs on from the early-to-mid 80s were component RGB monitors. This was great for the consoles and home computers of the time. In the US at the same time, TVs only had RF inputs, and only later on the mediocre composite and S-video inputs, and only in the late 90s - early 2000s, and on higher end TVs saw component input generalized. And then not RGB component, rather that inferior differential component. So SCART has forced european TVs a twenty years headstart on the quality of analog input and changed the experience of everyone with a TV-based home computer in the 80s.
Also it was bi-directionnal : a composite signal could travel from the TV to the peripheral and be simultaneously fed back from the peripheral to the TV. This allowed over-the-air pay-TV with a de-scrambler box that was simply plugged in on one of the SCARTs.
That's why it would be good if many people would re-implement these libs directly from the specs. That would weed out incompatible files. That's one of the points of open architectures that content specifications not be tied to implementation.
I don't understand this obsession in software development, of always considering that if a piece of code exists somewhere that does what you want to do, and somehow you have the right to use it, then you must use it. Bitmap image libraries do not represent any expertise or rocket science that you won't find in a freshman textbook and that anybody with a bit of time on their hands cannot re-implement. Yes it's a pain in the ass and in many cases some people like hobbyist programmers who are trying to put together the ultimate "linux desktop" and such are happy to find them ready-to-use as free software. But Google ? Come on, they're supposed to have all these brilliant minds around. Everybody knows that PNG and JPEG libraries are major vulnerabilities. Pulling one off the shelf just to hack something together and show it in trade shows ? That is so lame. Come on Google and everybody else : invest a little bit and do "re-invent the wheel", just for the chance to do it alittle bit better this time.
There is a market for IP, and when there is a market, there is a way to assess a market price. Granted, it's a very complex and relatively not liquid market but so is the real estate market, yet every homeowner has her property assessed by the county. Financial engineering has developped very complex tools for valuating things like stock options granted to employees and such. That would not be that different.
The formal specification for, say, liveness of an interlocking system is a one-liner in a typical temporal logic notation, and you can apply it without significant modification to any number of different implementations, of any number of different applications, whatever their complexity. This is leverage : you put your trust in a very short piece of "code" (the formal spec for your property) and in the tool itself (which is the same kind of trust you put in your compiler), and in return you get trust on a huge complicated piece of software that you wrote. Then you break down your testing into many, independent property checks that all validate one aspect of one big piece of inter-mangled software. That's hugely powerful.
I hope your prof failed you.