i can guarantee that you can't write a program in C that i can beat in length in lisp... and i bet that i could write a short lisp program that you wouldn't be able to write in C without significant effort.
No doubt, but that wouldn't really be a reasonable comparison because they are different classes of language. C is a machine-level language; Lisp is much higher-level. I guarantee I could write a fast, bit twiddling device driver in C with less effort than you could in Lisp. The results for the ICFP 2003 Programming Contest come out in a week or so. That is always interesting. From memory, Lisp entries have done well, but haven't excelled (though Dylan came second one year).
ML is strongly typed which means its immediately useless. OOP can viewed as a relaxation from strong types (in C) to slightly dynamic types. You watch, eventually most languages will be dynamically typed).
You can't just dismiss static typing so easily. (I think "static" is a more accurate term than "strong" here.) Static vs dynamic typing is the subject of endless religious wars, and there is no clear winner yet. Each has its advantages and disadvantages. My personal preference is to lean towards static typing but with type inferencing (like in ML) where you don't have to declare types unless you want to. I have done a fair amount of programming in dynamically typed languages (in fact, I've just convinced my bosses that a new app we are writing should be written in Python instead of C++. Hooray!).
First off, emacs is butter. Emacs is gorgoes. Emacs is so extensible and so efficient that its stupid. If you sat down and learned it for a week you would be twice as efficient than any windows editor. Emacs is built on lisp and for this reason all of emacs is extensible... everything!!! It's really amazing.
What I like about Windows style editors (CUA style editors) is the simplicity; there's pretty much just the cursor keys, ctrl, shift, etc. I can drive that with the lower parts of my brain -- I don't have to consciously think about it, which is good when programming because then I can think about the problem at hand. I've heard that complicated editors like emacs and vi are good for touch typists. That sounds reasonable, but I'm not a touch typist. And whenever I need to do some fancy text processing, I just write a script.
Concerning Lisp syntax, the thing about Lisp syntax is that its syntax is a parse tree where as most languages (none s-expression languages) require a token/parser stage to compile. Because you work on the sytax tree, macros are trivial to write because you are just performing modifications on the parse tree...
Like I said, it's conceptually elegant. But I still don't think it's readable. Lisp macros are something I'll get around to looking at sometime. I think I'll wait till Arc comes out; hopefully much of Lisp's historical baggage will have been sloughed off and it will be more pleasant to learn.
Oz borrows a lot of syntax from ML which possibly has one of the most barroque syntax I've ever seen.
That's odd because I find ML's sytax to be very sensible and simple (though I've only used OCaml and not any of the other ML dialects) and I find Lisp's unreadable. Many Lisp proponents like to say Lisp's is the one true syntax, but I don't see how that could be true when there must be a large subjective component to why one person prefers one syntax over another.
One of the problems I had when I tried to learn Lisp was that it needs special editor support. This meant, as far as I could tell, I had to learn emacs. I'm not interested in learning emacs. I prefer Windows style editors; their use is burned into my "finger memory" -- I'm used to them and am efficient with them. I did try to learn emacs once, but I found it very complicated. At that time at least, emacs could only scroll half a screen at a time. This seemed to me to be rather unergonomic! There was an add-on available to add line-by-line scrolling, but to install it I needed to Read The very large F'ing Manual and learn Lisp! I don't think it can be said that Lisp's syntax is easy if in order to learn it, you have to learn some complicated tool on top of learning the language itself.
In any case, just looking at Lisp code on paper, I find it hard to read. I can see that everything being S-expressions is conceptually very elegant, but to me there are not enough cues there for my human brain to parse the text easily. No doubt I could learn to read Lisp code with practice, but the initial hurdles are too great for me to bother with it just now, especially when there are so many interesting languages out there. And I find all the Lisp zealotry offputting. (I'm not saying you are a zealot; I'm sure most Lisp users are level-headed, lovely people. But there seem to be a lot of Lisp zealots about.) I find it hard to take any kind of zealot's word about anything being the One True Thing because they always have huge blind spots.
No doubt they learned a lot about operating systems by porting them. It sounds like this guy however is developing a product to sell; he will have to convince people that it is practical.
Re:There's one good thing about it.
on
Perl 1.0?
·
· Score: 5, Informative
Oh dear, it looks like you didn't close one of your HTML blocks properly;) In Python, you can join lines with a backslash:
if 1900 < year < 2100 and 1 <= month <= 12 \
and 1 <= day <= 31 and 0 <= hour < 24 \
and 0 <= minute < 60 and 0 <= second < 60: # Looks like a valid date
return 1
and you can split any kind of bracketed expression over multiple lines:
month_names = ['Januari', 'Februari', 'Maart', # These are the
'April', 'Mei', 'Juni', # Dutch names
'Juli', 'Augustus', 'September', # for the months
'Oktober', 'November', 'December'] # of the year
I don't know that you mean by inline error exception, but you can start comments at the fist column so that they stand out:
# *** inline error exception ***
"something..."
(Some of the formatting in the above examples is messed up a bit by some slashdot bogusness. Actually, there's an argument against Python's whitespace blocks for you -- things like slashdot can mess them up.)
Dijkstra was half joking, trolling even, in that remark about Basic. In the same piece are other hyperbolic comments such as "The use of COBOL cripples the mind; its teaching should, therefore, be regarded as a criminal offence." Have you really read any Dijkstra? He always seems level headed to me. Your rant seems to be about a different kind of person entirely. Are you seriously suggesting he should have been defending the status quo of Basic (and Fortran and Cobol)? That Basic is effective? If someone tells you goto is harmful, by all means try it out yourself, but also go and read the careful reasoning behind that little quote.
Actually, the best way I've found to get into the zone is to leave a compiltation error from the previous day. It is the mental equivalent of a stack pointer from which you can pop all of your registers, reinitialize the idle process, and start cranking.
I do that too; I put an XXX at any unfinished bits. It's visible, and you can search for it.
That scenario seems pretty unlikely precisely because people would find it annoying. It's not easy to sell stuff. People don't like handing over money unless they feel they are getting something of equal or greater value in return. You can't just slap a price tag -- even a tiny one -- on any old piece of crap and expect that people will buy it. They don't have to buy if they don't want to. If they are nagged for a payment every few minutes or otherwise annoyed, then they will go elsewhere!
A better analogy would be changing the funding structure for highways and roads. Instead of taxes, the funding would come from toll booths erected at every turn in town and every couple of miles on the highway. Would you feel annoyed and "nickle and dimed to death"? I would.
Why is that a better analogy? You've simply described a case where micropayments might not work so well. We can't buy road usage on the free market. No one is proposing that micropayments are the best model for everything.
I've been saying it for years, and I continue to be right. Micropayments don't solve a customer problem, they solve a provider problem. If you don't solve a customer problem, you don't have a success. Nobody wants to be nickle and dimed to death on the net. It's time to retire this monumentally dumb idea.
The customer problem to be solved is pretty clear: How can I buy some one-off thing on the Internet that has a small monetary value to me? Currently I can't really do that. Have you ever bought a newspaper or some chewing gum? Those are small transactions. No one forced you to buy them. Did you feel nickel-and-dimed to death?
Valis is amazing, but I don't think it'd be the best one for a PKD newbie to start with because it's so whacked out. Scanner Darkly would be a better one I think. If you liked Valis, BTW, you'd find Radio Free Albemuth interesting -- it's another version of his Valis story.
A list of pointers all pointing to the same value? Why don't you tell me why it would be useFUL? I can't see any good reason for using three copies of the same item, when you could just use that item by itself, saving memory...
Here's a scenario off the top of my head: Say you initialize a list of pointers so that each one points to some "default" object. Any list item can be replaced later by a non-default object (through user action or whatever). Suppose the program wants to modify some properties of the default item at some stage. You'd only need to do that in one place if there is a single default object.
How DO you create three separate lists with this time-saving syntax? Can you?
I don't think you can unless you have some sort of a loop. The most compact way I could think of was a list comprehension:
[x[:] for x in [[0]*2]*3]
Seems silly to me that if I want something useless, like three copies of the same list [...] I can save time typing
It's just a bit of syntactic sugar. I can't recall ever seeing it used much. It's not really important. Besides, why do you think such a list would be useless?
Am i one of the only people that actually DOES use VB6 on a regular basis??! (and, although i do live with my parents and can't afford REAL development tools, i'm only 17 so i think it is reasonable!)
You are in luck, my young friend. You can afford Python because it's free, and you are young enough that any brain damage caused by VB will be entirely reversible!:P
Once you strip away Tycho's rhetoric, he doesn't have much of an argument left. In fact it's not an argument at all, just an assertion: "If you have enough readers who care about your work to go through all that rigamaroll [of signing up for micropayments], you could succeed with any business model." No doubt Tycho knows a thing or two about the webcomics biz; so his opinion counts for something, but still, he doesn't really have anything to back up his assertion. I don't know much about marketing, but I do know that some items are price sensitive. Tycho thinks that if 10 people would pay $1, then 1 person would pay $10. But it's not true in general, and he doesn't tell us why he thinks it's true for webcomics.
I do know a thing or two about reading web comics. I even bought some merchandise once -- I bought some books from achewood.com. (I just mention that to establish that I'm not a complete cheap bastard.) Books and T-shirts are good items for macropayments. But what about other items? Let's stick with Achewood (because it's such a great comic) -- along with their daily free strips they have a weekly, subscription-based color strip. But that requires a macropayment, and I just can't bring myself to make a big payment for comics goddamnit. But I'm sure I'd pay some number of cents for a strip if it were possible. If I liked it, I'd pay again. If I found myself paying every week, I'd buy them in bulk with a subscription. Well fuck me if that isn't an existing business model employed to sells countless other things on this planet! Maybe it would work with comics, maybe it wouldn't. What makes Tycho (and countless Slashdotters) so sure that it wouldn't?
Quote from that link: "The source code for the X Windows System version of Mosaic is available under copyright for individual personal use, for use at an academic institution or for internal business use."
That is not free in an RMS sense. It's not BSD either. BSD is free in the RMS sense.
They do some incredible and brilliant things in the IOCCC, but I can't help thinking that deliberately obfuscating C is immoral somehow:P
What I like about the ICFP is the lack of restrictions. You can solve the problem using whatever tools and programming languages you like, as an individual or as part of a team. I don't think there are any other major programming contests like that.
That's a funny way of classifying things. Your hatred of Windows is sounding really irrational. Not all Unixes have single-machine kernel debugging. Which ones besides Linux do? And SoftICE on Windows allows single machine debugging (or so their blurb says* -- I haven't used it myself). And you didn't account for BeOS (which has a built-in kernel debugger) in your classification scheme;)
------------- * "Debug kernel-mode drivers with ease SoftICE is a powerful kernel mode debugger that supports device driver debugging on either a single or dual machine configuration. "
Ok, fair enough. I should have read the article. The kdb debugger is built in to the kernel. My experience with kernel debugging was using a seperate machine over a serial cable. But no one was suggesting adding a GUI to a built-in kernel debugger!
No, no, debuggers can have GUIs. Kernels can't have GUIs. "Windoze" people generally have strange view of computing? Are you really that much of a zealot, or have I just been trolled?
Huh? All he's saying is that GUI-fied debuggers are nicer than console ones. I've used windbg (one of the ones he mentions) to write and debug kernel-mode drivers on NT. I don't see why you think he hasn't a clue. Just because you use Linux doesn't mean you know everything there is to know about computers!
Slashdot Mirror
Print statements plus lots of assert statements works even better
i can guarantee that you can't write a program in C that i can beat in length in lisp... and i bet that i could write a short lisp program that you wouldn't be able to write in C without significant effort.
No doubt, but that wouldn't really be a reasonable comparison because they are different classes of language. C is a machine-level language; Lisp is much higher-level. I guarantee I could write a fast, bit twiddling device driver in C with less effort than you could in Lisp. The results for the ICFP 2003 Programming Contest come out in a week or so. That is always interesting. From memory, Lisp entries have done well, but haven't excelled (though Dylan came second one year).
ML is strongly typed which means its immediately useless. OOP can viewed as a relaxation from strong types (in C) to slightly dynamic types. You watch, eventually most languages will be dynamically typed).
You can't just dismiss static typing so easily. (I think "static" is a more accurate term than "strong" here.) Static vs dynamic typing is the subject of endless religious wars, and there is no clear winner yet. Each has its advantages and disadvantages. My personal preference is to lean towards static typing but with type inferencing (like in ML) where you don't have to declare types unless you want to. I have done a fair amount of programming in dynamically typed languages (in fact, I've just convinced my bosses that a new app we are writing should be written in Python instead of C++. Hooray!).
First off, emacs is butter. Emacs is gorgoes. Emacs is so extensible and so efficient that its stupid. If you sat down and learned it for a week you would be twice as efficient than any windows editor. Emacs is built on lisp and for this reason all of emacs is extensible... everything!!! It's really amazing.
What I like about Windows style editors (CUA style editors) is the simplicity; there's pretty much just the cursor keys, ctrl, shift, etc. I can drive that with the lower parts of my brain -- I don't have to consciously think about it, which is good when programming because then I can think about the problem at hand. I've heard that complicated editors like emacs and vi are good for touch typists. That sounds reasonable, but I'm not a touch typist. And whenever I need to do some fancy text processing, I just write a script.
Concerning Lisp syntax, the thing about Lisp syntax is that its syntax is a parse tree where as most languages (none s-expression languages) require a token/parser stage to compile. Because you work on the sytax tree, macros are trivial to write because you are just performing modifications on the parse tree...
Like I said, it's conceptually elegant. But I still don't think it's readable. Lisp macros are something I'll get around to looking at sometime. I think I'll wait till Arc comes out; hopefully much of Lisp's historical baggage will have been sloughed off and it will be more pleasant to learn.
Oz borrows a lot of syntax from ML which possibly has one of the most barroque syntax I've ever seen.
That's odd because I find ML's sytax to be very sensible and simple (though I've only used OCaml and not any of the other ML dialects) and I find Lisp's unreadable. Many Lisp proponents like to say Lisp's is the one true syntax, but I don't see how that could be true when there must be a large subjective component to why one person prefers one syntax over another.
One of the problems I had when I tried to learn Lisp was that it needs special editor support. This meant, as far as I could tell, I had to learn emacs. I'm not interested in learning emacs. I prefer Windows style editors; their use is burned into my "finger memory" -- I'm used to them and am efficient with them. I did try to learn emacs once, but I found it very complicated. At that time at least, emacs could only scroll half a screen at a time. This seemed to me to be rather unergonomic! There was an add-on available to add line-by-line scrolling, but to install it I needed to Read The very large F'ing Manual and learn Lisp! I don't think it can be said that Lisp's syntax is easy if in order to learn it, you have to learn some complicated tool on top of learning the language itself.
In any case, just looking at Lisp code on paper, I find it hard to read. I can see that everything being S-expressions is conceptually very elegant, but to me there are not enough cues there for my human brain to parse the text easily. No doubt I could learn to read Lisp code with practice, but the initial hurdles are too great for me to bother with it just now, especially when there are so many interesting languages out there. And I find all the Lisp zealotry offputting. (I'm not saying you are a zealot; I'm sure most Lisp users are level-headed, lovely people. But there seem to be a lot of Lisp zealots about.) I find it hard to take any kind of zealot's word about anything being the One True Thing because they always have huge blind spots.
No doubt they learned a lot about operating systems by porting them. It sounds like this guy however is developing a product to sell; he will have to convince people that it is practical.
Dijkstra was half joking, trolling even, in that remark about Basic. In the same piece are other hyperbolic comments such as "The use of COBOL cripples the mind; its teaching should, therefore, be regarded as a criminal offence." Have you really read any Dijkstra? He always seems level headed to me. Your rant seems to be about a different kind of person entirely. Are you seriously suggesting he should have been defending the status quo of Basic (and Fortran and Cobol)? That Basic is effective? If someone tells you goto is harmful, by all means try it out yourself, but also go and read the careful reasoning behind that little quote.
Yes. I try to make my code a little less pornographic.
Actually, the best way I've found to get into the zone is to leave a compiltation error from the previous day. It is the mental equivalent of a stack pointer from which you can pop all of your registers, reinitialize the idle process, and start cranking.
I do that too; I put an XXX at any unfinished bits. It's visible, and you can search for it.
That scenario seems pretty unlikely precisely because people would find it annoying. It's not easy to sell stuff. People don't like handing over money unless they feel they are getting something of equal or greater value in return. You can't just slap a price tag -- even a tiny one -- on any old piece of crap and expect that people will buy it. They don't have to buy if they don't want to. If they are nagged for a payment every few minutes or otherwise annoyed, then they will go elsewhere!
Bad analogy. You don't buy gum every few minutes.
A better analogy would be changing the funding structure for highways and roads. Instead of taxes, the funding would come from toll booths erected at every turn in town and every couple of miles on the highway. Would you feel annoyed and "nickle and dimed to death"? I would.
Why is that a better analogy? You've simply described a case where micropayments might not work so well. We can't buy road usage on the free market. No one is proposing that micropayments are the best model for everything.
People won't visit sites where they are going to be (literally) nickel-and-dimed to death.
That's true. People won't visit sites where they will be killed.
I've been saying it for years, and I continue to be right. Micropayments don't solve a customer problem, they solve a provider problem. If you don't solve a customer problem, you don't have a success. Nobody wants to be nickle and dimed to death on the net. It's time to retire this monumentally dumb idea.
The customer problem to be solved is pretty clear: How can I buy some one-off thing on the Internet that has a small monetary value to me? Currently I can't really do that. Have you ever bought a newspaper or some chewing gum? Those are small transactions. No one forced you to buy them. Did you feel nickel-and-dimed to death?
Valis is amazing, but I don't think it'd be the best one for a PKD newbie to start with because it's so whacked out. Scanner Darkly would be a better one I think. If you liked Valis, BTW, you'd find Radio Free Albemuth interesting -- it's another version of his Valis story.
A list of pointers all pointing to the same value? Why don't you tell me why it would be useFUL? I can't see any good reason for using three copies of the same item, when you could just use that item by itself, saving memory...
Here's a scenario off the top of my head: Say you initialize a list of pointers so that each one points to some "default" object. Any list item can be replaced later by a non-default object (through user action or whatever). Suppose the program wants to modify some properties of the default item at some stage. You'd only need to do that in one place if there is a single default object.
I don't think you can unless you have some sort of a loop. The most compact way I could think of was a list comprehension: Seems silly to me that if I want something useless, like three copies of the same list [...] I can save time typing
It's just a bit of syntactic sugar. I can't recall ever seeing it used much. It's not really important. Besides, why do you think such a list would be useless?
Am i one of the only people that actually DOES use VB6 on a regular basis??! (and, although i do live with my parents and can't afford REAL development tools, i'm only 17 so i think it is reasonable!)
:P
You are in luck, my young friend. You can afford Python because it's free, and you are young enough that any brain damage caused by VB will be entirely reversible!
Once you strip away Tycho's rhetoric, he doesn't have much of an argument left. In fact it's not an argument at all, just an assertion: "If you have enough readers who care about your work to go through all that rigamaroll [of signing up for micropayments], you could succeed with any business model." No doubt Tycho knows a thing or two about the webcomics biz; so his opinion counts for something, but still, he doesn't really have anything to back up his assertion. I don't know much about marketing, but I do know that some items are price sensitive. Tycho thinks that if 10 people would pay $1, then 1 person would pay $10. But it's not true in general, and he doesn't tell us why he thinks it's true for webcomics.
I do know a thing or two about reading web comics. I even bought some merchandise once -- I bought some books from achewood.com. (I just mention that to establish that I'm not a complete cheap bastard.) Books and T-shirts are good items for macropayments. But what about other items? Let's stick with Achewood (because it's such a great comic) -- along with their daily free strips they have a weekly, subscription-based color strip. But that requires a macropayment, and I just can't bring myself to make a big payment for comics goddamnit. But I'm sure I'd pay some number of cents for a strip if it were possible. If I liked it, I'd pay again. If I found myself paying every week, I'd buy them in bulk with a subscription. Well fuck me if that isn't an existing business model employed to sells countless other things on this planet! Maybe it would work with comics, maybe it wouldn't. What makes Tycho (and countless Slashdotters) so sure that it wouldn't?
Quote from that link: "The source code for the X Windows System version of Mosaic is available under copyright for individual personal use, for use at an academic institution or for internal business use."
That is not free in an RMS sense. It's not BSD either. BSD is free in the RMS sense.
They do some incredible and brilliant things in the IOCCC, but I can't help thinking that deliberately obfuscating C is immoral somehow :P
What I like about the ICFP is the lack of restrictions. You can solve the problem using whatever tools and programming languages you like, as an individual or as part of a team. I don't think there are any other major programming contests like that.
That's a funny way of classifying things. Your hatred of Windows is sounding really irrational. Not all Unixes have single-machine kernel debugging. Which ones besides Linux do? And SoftICE on Windows allows single machine debugging (or so their blurb says* -- I haven't used it myself). And you didn't account for BeOS (which has a built-in kernel debugger) in your classification scheme ;)
-------------
* "Debug kernel-mode drivers with ease
SoftICE is a powerful kernel mode debugger that supports device driver debugging on either a single or dual machine configuration. "
At home I use w2k as a workstation and FreeBSD as a server and Internet gateway. There are good things in both worlds!
Ok, fair enough. I should have read the article. The kdb debugger is built in to the kernel. My experience with kernel debugging was using a seperate machine over a serial cable. But no one was suggesting adding a GUI to a built-in kernel debugger!
No, no, debuggers can have GUIs. Kernels can't have GUIs. "Windoze" people generally have strange view of computing? Are you really that much of a zealot, or have I just been trolled?
Huh? All he's saying is that GUI-fied debuggers are nicer than console ones. I've used windbg (one of the ones he mentions) to write and debug kernel-mode drivers on NT. I don't see why you think he hasn't a clue. Just because you use Linux doesn't mean you know everything there is to know about computers!