I think the parent poster was talking about reaching back in time to grab a person from the 1920's, bringing them forward 80 years, and getting their reactions to how things have changed.
I know that I would be much more surprised by the apparent gross violation of the laws of physics involved in the actual time travel.
I bet if people from the 1920's see the world today, they would be alarmed by the technology and hitech gadgets (simple automatic doors for that matter) around them. I want to feel like that when I think of the future - not just some old computer capable of working really really fast.
My grandmother is from the 1920's, and if she is alarmed by automatic doors, she has never shown any sign of it. I suspect she doesn't know how motion sensors work, but then, she doesn't care. She is perplexed by the blinking 12:00 on her VCR, but that appears to be common in people from the 1970's as well. Mostly, she is alarmed that you can say "bitch" on TV and that her bursitis is acting up again.
She couldn't care less about computers, as she has no need for them. Her TV provides her with all of her "stories", and when she wants to contact someone in a hurry, she talks into a telephone, which she regards as a better use of energy than learning to type, buying an expensive computer, and paying for an ISP.
As near as I can tell, the only pieces of post-WW2 technology she uses are modern dentistry and insulin. Strangely, she seems more content than I am, so I have a feeling that when I'm her age, my disinterest in the latest gadgets will probably be minimal, too.
What sort of message would that be? Monkeys are assholes, and monkeys with large brains are assholes with nuclear weapons?
The "advanced civilization" that actually has to conduct that experiment would either be idiot-savants with a knack for genetic engineering or else rodeo clowns snatched from the future by alien overlords to act as viceroys over a slave class of biology graduate students.
Wikipedia has replaced Google as my favorite site. It's arguably the one site I would actually pay to access, and I'm so grateful I don't have to.
That being said, I don't like being a leech, but I don't have any spare money right now, so I'm working on a couple of articles, but mostly, I'm correcting grammatical and spelling errors whenever I see them. This is an excellent thing for everyone with good language skills to do, and it's almost effortless. Simply editing the text of an article to correct errors or to replace an awkward phrase doesn't require one to learn Wikipedia's peculiar markup system.
Of course, this only applies to you if you're part of the minority of Slashdot readers who know how to spell "ludicrous" and "ridiculous," can tell "e.g." from "i.e.," know that the expression is "just as soon," not "just assume," and understand that, unlike in C, the closing punctuation mark in English comes before the final quote, not after.
This attitude is so wrong in so many ways that it occasionally baffles me that it's so popular.
My feet ain't broke, either, but if I want to get from Portland to Atlanta in less than a year, I'm going to take a plane. My next big programming project probably isn't going to be in 6502 assembly language, either.
A better rule might be, "If it can be improved, improve it," with the caveat, often ignored, "Make sure it's really an improvement." Meanwhile, the don't fix it crowd can write SQL queries by the light of the fire they started by rubbing two sticks together when something better than SQL comes along.
On to the next part. XML serves its purpose very well.
Actually, it's fundamentally flawed. It would have been far better to define a specification language for describing the structure of existing binary representations, and some sort of procedural language for making transformations between them, instead of stuffing everything into a hierarchal format -- as if all, or even most, data was merely hierarchal -- that makes uuencoding look like a form of compression by comparison. But we probably do have to live with that now that the whole industry is on the bloody bandwagon, just like we have to accept the more ridiculous excesses of OO, another good idea carried to its reductio ad absurdum by obsessive-compulsive freaks bent on completeness and symmetry at the expense of basic common sense.
Life *is* open source already
on
Open Source Life?
·
· Score: 4, Insightful
What happens when a bio-cracker unleashes a plant virus on all the wheat in North America, and the genetic code to 'Wheat 2.0' is closed-source, patented code owned by a corporation? Should life be Open Source?
What, so Wheat 2.0's team of volunteer geneticists can rush out a quick patch? And would you like user-contributions to be in the form of digital sequences, or would you rather have us do a little quick PCR with the live stuff and ship you a test tube full of DNA?
Living organisms are open source already. Given the necessary hardware and the accompanying wads of cash, you can crack open any nucleus you want and sequence its chromosomes until the cows come home. Sure, it's uncommented, but it's not like Monsanto is sponsoring an annual obfuscated protein sequence contest, and if you're allergic to uncommented code, Open Source is definitely not your cup of tea.
If you're really concerned about engineered agricultural diseases, you might want to consider the solution that 3.8 billion years of evolution came up with: genetic diversity. If you don't have three midwestern states entirely covered with the same clones, it's going to be much, much harder to obliterate the whole crop.
Harlan's books are rarely posted to alt.binaries.ebooks. The only times i have seen it happen is after he has one of his legendary tanrums.
I can vouch for this. I download pretty much everything that appears in alt.binaries.ebooks. Not to read it, mind you -- I can't imagine reading anything book-length on a monitor -- but because it's the best source of contemporary English prose for my natural language processing experiments.
Harlan Ellison's books are seldom posted. I wish I could say the same for David Drake.
I suggest looking at getting reliable drives before looking at a RAID solution.
Although the original poster isn't very clear on the point, it sounds like this is more for personal or some other small-scale use than a high throughput, high availability system. In such a case, SCSI drives are a tremendous waste of money. 600 gigs of high-quality SCSI drives is much, much more expensive than 600 gigs of an IDE RAID-5 array.
One question to ask is what kind of applications will be using this array, and how bad will it be if you have to suffer through degraded performance between when a drive fails and when you get the replacement installed. If all you're doing is storing your monster MP3 collection and some word processor files, performance is not an issue with even the slowest modern drives. If you're serving documents to a busy office of dozens or hundreds of users, then performance must be taken into account. If you're going to store the active copy of your corporate financial database, then you probably are talking SCSI and RAID-5, plus offsite backups.
So you want to hear these lame proposals so you can scoff at them and feel superior? Or what?
I think it's probably the "somebody do something" reflex that gives us so many poorly-conceived laws.
The current email system is fundamentally flawed. You can heap all the additional crap on top of it that you want, but then it will just be a fundamentally flawed system with additional complexity. None of the schemes thus far proposed are workable, and none of them address the root problems.
Worse, some of them ignore other problems with email in their haste to deal with spam. Take, for example, the fact that Joe User (read: most people) use email as a file transfer protocol. Given that these turkeys are never going to adapt to using ftp, how about we at least implement an 8-bit clean mail system so files can be sent without being bloated by MIME encoding?
As far as spam goes, no law is going to help, as the Internet is not bound by borders -- just ask the censors in mainland China who, with considerable more police power at their disposal than any free state, still can't keep a lid on things. For the same reason, being able to authenticate the source of an email is at best going to aid in automated filtering via blacklists and whitelists. Just because you're on the US do-not-spam list doesn't mean that some turkey in Thailand isn't going to send you tons of traceable spam.
And I don't know about you, but I'd rather keep feeding my copy of bogofilter than have my email traffic closely scrutinized by corporate interests and open to subpoenas by overzealous prosecutors.
Of course, I'm still limited to doing very basic things, but I'm beginning to understand the power of POV--especially the fact that it's a complete language. I find it amazing that people have written macros that will automatically generate everything from trees to whole cities.
That, really, is the beauty of POV-Ray. I also use various commercial modelers, but when I need (or just want) to do something algorithmically, I turn to POV-Ray. Personally, I find the macro language annoying, so I usually use Perl to generate POV code. My last for-fun project was a series of Perl scripts that generated fairly detailed office and apartment buildings in a variety of architectural styles, with most of the parameters being tweakable with commandline arguments. I first used POV back in 1990 or so, while it was still called DKB-Trace, and used it as a backend for a C program which took spreadsheet data and made bar and pie graphs that completely blew away the stuff that came out of Excel or Quattro Pro. (Of course, a simple bar graph took something like three hours to render on my 386, but that's beside the point.)
There are a number of free and cheap commercial utilities to convert from common commercial 3D file formats to POV if you prefer to use a modeler. For many years, I used POV to do renderings from 3D Studio models because the 3D Studio renderer sucked by comparison.
People are familiar with the interface of Lotus 123. We shouldn't try to change that fact and expect people to learn how to use a mouse.
The current interface of Excel and the interface of Lotus 123 differ very little. In fact, the only significant change has been mouse support and cosmetic details. And therein lies an important point -- you can get away with adding to an interface; what confuses users is when you remove existing features or substantially change the way they work.
I really can't understand arguments like the one OSNews makes. If people hate the interface then they hate the interface.
I don't understand it either. When Microsoft did this kind of thing with some dubious interface decisions in the Win9x series, I could understand it -- what does a monopoly care? Of course, I also switched to Linux around that time.
When Nautilus started appearing as the default desktop in some distros, I spent a day or two with it and then figured out how to kill it and installed an obscure file manager from source that basically imitated Windows Explorer. I found Nautilus to be exactly what you'd expect coming from people who thought they were going to revolutionize the industry with, um, a file browser.
The point that so many UI programmers fail to understand is that nobody gives a shit about the interface except UI programmers. Everyone else just wants their next piece of software to have basically the same interface as their last piece of software. They may want a few minor tweaks to soothe some annoyance or other, but they don't want to spend time relearning the interface because the interface itself is not important to them. It's just a way of getting to the features that they need. Period.
Windows users want to migrate to Linux and use it the way they're used to using Windows. One presumes that Mac users, Amiga users, and every other kind of user out there would like to avoid a serious learning curve, too. People spend years developing working habits with their systems, and they will need a stronger incentive than the philosophical ramblings of UI programmers to do otherwise.
Why it's so damn hard just to take a familiar model and clone it is beyond me, especially when that's the primary barrier that stands between FOSS and the demise of Microsoft.
"if (flag)" and "if (flag != 0)" are exactly equivalent, and a compiler that produces different code for them is severely brain-damaged.
Then the overwhelming majority of compilers are severely brain-damaged, but then, you'd have to know assembly language to realize that. (Or type promotion rules in C.)
We have compilers for a reason, to produce assembly code as efficiently as possible for a higher level language. Most 99% of the time, the compiler will optomize the code just as well, or better than you can.
Compilers don't optimize everything, though. No compiler will correct for your inefficient use of memory, for example. Compilers won't compensate for your failure to understand how the CPU cache works, secondary storage and network latency, etc. Neither will assembly, mind you, but when writing in assembly, you have to think about these things, and when you later spend 99% of your time in a high-level language, you will at least be conscious of them.
One thing I see in programs all the time that just boggles my mind are elaborate sets of conditionals revolving around the value of several boolean variables -- usually implemented in C as ints. You'll see something like this...
...which wastes (on a 32-bit machine) 128 bits to store what only needs four bits, and a bunch of clock cycles to load and test four separate values before executing a jump. If all these values were bits in a single int (or short int, for that matter), you could do a bitwise mask by ORing several constants together...
if(flagvar & (FLAG1|FLAG2|FLAG3|FLAG4)) { /* something happens here */ }
...and get your result much more efficiently. This is going to be flamingly obvious to someone with assembly language experience, but will probably not occur to someone who recently learned C++ or Java as a first language in school. Nor will it be obvious that
if(flag)
is more efficient than
if(flag != 0)
nor will it be clear why.
Now, none of this is a recommendation that everything be written in assembly language. You should probably always use the highest-level language that meets your needs. But knowing what's happening "down there" will help with those high-level languages, and won't necessarily make your code unclear.
With more than 30,000 restaurants around the globe and more than 1,200 in Germany alone, McDonald's is the undisputed market leader in the fast food sector.
You know, until I saw this, I never consciously realized that I had never, ever thought of McDonald's locations as restaurants. I think the proper term here is joint.
Apparently the market noticed. Volume was high and the price of SCOX shares declined by 10.55%. Anyone still along for the ride at this point is either not paying attention or else hoping that Darl will pull a penguin out of his ass at the last minute.
the grey goo scenario IS NOT POSSIBLE because it has not happened
Given the infinitesimal fraction of the universe we can observe directly in detail, the preceding statement is a bit like, "The Chinese are not possible, because there are no Chinese in my living room."
anyone who who seriously thought a-bomb tests would ignite the atmosphere was applying as much logial brain power as those people who thought humans would suffocate at the dizzying speeds of 30mph on the early steam trains.
Yes, the idea of igniting the atmosphere was stupid. But the idea that nuclear fission was going to provide cheap, clean, limitless power was also stupid, but at one point widely believed. The idea that nanotech will be a miracle technology with no dangers might well be equally naive.
And while no one suffocated on a 19th century train travelling at 30mph, both the rate of acceleration and velocities achieved by 20th century military aircraft are capable of causing injury or death, hence closed cockpits and flight suits. The full potential of a technology is seldom immediately obvious in its early prototypes.
Being irrationally fearless is no better than being irrationally fearful.
I use both. The eclipse development environment got this feature WELL after the dot net betas had it. However, I think they both cloned it from NetBeans...
I hate to point out the obvious, but this is a lot older than Eclipse or NetBeans. I don't know what it's called where you're from, but my people call it "grep".
Of course, our grep is a generalized tool, so not only can we build lists of "TODO" comments, we can hunt down "FIXME" and even "This is an ugly hack" and "I am so ashamed of this".
1. Money. Windows is expensive. It's an operating system, for crying out loud. Why should I have to pay for an operating system?
2. Security. I don't mean security from "hackers". I mean I want to be sure that my OS isn't reporting information back to HQ.
3. DRM. Don't want it.
4. Power. Linux comes with an amazing array of development tools. I know this probably doesn't matter to Joe User, but when I got into computers, "user" and "programmer" were synonymous. I'm still a programmer. And I still don't want to pay hundreds or thousands of dollars for compilers.
5. Stability. Frankly, Windows' bugginess doesn't bother me too much on a desktop. You get used to it. But I wouldn't want to run a server on it.
6. Efficiency. I don't like to buy new machines any more often than I have to. To quote Bill Gates, "What do I look like? The queen?" If I have to upgrade my hardware, it better be because of an actual application, not my freaking OS.
7. Accountability. Closed-source companies are accountable to no one. If they close up shop, I'm screwed if I need their app. With open source, that can't happen to me.
All Windows has that I can't live without comes from Adobe. When Adobe sees the light or WINE supports Photoshop, MS can kiss my skinny white butt.
Every year at the University of Waterloo the Computer Engineering and Computer Science students personally build their own operating systems (including documentation) in less than four months. This is done without any prior knowledge of how OSes work and without being taught C.
Which is what I've been saying since this crap came out. Ken Brown might be surprised to learn that many, many students have written compilers as class projects, too. (Having done both, frankly, the compiler was harder.)
The really amazing thing is just how many free (both as in beer and as in speech) operating systems there are out there. Last time I checked, there were dozens and dozens in an operational state, some of which are a hell of a lot better than Linux 0.01 was, many of which are maintained by one or two people alone, and a few of which are actually written in hand-coded assembly language, which is a damn sight harder to do than cranking out C. More than a couple of them have been featured on Slashdot.
Yes, writing a full-featured, mature UNIX-like operating system is hard. The reason Linux is one such OS is not because Linus Torvalds cranked out a buggy, minimalistic stub of an operating system all these years ago. Any halfway decent programmer with spare time and motivation can do that. It's because thousands and thousands of talented programmers took that buggy, minimalistic stub and have been cheerfully adding to it for a decade.
You can grow a garden in horseshit, Mr. Brown, but that doesn't mean you pulled the garden straight out of the horse's ass. Your book, on the other hand...
Slashdot Mirror
I think the parent poster was talking about reaching back in time to grab a person from the 1920's, bringing them forward 80 years, and getting their reactions to how things have changed.
I know that I would be much more surprised by the apparent gross violation of the laws of physics involved in the actual time travel.
Then I would ask Captain Kirk for his autograph.
I bet if people from the 1920's see the world today, they would be alarmed by the technology and hitech gadgets (simple automatic doors for that matter) around them. I want to feel like that when I think of the future - not just some old computer capable of working really really fast.
My grandmother is from the 1920's, and if she is alarmed by automatic doors, she has never shown any sign of it. I suspect she doesn't know how motion sensors work, but then, she doesn't care. She is perplexed by the blinking 12:00 on her VCR, but that appears to be common in people from the 1970's as well. Mostly, she is alarmed that you can say "bitch" on TV and that her bursitis is acting up again.
She couldn't care less about computers, as she has no need for them. Her TV provides her with all of her "stories", and when she wants to contact someone in a hurry, she talks into a telephone, which she regards as a better use of energy than learning to type, buying an expensive computer, and paying for an ISP.
As near as I can tell, the only pieces of post-WW2 technology she uses are modern dentistry and insulin. Strangely, she seems more content than I am, so I have a feeling that when I'm her age, my disinterest in the latest gadgets will probably be minimal, too.
Who knows where we'll be 10 years from now[?]
At this rate, we'll all be sitting in cube farms with a sign reading "Longhorn macht frei" hanging over our heads.
Are we the message?
What sort of message would that be? Monkeys are assholes, and monkeys with large brains are assholes with nuclear weapons?
The "advanced civilization" that actually has to conduct that experiment would either be idiot-savants with a knack for genetic engineering or else rodeo clowns snatched from the future by alien overlords to act as viceroys over a slave class of biology graduate students.
including one from 1826 for the first internal combustion engine...
Well, of course-- that would be the original SCO internal combustion engine, the principles of which have been stolen by every car on the planet!
Wikipedia has replaced Google as my favorite site. It's arguably the one site I would actually pay to access, and I'm so grateful I don't have to.
That being said, I don't like being a leech, but I don't have any spare money right now, so I'm working on a couple of articles, but mostly, I'm correcting grammatical and spelling errors whenever I see them. This is an excellent thing for everyone with good language skills to do, and it's almost effortless. Simply editing the text of an article to correct errors or to replace an awkward phrase doesn't require one to learn Wikipedia's peculiar markup system.
Of course, this only applies to you if you're part of the minority of Slashdot readers who know how to spell "ludicrous" and "ridiculous," can tell "e.g." from "i.e.," know that the expression is "just as soon," not "just assume," and understand that, unlike in C, the closing punctuation mark in English comes before the final quote, not after.
Every time U2 comes out with an album, I mail a money order for the cost of the CD to Negativland.
It aint broke and don't need fixin'
This attitude is so wrong in so many ways that it occasionally baffles me that it's so popular.
My feet ain't broke, either, but if I want to get from Portland to Atlanta in less than a year, I'm going to take a plane. My next big programming project probably isn't going to be in 6502 assembly language, either.
A better rule might be, "If it can be improved, improve it," with the caveat, often ignored, "Make sure it's really an improvement." Meanwhile, the don't fix it crowd can write SQL queries by the light of the fire they started by rubbing two sticks together when something better than SQL comes along.
On to the next part. XML serves its purpose very well.
Actually, it's fundamentally flawed. It would have been far better to define a specification language for describing the structure of existing binary representations, and some sort of procedural language for making transformations between them, instead of stuffing everything into a hierarchal format -- as if all, or even most, data was merely hierarchal -- that makes uuencoding look like a form of compression by comparison. But we probably do have to live with that now that the whole industry is on the bloody bandwagon, just like we have to accept the more ridiculous excesses of OO, another good idea carried to its reductio ad absurdum by obsessive-compulsive freaks bent on completeness and symmetry at the expense of basic common sense.
What happens when a bio-cracker unleashes a plant virus on all the wheat in North America, and the genetic code to 'Wheat 2.0' is closed-source, patented code owned by a corporation? Should life be Open Source?
What, so Wheat 2.0's team of volunteer geneticists can rush out a quick patch? And would you like user-contributions to be in the form of digital sequences, or would you rather have us do a little quick PCR with the live stuff and ship you a test tube full of DNA?
Living organisms are open source already. Given the necessary hardware and the accompanying wads of cash, you can crack open any nucleus you want and sequence its chromosomes until the cows come home. Sure, it's uncommented, but it's not like Monsanto is sponsoring an annual obfuscated protein sequence contest, and if you're allergic to uncommented code, Open Source is definitely not your cup of tea.
If you're really concerned about engineered agricultural diseases, you might want to consider the solution that 3.8 billion years of evolution came up with: genetic diversity. If you don't have three midwestern states entirely covered with the same clones, it's going to be much, much harder to obliterate the whole crop.
Harlan's books are rarely posted to alt.binaries.ebooks. The only times i have seen it happen is after he has one of his legendary tanrums.
I can vouch for this. I download pretty much everything that appears in alt.binaries.ebooks. Not to read it, mind you -- I can't imagine reading anything book-length on a monitor -- but because it's the best source of contemporary English prose for my natural language processing experiments.
Harlan Ellison's books are seldom posted. I wish I could say the same for David Drake.
I suggest looking at getting reliable drives before looking at a RAID solution.
Although the original poster isn't very clear on the point, it sounds like this is more for personal or some other small-scale use than a high throughput, high availability system. In such a case, SCSI drives are a tremendous waste of money. 600 gigs of high-quality SCSI drives is much, much more expensive than 600 gigs of an IDE RAID-5 array.
One question to ask is what kind of applications will be using this array, and how bad will it be if you have to suffer through degraded performance between when a drive fails and when you get the replacement installed. If all you're doing is storing your monster MP3 collection and some word processor files, performance is not an issue with even the slowest modern drives. If you're serving documents to a busy office of dozens or hundreds of users, then performance must be taken into account. If you're going to store the active copy of your corporate financial database, then you probably are talking SCSI and RAID-5, plus offsite backups.
So you want to hear these lame proposals so you can scoff at them and feel superior? Or what?
I think it's probably the "somebody do something" reflex that gives us so many poorly-conceived laws.
The current email system is fundamentally flawed. You can heap all the additional crap on top of it that you want, but then it will just be a fundamentally flawed system with additional complexity. None of the schemes thus far proposed are workable, and none of them address the root problems.
Worse, some of them ignore other problems with email in their haste to deal with spam. Take, for example, the fact that Joe User (read: most people) use email as a file transfer protocol. Given that these turkeys are never going to adapt to using ftp, how about we at least implement an 8-bit clean mail system so files can be sent without being bloated by MIME encoding?
As far as spam goes, no law is going to help, as the Internet is not bound by borders -- just ask the censors in mainland China who, with considerable more police power at their disposal than any free state, still can't keep a lid on things. For the same reason, being able to authenticate the source of an email is at best going to aid in automated filtering via blacklists and whitelists. Just because you're on the US do-not-spam list doesn't mean that some turkey in Thailand isn't going to send you tons of traceable spam.
And I don't know about you, but I'd rather keep feeding my copy of bogofilter than have my email traffic closely scrutinized by corporate interests and open to subpoenas by overzealous prosecutors.
Of course, I'm still limited to doing very basic things, but I'm beginning to understand the power of POV--especially the fact that it's a complete language. I find it amazing that people have written macros that will automatically generate everything from trees to whole cities.
That, really, is the beauty of POV-Ray. I also use various commercial modelers, but when I need (or just want) to do something algorithmically, I turn to POV-Ray. Personally, I find the macro language annoying, so I usually use Perl to generate POV code. My last for-fun project was a series of Perl scripts that generated fairly detailed office and apartment buildings in a variety of architectural styles, with most of the parameters being tweakable with commandline arguments. I first used POV back in 1990 or so, while it was still called DKB-Trace, and used it as a backend for a C program which took spreadsheet data and made bar and pie graphs that completely blew away the stuff that came out of Excel or Quattro Pro. (Of course, a simple bar graph took something like three hours to render on my 386, but that's beside the point.)
There are a number of free and cheap commercial utilities to convert from common commercial 3D file formats to POV if you prefer to use a modeler. For many years, I used POV to do renderings from 3D Studio models because the 3D Studio renderer sucked by comparison.
People are familiar with the interface of Lotus 123. We shouldn't try to change that fact and expect people to learn how to use a mouse.
The current interface of Excel and the interface of Lotus 123 differ very little. In fact, the only significant change has been mouse support and cosmetic details. And therein lies an important point -- you can get away with adding to an interface; what confuses users is when you remove existing features or substantially change the way they work.
I really can't understand arguments like the one OSNews makes. If people hate the interface then they hate the interface.
I don't understand it either. When Microsoft did this kind of thing with some dubious interface decisions in the Win9x series, I could understand it -- what does a monopoly care? Of course, I also switched to Linux around that time.
When Nautilus started appearing as the default desktop in some distros, I spent a day or two with it and then figured out how to kill it and installed an obscure file manager from source that basically imitated Windows Explorer. I found Nautilus to be exactly what you'd expect coming from people who thought they were going to revolutionize the industry with, um, a file browser.
The point that so many UI programmers fail to understand is that nobody gives a shit about the interface except UI programmers. Everyone else just wants their next piece of software to have basically the same interface as their last piece of software. They may want a few minor tweaks to soothe some annoyance or other, but they don't want to spend time relearning the interface because the interface itself is not important to them. It's just a way of getting to the features that they need. Period.
Windows users want to migrate to Linux and use it the way they're used to using Windows. One presumes that Mac users, Amiga users, and every other kind of user out there would like to avoid a serious learning curve, too. People spend years developing working habits with their systems, and they will need a stronger incentive than the philosophical ramblings of UI programmers to do otherwise.
Why it's so damn hard just to take a familiar model and clone it is beyond me, especially when that's the primary barrier that stands between FOSS and the demise of Microsoft.
"if (flag)" and "if (flag != 0)" are exactly equivalent, and a compiler that produces different code for them is severely brain-damaged.
Then the overwhelming majority of compilers are severely brain-damaged, but then, you'd have to know assembly language to realize that. (Or type promotion rules in C.)
Compilers don't optimize everything, though. No compiler will correct for your inefficient use of memory, for example. Compilers won't compensate for your failure to understand how the CPU cache works, secondary storage and network latency, etc. Neither will assembly, mind you, but when writing in assembly, you have to think about these things, and when you later spend 99% of your time in a high-level language, you will at least be conscious of them.
One thing I see in programs all the time that just boggles my mind are elaborate sets of conditionals revolving around the value of several boolean variables -- usually implemented in C as ints. You'll see something like this......which wastes (on a 32-bit machine) 128 bits to store what only needs four bits, and a bunch of clock cycles to load and test four separate values before executing a jump. If all these values were bits in a single int (or short int, for that matter), you could do a bitwise mask by ORing several constants together......and get your result much more efficiently. This is going to be flamingly obvious to someone with assembly language experience, but will probably not occur to someone who recently learned C++ or Java as a first language in school. Nor will it be obvious thatis more efficient thannor will it be clear why.
Now, none of this is a recommendation that everything be written in assembly language. You should probably always use the highest-level language that meets your needs. But knowing what's happening "down there" will help with those high-level languages, and won't necessarily make your code unclear.
Do artists have a viable choice?
Sure. I see them working in record stores all the time!
With more than 30,000 restaurants around the globe and more than 1,200 in Germany alone, McDonald's is the undisputed market leader in the fast food sector.
You know, until I saw this, I never consciously realized that I had never, ever thought of McDonald's locations as restaurants. I think the proper term here is joint.
Apparently the market noticed. Volume was high and the price of SCOX shares declined by 10.55%. Anyone still along for the ride at this point is either not paying attention or else hoping that Darl will pull a penguin out of his ass at the last minute.
I wish to hell I'd sold short yesterday.
the grey goo scenario IS NOT POSSIBLE because it has not happened
Given the infinitesimal fraction of the universe we can observe directly in detail, the preceding statement is a bit like, "The Chinese are not possible, because there are no Chinese in my living room."
anyone who who seriously thought a-bomb tests would ignite the atmosphere was applying as much logial brain power as those people who thought humans would suffocate at the dizzying speeds of 30mph on the early steam trains.
Yes, the idea of igniting the atmosphere was stupid. But the idea that nuclear fission was going to provide cheap, clean, limitless power was also stupid, but at one point widely believed. The idea that nanotech will be a miracle technology with no dangers might well be equally naive.
And while no one suffocated on a 19th century train travelling at 30mph, both the rate of acceleration and velocities achieved by 20th century military aircraft are capable of causing injury or death, hence closed cockpits and flight suits. The full potential of a technology is seldom immediately obvious in its early prototypes.
Being irrationally fearless is no better than being irrationally fearful.
I use both. The eclipse development environment got this feature WELL after the dot net betas had it. However, I think they both cloned it from NetBeans...
I hate to point out the obvious, but this is a lot older than Eclipse or NetBeans. I don't know what it's called where you're from, but my people call it "grep".
Of course, our grep is a generalized tool, so not only can we build lists of "TODO" comments, we can hunt down "FIXME" and even "This is an ugly hack" and "I am so ashamed of this".
1. Money. Windows is expensive. It's an operating system, for crying out loud. Why should I have to pay for an operating system?
2. Security. I don't mean security from "hackers". I mean I want to be sure that my OS isn't reporting information back to HQ.
3. DRM. Don't want it.
4. Power. Linux comes with an amazing array of development tools. I know this probably doesn't matter to Joe User, but when I got into computers, "user" and "programmer" were synonymous. I'm still a programmer. And I still don't want to pay hundreds or thousands of dollars for compilers.
5. Stability. Frankly, Windows' bugginess doesn't bother me too much on a desktop. You get used to it. But I wouldn't want to run a server on it.
6. Efficiency. I don't like to buy new machines any more often than I have to. To quote Bill Gates, "What do I look like? The queen?" If I have to upgrade my hardware, it better be because of an actual application, not my freaking OS.
7. Accountability. Closed-source companies are accountable to no one. If they close up shop, I'm screwed if I need their app. With open source, that can't happen to me.
All Windows has that I can't live without comes from Adobe. When Adobe sees the light or WINE supports Photoshop, MS can kiss my skinny white butt.
Every year at the University of Waterloo the Computer Engineering and Computer Science students personally build their own operating systems (including documentation) in less than four months. This is done without any prior knowledge of how OSes work and without being taught C.
Which is what I've been saying since this crap came out. Ken Brown might be surprised to learn that many, many students have written compilers as class projects, too. (Having done both, frankly, the compiler was harder.)
The really amazing thing is just how many free (both as in beer and as in speech) operating systems there are out there. Last time I checked, there were dozens and dozens in an operational state, some of which are a hell of a lot better than Linux 0.01 was, many of which are maintained by one or two people alone, and a few of which are actually written in hand-coded assembly language, which is a damn sight harder to do than cranking out C. More than a couple of them have been featured on Slashdot.
Yes, writing a full-featured, mature UNIX-like operating system is hard. The reason Linux is one such OS is not because Linus Torvalds cranked out a buggy, minimalistic stub of an operating system all these years ago. Any halfway decent programmer with spare time and motivation can do that. It's because thousands and thousands of talented programmers took that buggy, minimalistic stub and have been cheerfully adding to it for a decade.
You can grow a garden in horseshit, Mr. Brown, but that doesn't mean you pulled the garden straight out of the horse's ass. Your book, on the other hand...
"I plan to donate a grape iMac..."
Only a Mac user would consider the color of the case to be a technical detail
(ducking)