I did that for almost a year, except with an actual desktop system. I did get funny looks lugging that thing into the office every morning, but it worked out pretty well. Maybe geeks would lose the wimp image if they had to lug a full tower to work every day.
I haven't read his book; his site deals with coding issues, not interface ones. And I don't agree with a lot of what he says, but he's very practically-minded, and a good antidote to academic thinking.
I haven't read Starship Troopers, only seen the movie (hides face in shame). But it seemed to me like heavy-handed sarcasm, hardly in support of war. Like what you get at Adequacy.
I feel your pain, but once you've been using it for a while you'll have discovered all kinds of things that will have you trying non-functional key combinations on your Windows box and getting frustrated when you don't work. (Small example, in just about any window manager, including KDE, you can customize every key combination. So you can even have your alt-F4 to close windows if you like that. Or the infamous "emacs keys" which just about every system but windows and macs has standardized on for text editing.)
"The syntax makes sense to me." -- The opposite is precisely the reason I never got too deep into Perl, although I've done a couple projects with it. There was just SO MUCH syntax, and so many wierd things that were hard to remember. I actually learned to program in Pascal, then a bit of C++, and most of my experience in PHP (how ironic), but to me Python has near-perfect syntax; writing Python feels to me like writing pseudo-code, like breathing.
Perl, NEVER. I wonder what it is about some people that make them call Perl's syntax intuitive; we can't be that different. Perhaps it's that the syntax takes forever to learn, but once you've learned it you've also learned everything else about the language. Whereas in Python, anyone can learn all the syntax in a few hours, but mastering the language takes just as long as any other.
I'm learning German and I think the difference is similar; I absolutely hate all the grammar, genders, cases, stuff like that, but once you've got that down you mostly have the language. Whereas English is the opposite; desceptively simple at a low level, but terribly hard to master. (This implies German:Perl::English:Python, though, and I'd say it's almost the opposite, so maybe my argument is confused.)
If you haven't read the original Gentle Art of Verbal Self Defense by the same author,
check it out. Like the book you mention, it makes you think a lot about the motives behind your words and others', but on the person-to-person level. Very interesting stuff.
I agree 100%! This is why power users love *nix so much, even though it's basically 20 year-old interface technology; because the interface is much closer to the actual capabilities of the computer. If the interface isn't smart enough to help me, then it should at least stay out of the way.
I disagree with you in one important respect: "normals" are capable of transcending the limitations you have ascribed to them. To label them as a group "stupid" and "unimaginative" is making the same mistake as labeling autistics "retarded" and "socially inept." Some are even more interesting because of their fight against the "normal" tendencies which you are free of.
Everyone feels that their viewpoint is somehow superior to everyone else's -- this cannot be helped, it is the nature of being conscious. The trick is to seperate the feeling of rightness from the intellectual certainty thereof.
Get laid, are you kidding? He just has to snap his fingers. If you mean have a meaningful relationship, well of course that's a whole 'nother ball of wax.
But the point is Feynman was still abnormal. He was like a big kid. His relationships with women were unstable; he couldn't use diplomacy in dealing with people; he didn't like to think about the consequences of his research but did it just because it was fun. And these things made him a good scientist and a great person, but there's no way he could have become President, for example.
First, my argument is not with your characterization of Linux vs Windows vs Mac. It is with the implication that user interface research has discovered anything interesting since the 60s. You didn't actually argue against that (usability testing doesn't count as an innovative discovery), so I can't argue back.
Here's a good analogy; interface design = psychology. Neither is actually a science. Neither admits measurable progress. Neither has any formal theoretical basis. Heck, HF doesn't even have an interesting informal theoretical basis like Freud.
"If you're aiming for a world dominated by user-friendly GUIs, not geek-friendly command lines, that sort of integration [into the modern GUI environment] is essential."
As long as there is a difference between "geek-friendly" and "user-friendly," interface design -- as an art or a science -- is lacking.
"Or, to put it another way: the category of the process is not of neccessity related to the category of the product."
Ah, I did not confuse the two although you put it much more clearly than I did. My point can now be simply stated as: "art" is always a process and never a product.
I say that simply because I can think of no reasonable, universal way of defining "art" for a product. So while I agree that a painting and a computer program may not both be artistic products, to elucidate the difference will require more than a simple one-word categorization.
Any definition of art based on what's "aesthetically pleasing" will be agreed upon by noone and fail to cover half the things most poeple agree are art anyways.
Here's why coding is an art: it can't (yet) be done well by a computer. That is, we don't know all the rules that go into making a good program, we haven't codified them. This is in contrast to most real engineering disciplines, where it is very easy to quantify a superior solution; the only trick is to find it.
What makes something art is how it is made, not what it is ultimately used for. As long as some part of writing good code remains locked in the coder's head it is as much a deep and mysterious art as writing a symphony or creating a painting.
Being that the relational model is a superset of the hierarchical model, it needn't be more complicated, and it's much closer to the way we think. Just imagine an interface where instead of remembering some complex hierarchical structure the user can go, "hmm, it was a business letter, and it was about fred," click on a business icon and a fred icon, and there's his file.
You make the incorrect (very Western) assumption that "old is worse". Like we've actually made advances in UI design in the past 30 years. What's happened is that people have had a long time to get used to the current paradigm (which is, actually, still basically what we had 30 years ago).
Pick up The Humane Interface by Jeff Raskin sometime. And don't assume that Progress inevitably makes things Better.
Oh that just works great when you switch companies, go to the computer lab at school...
First thing, just looking different isn't that big a deal to casual users. Everyone I know can go between Mac and Windows if they have to, and those two have some really big differences in functionality. (Like how to quit an application.)
Second thing, the beauty of Linux is that with rc-files I can fit everything I need to set up my perfect interface (i.e. mutt, fetchmail, window manager config, emacs, vi, etc.) on a floppy. I can have any reasonably-equipped machine behaving just like mine in under a minute. Try doing that on Windows...
White space: I'm tired of people referring to this as something Python "got wrong". You may not agree with it, but it's intentional and some people find it superior. I've never met anyone who disliked it after having actually tried it.
I think part of the problem is that "corporate interests" translates into "anybody who wants to make money." The internet is like a heat engine for information; by moving it from one place to another, it literally makes money. Nothing will stop people from capitalizing on that.
Essentially don't DNA computers and quantum computers operate in the same way: simultaneously computing over all possible values (or in DNA, a whole heck of a lot of values), and retrieving the single correct answer at the end (via collapse in QM, via chemical filtering in DNA)?
Sounds like they took relational (database) theory, changed some terms around, and slapped a cute interface on it.
The comments saying it's been done before are right, but hierarchical filesystems ain't it. Although I will be really excited when we get a fully relational filesystem.
Slashdot Mirror
Somebody mod this abstract nonsense down.
I did that for almost a year, except with an actual desktop system. I did get funny looks lugging that thing into the office every morning, but it worked out pretty well. Maybe geeks would lose the wimp image if they had to lug a full tower to work every day.
I haven't read his book; his site deals with coding issues, not interface ones. And I don't agree with a lot of what he says, but he's very practically-minded, and a good antidote to academic thinking.
I haven't read Starship Troopers, only seen the movie (hides face in shame). But it seemed to me like heavy-handed sarcasm, hardly in support of war. Like what you get at Adequacy.
He says the exact same thing as you, only better.
I feel your pain, but once you've been using it for a while you'll have discovered all kinds of things that will have you trying non-functional key combinations on your Windows box and getting frustrated when you don't work. (Small example, in just about any window manager, including KDE, you can customize every key combination. So you can even have your alt-F4 to close windows if you like that. Or the infamous "emacs keys" which just about every system but windows and macs has standardized on for text editing.)
Perl, NEVER. I wonder what it is about some people that make them call Perl's syntax intuitive; we can't be that different. Perhaps it's that the syntax takes forever to learn, but once you've learned it you've also learned everything else about the language. Whereas in Python, anyone can learn all the syntax in a few hours, but mastering the language takes just as long as any other.
I'm learning German and I think the difference is similar; I absolutely hate all the grammar, genders, cases, stuff like that, but once you've got that down you mostly have the language. Whereas English is the opposite; desceptively simple at a low level, but terribly hard to master. (This implies German:Perl::English:Python, though, and I'd say it's almost the opposite, so maybe my argument is confused.)
If you haven't read the original Gentle Art of Verbal Self Defense by the same author, check it out. Like the book you mention, it makes you think a lot about the motives behind your words and others', but on the person-to-person level. Very interesting stuff.
I agree 100%! This is why power users love *nix so much, even though it's basically 20 year-old interface technology; because the interface is much closer to the actual capabilities of the computer. If the interface isn't smart enough to help me, then it should at least stay out of the way.
Everyone feels that their viewpoint is somehow superior to everyone else's -- this cannot be helped, it is the nature of being conscious. The trick is to seperate the feeling of rightness from the intellectual certainty thereof.
Get laid, are you kidding? He just has to snap his fingers. If you mean have a meaningful relationship, well of course that's a whole 'nother ball of wax.
But the point is Feynman was still abnormal. He was like a big kid. His relationships with women were unstable; he couldn't use diplomacy in dealing with people; he didn't like to think about the consequences of his research but did it just because it was fun. And these things made him a good scientist and a great person, but there's no way he could have become President, for example.
First, my argument is not with your characterization of Linux vs Windows vs Mac. It is with the implication that user interface research has discovered anything interesting since the 60s. You didn't actually argue against that (usability testing doesn't count as an innovative discovery), so I can't argue back.
Here's a good analogy; interface design = psychology. Neither is actually a science. Neither admits measurable progress. Neither has any formal theoretical basis. Heck, HF doesn't even have an interesting informal theoretical basis like Freud.
"If you're aiming for a world dominated by user-friendly GUIs, not geek-friendly command lines, that sort of integration [into the modern GUI environment] is essential."
As long as there is a difference between "geek-friendly" and "user-friendly," interface design -- as an art or a science -- is lacking.
Ah, I did not confuse the two although you put it much more clearly than I did. My point can now be simply stated as: "art" is always a process and never a product.
I say that simply because I can think of no reasonable, universal way of defining "art" for a product. So while I agree that a painting and a computer program may not both be artistic products, to elucidate the difference will require more than a simple one-word categorization.
Is the human brain capable of solving NP-complete problems? That doesn't sound right.
Here's why coding is an art: it can't (yet) be done well by a computer. That is, we don't know all the rules that go into making a good program, we haven't codified them. This is in contrast to most real engineering disciplines, where it is very easy to quantify a superior solution; the only trick is to find it.
What makes something art is how it is made, not what it is ultimately used for. As long as some part of writing good code remains locked in the coder's head it is as much a deep and mysterious art as writing a symphony or creating a painting.
Check out Reiser'S whitepaper.
Pick up The Humane Interface by Jeff Raskin sometime. And don't assume that Progress inevitably makes things Better.
First thing, just looking different isn't that big a deal to casual users. Everyone I know can go between Mac and Windows if they have to, and those two have some really big differences in functionality. (Like how to quit an application.)
Second thing, the beauty of Linux is that with rc-files I can fit everything I need to set up my perfect interface (i.e. mutt, fetchmail, window manager config, emacs, vi, etc.) on a floppy. I can have any reasonably-equipped machine behaving just like mine in under a minute. Try doing that on Windows...
I'm confused, how is it possible to eliminate acceleration and still change velocity? Fuzzy controllers or not.
White space: I'm tired of people referring to this as something Python "got wrong". You may not agree with it, but it's intentional and some people find it superior. I've never met anyone who disliked it after having actually tried it.
I think part of the problem is that "corporate interests" translates into "anybody who wants to make money." The internet is like a heat engine for information; by moving it from one place to another, it literally makes money. Nothing will stop people from capitalizing on that.
Essentially don't DNA computers and quantum computers operate in the same way: simultaneously computing over all possible values (or in DNA, a whole heck of a lot of values), and retrieving the single correct answer at the end (via collapse in QM, via chemical filtering in DNA)?
Damn, does this mean I won't be able to watch porn when I've got a sore throat or a cold?
The comments saying it's been done before are right, but hierarchical filesystems ain't it. Although I will be really excited when we get a fully relational filesystem.