Now msword is struggling to compete in the changed ecosystem, and has avoided extinction only by virtue of the vicious protectiveness of its mommy.
At this point, new word processor adaptations are more likely to succeed if they provide an interface familiar to the potential users; that doesn't mean they're the same in all ways.
I disagree with your basic premise. Word hasn't avoided extinction because of its mommy's protectiveness. It has avoided extinction because it is the best word processor available today, bar none. AFAICS, those new word processors are all trying to be the same in all ways, but of course they'll always be one step behind because they're letting Microsoft define the game. Can you name a single serious feature that something like OO Writer or AbiWord has that Word doesn't do at least as well? (Please don't anyone shout the "PDF export" mantra. Leaving aside recent developments in Microsoft land, the PDF export in OO Writer is so crippled as to be a liability anyway.)
How many people outside of fields like engineering and other math-specialty careers even need to be able to do much beyond the basic four functions anyway?
Well, let's see...
Here in the UK, much of our economy is in serious danger of tanking because of record consumer debts. At the same time, we have people stretching to get the biggest mortgage they can today, with no concept whatsoever of the impact the expected 0.25% rise in interest rates next month will have on their repayments. We have people making the minimum payment on their credit card bill each month, even when they could readily afford more, because they want to put something in their savings account. And of course, the majority of adults have their money in bank current accounts that pay interest at 0.1% pa when the rival bank next door offers 2.5% and typical savings accounts offer 5–6%.
At this point, I could start on the examples of how much time and material countless tradesmen waste because of an inability to do basic trigonometry, but I think the point is already made.
The catch is that even people who could be quite good at maths are being directed elsewhere, leaving a shortage of mathematical knowledge when they get to higher education and could really use it.
In the UK, the standards-based approach has been bad for education. This is the view of people I know involved in staff recruitment (I work in software dev). It is also the view of people I know involved in the university scene (I live in Cambridge, UK, and many of my friends are staff or postgrad researchers at the university). And it is most certainly the view of people I know involved in teaching at school age (those that haven't simply left the profession in disgust, that is).
The argument about standards-based testing would have merit if the approach worked in practice, but unfortunately, we can clearly see now that school league tables have not had the desired effect. Instead of motivating schools to teach to higher standards, what they have actually done is motivate schools to play the system.
Today, schools will encourage weaker students to take subjects where they are likely to get better grades rather than more difficult subjects, as with mathematics in the case of TFA. Similar things hold for sciences, modern languages, etc. This is caused, in no small part, by giving all subjects equal weight in the statistics (give or take special statistics for things like English and maths, which they play around with every couple of years).
Today, schools will focus on teaching pupils to pass their exams with as high a grade as possible, not on teaching pupils their subject and letting exams simply be a measure of how well the pupils have learned. Revision is all about exam strategy now.
Today, schools will actively discourage pupils from taking courses where they may pass but without gaining a high grade. No grade at all damages the averages less than a D or E grade, and so doesn't corrupt the school's precious "percentage of examinations taken that were passed at grades A*-C" type statistics.
The bottom line is that instead of teaching pupils real understanding in key subjects, and playing a role in their personal and social development along the way, today's schools are simply machines geared to generating exam passes, and today's pupils are simply fuel for the machine. Consequently, you can get straight-A students who don't know their subjects. You get universities inventing their own entrance examinations and/or stating bluntly that they will ignore certain A-level subjects entirely when considering applications, simply because otherwise everyone applying is a straight-A student and the admissions tutors can't distinguish between them. And you get people applying for jobs with great qualifications on paper, who can't do now with an A-level in a subject what someone twenty years older could do after gaining an O-level.
This isn't education, it's product marketing for the New Labour administration. And like much of marketing, most of it is simply lying with statistics, and finding excuses to deny a reality that is self-evident to any qualified observer who takes the time to look.
And as for firing teachers, consider this: so many old-school, teach-the-subject veterans are now leaving the profession (often through early retirement deals because they are much more expensive to employ as teachers than green youngsters fresh from university) that all the accumulated wisdom of generations of teachers is rapidly disappearing. We are being left only with youngsters who have found trendy new methods like synthetic phonics to increase results (no, wait, that one's decades old!) and think they're very clever. Unfortunately, the ones who are very clever rapidly get disillusioned and leave the profession, as several highly qualified and very smart friends who graduated in my university generation all did within two years of starting work as teachers. You don't have to fire anyone in this scheme, because the good people — young and old alike — have already left in disgust.
Because tagging UI is more complicated than folders, particularly if you usually organise with subfolders and need to emulate the relationship using multiple tags? Simple as that, I suspect...
I suspect the point was that nothing you've listed intrinsically requires e-mail. It just happens to be a convenient conduit for sending specially formatted messages representing meeting invitations and accept/decline responses, and if you're going to use e-mail to send/receive such messages, then obviously your calendaring UI has to be integrated with your e-mail system somewhere along the line.
In a newsgroup or forum context, top quoting is more appropriate, as one is posting a reply for all to see, and the "conversation" shouldn't turn into a long, tit-for-tat dialog. [...] In one-on-one email exchanges, the opposite is true. One doesn't need to review the whole conversation, but only the last reply. Bottom quoting fits in with this - put the most relevant info latest, but keep the rest of the stuff for a conversation record. Trimming is only important to get rid of the fluff put on by overly long sigs and advertisements, but otherwise, keeping the text of the conversation in one place is better than having to find the particular email where a certain portion of the conversation took place.
The only use for bottom-quoting I've ever seen is when forwarding an entire e-mail thread to someone who hasn't been involved in the discussion so far. Decent tools for tracking the conversation over several mails and building a digest would be a far superior way to do this.
Other than that, IMHO top-posting is a serious PITA even in one-on-one conversations. I'm dealing with several groups containing people elsewhere in the company right now, and I have double figures of e-mail discussions on the go on any given day. I hate it when people just send me something that goes straight into text, because the first thing I have to do is scroll down to the previous comments to remind myself what was said before. Then I have to try to work out which part of often quite long previous mails each comment refers to. None of this would be necessary if people just quoted properly. And don't even get me started on Outlook-style quoting in HTML e-mails, which makes it almost impossible to quote properly in reply even if you want to...
Visual Studio is far from perfect, but for a tool that I used day in and day out, I don't have many complaints overall. How many programs can you really say that about?
Hang on...
Windows XP
Word
Excel
LaTeX
Firefox
Visual C++ (though it's actually gone backwards in many ways since all the.Net stuff was introduced)
Yep, that's about it. Interesting that without any prior intent, I've come up with four big name MS products I use routinely, two big name products that are freely available, and no commercial software from vendors other than Microsoft.
The list of big name software I use regularly but wouldn't classify as "far from perfect, but not many complaints overall" is also quite intriguing, but I'll omit it here to avoid the flame-war. Suffice it to say that several big name products in OSS world and several very well known commercial vendors would feature prominently.
The irony of your post is that your biology analogy doesn't really work.
I'm surprised that in the hundreds of posts to this thread, no-one yet seems to have noticed that one good possibility is worth more than a choice among many poor options.
MS Office is my usual example. While some OSS has been very successful, there is more creativity in the search bar in Firefox than in the entirety of OpenOffice Writer. It is unashamedly written as a Word rip-off. So are most of the other big-name OSS word processors. The only real adaptation most of these products do is to run on a non-Windows platform. They aren't creative at all, for the most part at least. Resistance to disease? Like what, the macro viruses everyone complains about yet strangely I never seem to encounter?
If you're fond of analogies, try a few different ones to see what fits. You might like to consider the old redundancy for safety vs. wasted effort conundrum as a starting point.
That is unfortunate, because people like you and posts like the one you just wrote are exactly the reason Linux is not catching on as fast as might otherwise be expected. You are the modern equivalent of the "RTFM, n00b!" morons on IRC. You are marginally more polite, but your immediate assumption is that the GP post is trolling, and not simply the real life experience of a somewhat technical but non-expert user who is considering moving from the tried-and-tested ground of Windows to the brave new world of Linux, and who isn't going to bow down and worship at the altar of Ubuntu before he's seen his miracles.
Instead of providing constructive advice about Python, for example, you mock the poster for his lack of knowledge, and make a flippant reference to Google. Of course people like the GP poster know how to search using Google. The point is that on familiar ground (Windows in this case), he doesn't have to. Thus expecting him to search on Google every five minutes for the first several weeks of setting up his new system is unrealistic. He won't waste his time doing that for long, he will simply switch back to using Windows, where the stuff he wants to do can be done easily in a way he already understands.
This is the challenge facing the Linux community, and smugly, arrogantly, pretentiously denying it isn't going to convince anyone about the merits of Linux as a platform — quite the opposite, most likely.
Good code now days has more to do with structure and maintainability as opposed to squeezing every possisble extra CPU out of a procedure.
You write as if the two are mutually exclusive. IME, while this can be true, it rarely is in practice. The number of people who selectively quote part of Hoare's warning about premature optimisation, and use this as an excuse not to consider efficiency in their programming, is a major contributory factor in the amount of sucky code in the world today.
Instead of wasting time learning C/assembler, which you probably wouldn't get any use out of, I would go with an OO language such as Java or.net. Stick to the high level languages, no matter what the code programming pursists say.
Please yourself. In five years' time, when the rest of us are writing concurrent programs in languages that actually support concurrency seriously without introducing race conditions and deadlocks every other function, we "purists" will try not to laugh at you too much for going with popularity rather than understanding and choosing the right tools for the job.:-)
I disagree. If a government took away the rights (ethical, legal, or otherwise) of individual citizens in this way, then you would be moving towards a more socialist political framework. However, governments should not be required to accord the same respect to artificial entities like corporations that they should towards real people. Businesses are commercial entities, and should be permitted by law to exist and to act exactly to the extent that they serve the people by doing so. It is a very important role of government, particularly in basically capitalist societies, to provide the checks and balances that keep businesses doing so. The alternative, where capitalism is allowed to run its course unchecked, is that every major industry ultimately converges on an effective monopoly, which lacking effective competition or regulation then exploits the people forever (cf. the current US government).
High level languages vs. assembly
on
Beginning Ruby
·
· Score: 1
The rules of the game have changed in two important ways over the intervening years, though:
Writing efficient assembly language by hand got harder. With the move to the mainstream of pipelining and caching technologies, you can't just write assembly instructions in any old order to get the fastest results any more. And of course, we have many extra instructions available on modern chips: the Intel/AMD family alone now have something like seven or eight supplementary instruction sets for specialist numerical applications, on top of the basic IA-32 stuff that remains the lingua franca of PC assembly coding. Manually generating optimal assembly code is therefore much more difficult than it used to be, and very much a specialist activity. (Those specialists are typically employed writing optimising compilers.)
Optimiser technology got better. Up until a few years ago, even the most powerful optimisations in separate compilation languages like C and C++ were pretty much local to a particular file. Today, improved compiler/linker technology allows for global optimisations. Optimisation based on real world profiling data is possible. And of course, the basic underlying optimisations are more varied and more powerful; note the adoption of SSA within GCC, for example.
This isn't to say that hand-crafted assembly language no longer has a place. C and C++ are pretty low-level languages, and don't convey a lot of semantic information that the programmer has. Thus the compiler and optimiser are constrained in the assumptions they can safely make, with aliasing being the classic example problem. Even when they do "understand", their generic code generators may not give 100% optimal assembly in every single case. I work in high performance numerics, and I've seen my share of little functions that surprisingly weren't inlined (until, knowing that they aren't and thinking about it more deeply, you realise what assumption couldn't be made by the compiler and it's not surprising any more). I've certainly seen needlessly slow assembly language emitted in critical functions.
But the rules have changed, and the market for lovingly hand-crafted assembly is much smaller than it used to be. Software development is, in practice, all about "good enough". With today's rules, compilers are good enough a lot more often than they used to be. In contrast, writing better enough assembly language by hand is an increasingly difficult job, and the time spent doing so might well be better spend optimising some algorithm or data structure at a higher level, or filing a performance bug with your compiler vendor so they can fix their assembly language generation more generally.
Having suffered for years with endless rants and moans about how lame XP is, suddenly people are beginning to show it love.
Unfortunately for Microsoft, people are also beginning to show a bit more love to Apple and Linux boxes.
I have noticed a distinct shift in recent months, with more friends asking me — as resident geek — about Apple stuff. I don't personally know much about Apple as a platform, so I direct them elsewhere for advice. I haven't yet worked out the catalyst for this, but I doubt it was Vista; the trend started earlier. Maybe the relatively high-profile "switch" campaign worked to some extent, or increased visibility because of iPods and the like played a role. In any case, Vista seems to be doing little to win back the hearts and minds Microsoft have been losing to Apple.
It pretty much goes without saying that Vista with all its DRM is a big turn-off for geeks, and also that Linux has been improving steadily as a desktop platform. Again, I'm not sufficiently expert to know for sure, but I get the feeling that the traditional hang-ups about having to hack text files and poor hardware support are becoming less relevant as the major distros up their game. The time of "Linux on the desktop" may not quite have arrived yet, but with the traditional Windows dominance in the games market becoming less relevant when 80% of that market is based on consoles anyway, I doubt it will be many more years before Linux becomes a serious contender for the desktop space of home users.
Anyway, if my experience is typical (obviously I can't be sure) then Microsoft are in for a rough ride over the next couple of years. I doubt this is because Vista is seen as a backward step by much of the user base, though of course it has some big problems that are starting to attract attention. But I think right now, it's simply that the alternatives are becoming better, and Vista isn't seen as a forward step either.
Just because I claim one thing is far too much like hard work, please don't think I'm implying the other isn't.:-)
It's definitely an advantage to have some sort of worthwhile programming/macro facilities when you're building a document template, but it's also useful to be able to override such things and tweak manually when the situation warrants it. The problem today is that with the TeX family, the programming is there (albeit using obscure black magic known only to about three witch doctors in the world) but overriding it when it doesn't give a good result is a royal PITA. The typical DTP approach is at the opposite extreme: you can position everything by hand if you want to... and if you don't want to, that's tough, because you're probably going to have to anyway.
I do a fair amount of document production, formal technical things at work, publicity for local clubs, that sort of thing, but nothing really on the scale you're talking about as a professional. I know I've found it frustrating to be torn between these two ways of getting results — made worse by the fact that once you've committed to one, it's very hard to change to the other without starting from scratch — so I certainly sympathise with those of you who have to deal with this on an "industrial" scale!
How do you know that the Quicktime DLLs aren't escalating privileges to run commands in supervisor mode?
I don't. But any operating system that lets a simple user application like movie playing escalate privileges far enough to take out the entire system is broken.
In theory, yes it is. Unfortunately, in practice it doesn't work, at least in the LaTeX family. I've seen plenty of cases where a page break being emitted at an unfortunate point resulted in the first text at the top of the following page incorrectly picking up a colour used somewhere else in the vicinity, for example. The unpredictable nature of page breaks in the LaTeX model seems to make this unavoidable; the 'net is full of people with the same problem, but I never found a reliable solution.
Placed graphics usually ``just work'' and of course, one can handle things programmatically
Sure, but again they only “just work” if the behaviour you want fits into the very limited categories offered by LaTeX. I have lost track of the number of friends I've helped to hack a LaTeX source file for a paper or thesis because it was doing outright stupid things with float placement. If you're intimately familiar with all the rules used to determine here-placement and the various tolerances you can configure for the whitespace areas around a float with the different placement options, you might be able to get good results first time most of the time, but it's far too much like hard work.
You're right that programmatic handling is highly desirable for a lot of things when you are, effectively, building a document template into which you can flow content. This possibility is obviously one of the big strengths of the TeX-based systems. But it should always be easy to override those rules when the results of following them dogmatically are undesirable, and this is difficult to almost impossible for a typical LaTeX project. And speaking as a professional software developer, the prospect of trying to write new logic using the TeX programming primitives is enough to dampen my will to live...
Fair enough. I'd guess your requirements are beyond any of the typical DTP packages today, and I know there are some pretty major limitations on what InDesign can do yet in this area so far. Personally, the idea of trying to produce that sort of document using (La)TeX fills me with horror, but maybe your particular requirements don't hit any of the areas where the TeX family are completely screwed up (use of any sort of colour or floating figures, for example).
I suspect we can all agree that no-one has yet produced typesetting/DTP software with the kind of power and flexibility we'd like, but also the ease of use and efficiency.:-(
...except DRM measures don't do squat to the market.
But that doesn't follow, because the entire black market is not bootleggers and geeks. You're ignoring the vast numbers of people who simply borrow media from a friend to rip it, but don't really have a clue what they're doing. You're also ignoring the unfortunate reality that while paying customers may be annoyed by this stuff, right now most of them don't care enough to stop paying.
If you get a complete system crash from something like this, it is always the operating system's fault. You can shout and whinge about kernel mode drivers and catching access violations and real time this, that or the other. However, the bottom line is that the kernel is the heart of any operating system. Its only purpose is to construct a framework within which other software can run, which provides some basic guarantees about performance, resource management, security, etc. There is little, if any, reason to compromise the integrity of the kernel and allow any other code to run at that privilege level in a desktop OS like Vista. It certainly isn't necessary to play back some video file in a user application.
Since the only excuse for breaking backward compatibility with just about the whole universe, which on reports to date pretty much sounds like what Vista is doing, is because they're putting a serious security model in place, this sort of thing simply shouldn't be possible. If it is, then by definition their new, serious security model is fundamentally broken.
Owning patents yourself in hopes you can countersue is pointless; it doesn't work.
That's not entirely fair. It appears to work very well as a way for large companies to push small companies out of a market. (Not that this is something to be welcomed, of course, given that it defeats the whole point of having patents...) But you're right, patent-hoarding isn't a terribly effective defence against patent trolls who exist purely to litigate.
We would probably need some sort of "bad faith" mechanism to nullify a patent to beat the non-product businesses. I suspect "We're actively developing something based on it" would have to be an absolute defence to that mechanism, to avoid it being used to undermine legitimate patents or create yet more litigate-until-they're-dead lawsuits.
So how far reaching is this rediculous patent system. The USA is a write-off i know. Things are similar in europe yes?
No. The European legislature has so far rejected attempts to force the EU member states to implement software patents.
This is not to say that no patents have ever been granted on software in EU states. Indeed, one of the few good things about the various failed proposals was that they would have regulated an area that is somewhat messy in some countries at present. But those patents may or may not stand up — I don't think have been any relevant test cases yet — and they are relatively few. We don't have one-click and the like here.
Slashdot Mirror
I disagree with your basic premise. Word hasn't avoided extinction because of its mommy's protectiveness. It has avoided extinction because it is the best word processor available today, bar none. AFAICS, those new word processors are all trying to be the same in all ways, but of course they'll always be one step behind because they're letting Microsoft define the game. Can you name a single serious feature that something like OO Writer or AbiWord has that Word doesn't do at least as well? (Please don't anyone shout the "PDF export" mantra. Leaving aside recent developments in Microsoft land, the PDF export in OO Writer is so crippled as to be a liability anyway.)
Well, let's see...
Here in the UK, much of our economy is in serious danger of tanking because of record consumer debts. At the same time, we have people stretching to get the biggest mortgage they can today, with no concept whatsoever of the impact the expected 0.25% rise in interest rates next month will have on their repayments. We have people making the minimum payment on their credit card bill each month, even when they could readily afford more, because they want to put something in their savings account. And of course, the majority of adults have their money in bank current accounts that pay interest at 0.1% pa when the rival bank next door offers 2.5% and typical savings accounts offer 5–6%.
At this point, I could start on the examples of how much time and material countless tradesmen waste because of an inability to do basic trigonometry, but I think the point is already made.
The catch is that even people who could be quite good at maths are being directed elsewhere, leaving a shortage of mathematical knowledge when they get to higher education and could really use it.
In the UK, the standards-based approach has been bad for education. This is the view of people I know involved in staff recruitment (I work in software dev). It is also the view of people I know involved in the university scene (I live in Cambridge, UK, and many of my friends are staff or postgrad researchers at the university). And it is most certainly the view of people I know involved in teaching at school age (those that haven't simply left the profession in disgust, that is).
The argument about standards-based testing would have merit if the approach worked in practice, but unfortunately, we can clearly see now that school league tables have not had the desired effect. Instead of motivating schools to teach to higher standards, what they have actually done is motivate schools to play the system.
Today, schools will encourage weaker students to take subjects where they are likely to get better grades rather than more difficult subjects, as with mathematics in the case of TFA. Similar things hold for sciences, modern languages, etc. This is caused, in no small part, by giving all subjects equal weight in the statistics (give or take special statistics for things like English and maths, which they play around with every couple of years).
Today, schools will focus on teaching pupils to pass their exams with as high a grade as possible, not on teaching pupils their subject and letting exams simply be a measure of how well the pupils have learned. Revision is all about exam strategy now.
Today, schools will actively discourage pupils from taking courses where they may pass but without gaining a high grade. No grade at all damages the averages less than a D or E grade, and so doesn't corrupt the school's precious "percentage of examinations taken that were passed at grades A*-C" type statistics.
The bottom line is that instead of teaching pupils real understanding in key subjects, and playing a role in their personal and social development along the way, today's schools are simply machines geared to generating exam passes, and today's pupils are simply fuel for the machine. Consequently, you can get straight-A students who don't know their subjects. You get universities inventing their own entrance examinations and/or stating bluntly that they will ignore certain A-level subjects entirely when considering applications, simply because otherwise everyone applying is a straight-A student and the admissions tutors can't distinguish between them. And you get people applying for jobs with great qualifications on paper, who can't do now with an A-level in a subject what someone twenty years older could do after gaining an O-level.
This isn't education, it's product marketing for the New Labour administration. And like much of marketing, most of it is simply lying with statistics, and finding excuses to deny a reality that is self-evident to any qualified observer who takes the time to look.
And as for firing teachers, consider this: so many old-school, teach-the-subject veterans are now leaving the profession (often through early retirement deals because they are much more expensive to employ as teachers than green youngsters fresh from university) that all the accumulated wisdom of generations of teachers is rapidly disappearing. We are being left only with youngsters who have found trendy new methods like synthetic phonics to increase results (no, wait, that one's decades old!) and think they're very clever. Unfortunately, the ones who are very clever rapidly get disillusioned and leave the profession, as several highly qualified and very smart friends who graduated in my university generation all did within two years of starting work as teachers. You don't have to fire anyone in this scheme, because the good people — young and old alike — have already left in disgust.
Because tagging UI is more complicated than folders, particularly if you usually organise with subfolders and need to emulate the relationship using multiple tags? Simple as that, I suspect...
I suspect the point was that nothing you've listed intrinsically requires e-mail. It just happens to be a convenient conduit for sending specially formatted messages representing meeting invitations and accept/decline responses, and if you're going to use e-mail to send/receive such messages, then obviously your calendaring UI has to be integrated with your e-mail system somewhere along the line.
The only use for bottom-quoting I've ever seen is when forwarding an entire e-mail thread to someone who hasn't been involved in the discussion so far. Decent tools for tracking the conversation over several mails and building a digest would be a far superior way to do this.
Other than that, IMHO top-posting is a serious PITA even in one-on-one conversations. I'm dealing with several groups containing people elsewhere in the company right now, and I have double figures of e-mail discussions on the go on any given day. I hate it when people just send me something that goes straight into text, because the first thing I have to do is scroll down to the previous comments to remind myself what was said before. Then I have to try to work out which part of often quite long previous mails each comment refers to. None of this would be necessary if people just quoted properly. And don't even get me started on Outlook-style quoting in HTML e-mails, which makes it almost impossible to quote properly in reply even if you want to...
Hang on...
Yep, that's about it. Interesting that without any prior intent, I've come up with four big name MS products I use routinely, two big name products that are freely available, and no commercial software from vendors other than Microsoft.
The list of big name software I use regularly but wouldn't classify as "far from perfect, but not many complaints overall" is also quite intriguing, but I'll omit it here to avoid the flame-war. Suffice it to say that several big name products in OSS world and several very well known commercial vendors would feature prominently.
The irony of your post is that your biology analogy doesn't really work.
I'm surprised that in the hundreds of posts to this thread, no-one yet seems to have noticed that one good possibility is worth more than a choice among many poor options.
MS Office is my usual example. While some OSS has been very successful, there is more creativity in the search bar in Firefox than in the entirety of OpenOffice Writer. It is unashamedly written as a Word rip-off. So are most of the other big-name OSS word processors. The only real adaptation most of these products do is to run on a non-Windows platform. They aren't creative at all, for the most part at least. Resistance to disease? Like what, the macro viruses everyone complains about yet strangely I never seem to encounter?
If you're fond of analogies, try a few different ones to see what fits. You might like to consider the old redundancy for safety vs. wasted effort conundrum as a starting point.
It's not just you.
That is unfortunate, because people like you and posts like the one you just wrote are exactly the reason Linux is not catching on as fast as might otherwise be expected. You are the modern equivalent of the "RTFM, n00b!" morons on IRC. You are marginally more polite, but your immediate assumption is that the GP post is trolling, and not simply the real life experience of a somewhat technical but non-expert user who is considering moving from the tried-and-tested ground of Windows to the brave new world of Linux, and who isn't going to bow down and worship at the altar of Ubuntu before he's seen his miracles.
Instead of providing constructive advice about Python, for example, you mock the poster for his lack of knowledge, and make a flippant reference to Google. Of course people like the GP poster know how to search using Google. The point is that on familiar ground (Windows in this case), he doesn't have to. Thus expecting him to search on Google every five minutes for the first several weeks of setting up his new system is unrealistic. He won't waste his time doing that for long, he will simply switch back to using Windows, where the stuff he wants to do can be done easily in a way he already understands.
This is the challenge facing the Linux community, and smugly, arrogantly, pretentiously denying it isn't going to convince anyone about the merits of Linux as a platform — quite the opposite, most likely.
You write as if the two are mutually exclusive. IME, while this can be true, it rarely is in practice. The number of people who selectively quote part of Hoare's warning about premature optimisation, and use this as an excuse not to consider efficiency in their programming, is a major contributory factor in the amount of sucky code in the world today.
Please yourself. In five years' time, when the rest of us are writing concurrent programs in languages that actually support concurrency seriously without introducing race conditions and deadlocks every other function, we "purists" will try not to laugh at you too much for going with popularity rather than understanding and choosing the right tools for the job. :-)
I disagree. If a government took away the rights (ethical, legal, or otherwise) of individual citizens in this way, then you would be moving towards a more socialist political framework. However, governments should not be required to accord the same respect to artificial entities like corporations that they should towards real people. Businesses are commercial entities, and should be permitted by law to exist and to act exactly to the extent that they serve the people by doing so. It is a very important role of government, particularly in basically capitalist societies, to provide the checks and balances that keep businesses doing so. The alternative, where capitalism is allowed to run its course unchecked, is that every major industry ultimately converges on an effective monopoly, which lacking effective competition or regulation then exploits the people forever (cf. the current US government).
The rules of the game have changed in two important ways over the intervening years, though:
This isn't to say that hand-crafted assembly language no longer has a place. C and C++ are pretty low-level languages, and don't convey a lot of semantic information that the programmer has. Thus the compiler and optimiser are constrained in the assumptions they can safely make, with aliasing being the classic example problem. Even when they do "understand", their generic code generators may not give 100% optimal assembly in every single case. I work in high performance numerics, and I've seen my share of little functions that surprisingly weren't inlined (until, knowing that they aren't and thinking about it more deeply, you realise what assumption couldn't be made by the compiler and it's not surprising any more). I've certainly seen needlessly slow assembly language emitted in critical functions.
But the rules have changed, and the market for lovingly hand-crafted assembly is much smaller than it used to be. Software development is, in practice, all about "good enough". With today's rules, compilers are good enough a lot more often than they used to be. In contrast, writing better enough assembly language by hand is an increasingly difficult job, and the time spent doing so might well be better spend optimising some algorithm or data structure at a higher level, or filing a performance bug with your compiler vendor so they can fix their assembly language generation more generally.
Unfortunately for Microsoft, people are also beginning to show a bit more love to Apple and Linux boxes.
I have noticed a distinct shift in recent months, with more friends asking me — as resident geek — about Apple stuff. I don't personally know much about Apple as a platform, so I direct them elsewhere for advice. I haven't yet worked out the catalyst for this, but I doubt it was Vista; the trend started earlier. Maybe the relatively high-profile "switch" campaign worked to some extent, or increased visibility because of iPods and the like played a role. In any case, Vista seems to be doing little to win back the hearts and minds Microsoft have been losing to Apple.
It pretty much goes without saying that Vista with all its DRM is a big turn-off for geeks, and also that Linux has been improving steadily as a desktop platform. Again, I'm not sufficiently expert to know for sure, but I get the feeling that the traditional hang-ups about having to hack text files and poor hardware support are becoming less relevant as the major distros up their game. The time of "Linux on the desktop" may not quite have arrived yet, but with the traditional Windows dominance in the games market becoming less relevant when 80% of that market is based on consoles anyway, I doubt it will be many more years before Linux becomes a serious contender for the desktop space of home users.
Anyway, if my experience is typical (obviously I can't be sure) then Microsoft are in for a rough ride over the next couple of years. I doubt this is because Vista is seen as a backward step by much of the user base, though of course it has some big problems that are starting to attract attention. But I think right now, it's simply that the alternatives are becoming better, and Vista isn't seen as a forward step either.
Just because I claim one thing is far too much like hard work, please don't think I'm implying the other isn't. :-)
It's definitely an advantage to have some sort of worthwhile programming/macro facilities when you're building a document template, but it's also useful to be able to override such things and tweak manually when the situation warrants it. The problem today is that with the TeX family, the programming is there (albeit using obscure black magic known only to about three witch doctors in the world) but overriding it when it doesn't give a good result is a royal PITA. The typical DTP approach is at the opposite extreme: you can position everything by hand if you want to... and if you don't want to, that's tough, because you're probably going to have to anyway.
I do a fair amount of document production, formal technical things at work, publicity for local clubs, that sort of thing, but nothing really on the scale you're talking about as a professional. I know I've found it frustrating to be torn between these two ways of getting results — made worse by the fact that once you've committed to one, it's very hard to change to the other without starting from scratch — so I certainly sympathise with those of you who have to deal with this on an "industrial" scale!
Which one, Bill or Steve?
I don't. But any operating system that lets a simple user application like movie playing escalate privileges far enough to take out the entire system is broken.
In theory, yes it is. Unfortunately, in practice it doesn't work, at least in the LaTeX family. I've seen plenty of cases where a page break being emitted at an unfortunate point resulted in the first text at the top of the following page incorrectly picking up a colour used somewhere else in the vicinity, for example. The unpredictable nature of page breaks in the LaTeX model seems to make this unavoidable; the 'net is full of people with the same problem, but I never found a reliable solution.
Sure, but again they only “just work” if the behaviour you want fits into the very limited categories offered by LaTeX. I have lost track of the number of friends I've helped to hack a LaTeX source file for a paper or thesis because it was doing outright stupid things with float placement. If you're intimately familiar with all the rules used to determine here-placement and the various tolerances you can configure for the whitespace areas around a float with the different placement options, you might be able to get good results first time most of the time, but it's far too much like hard work.
You're right that programmatic handling is highly desirable for a lot of things when you are, effectively, building a document template into which you can flow content. This possibility is obviously one of the big strengths of the TeX-based systems. But it should always be easy to override those rules when the results of following them dogmatically are undesirable, and this is difficult to almost impossible for a typical LaTeX project. And speaking as a professional software developer, the prospect of trying to write new logic using the TeX programming primitives is enough to dampen my will to live...
Fair enough. I'd guess your requirements are beyond any of the typical DTP packages today, and I know there are some pretty major limitations on what InDesign can do yet in this area so far. Personally, the idea of trying to produce that sort of document using (La)TeX fills me with horror, but maybe your particular requirements don't hit any of the areas where the TeX family are completely screwed up (use of any sort of colour or floating figures, for example).
I suspect we can all agree that no-one has yet produced typesetting/DTP software with the kind of power and flexibility we'd like, but also the ease of use and efficiency. :-(
That's true.
So is that.
And that.
But that doesn't follow, because the entire black market is not bootleggers and geeks. You're ignoring the vast numbers of people who simply borrow media from a friend to rip it, but don't really have a clue what they're doing. You're also ignoring the unfortunate reality that while paying customers may be annoyed by this stuff, right now most of them don't care enough to stop paying.
If you get a complete system crash from something like this, it is always the operating system's fault. You can shout and whinge about kernel mode drivers and catching access violations and real time this, that or the other. However, the bottom line is that the kernel is the heart of any operating system. Its only purpose is to construct a framework within which other software can run, which provides some basic guarantees about performance, resource management, security, etc. There is little, if any, reason to compromise the integrity of the kernel and allow any other code to run at that privilege level in a desktop OS like Vista. It certainly isn't necessary to play back some video file in a user application.
Since the only excuse for breaking backward compatibility with just about the whole universe, which on reports to date pretty much sounds like what Vista is doing, is because they're putting a serious security model in place, this sort of thing simply shouldn't be possible. If it is, then by definition their new, serious security model is fundamentally broken.
That's not entirely fair. It appears to work very well as a way for large companies to push small companies out of a market. (Not that this is something to be welcomed, of course, given that it defeats the whole point of having patents...) But you're right, patent-hoarding isn't a terribly effective defence against patent trolls who exist purely to litigate.
We would probably need some sort of "bad faith" mechanism to nullify a patent to beat the non-product businesses. I suspect "We're actively developing something based on it" would have to be an absolute defence to that mechanism, to avoid it being used to undermine legitimate patents or create yet more litigate-until-they're-dead lawsuits.
The not-as-clever-as-he-liked-to-think inventor was older than 12, you insensitive clod!
Hey, did you always have that red dot on your head? It's like, someone's pointing a laser sight at Slashdot.
Oh...
No. The European legislature has so far rejected attempts to force the EU member states to implement software patents.
This is not to say that no patents have ever been granted on software in EU states. Indeed, one of the few good things about the various failed proposals was that they would have regulated an area that is somewhat messy in some countries at present. But those patents may or may not stand up — I don't think have been any relevant test cases yet — and they are relatively few. We don't have one-click and the like here.