If you a decent level of tecnical ability and are familiar with the man, you know Bill Joy is not stupid.
As for a liar, I don't know him personally nor do you, so I don't think you should accuse him of such a character defect. You might disagree with his views, and he might be incorrect (or correct), but that doesn't mean that he is a liar. I don't know of anything publicacly that he's done for you to accuse him like that.
Yes, I remember a discussion about that issue in www.javalobby.org . The C++ server connected to clients which were in turn applets using CORBA. This worked kind of good but had two major flaws.
1) Firewalls
2) ORB init time (at least orbix)
3) Applet/client code was bloated due to CORBA libraries
The new server still uses an ORB to connect to another server, but to talk to clients it uses serialization through servlets and HTTP. This works really well, but when you think about it, the cost of HTTP socket creation all the time could be bad. It turns out that for our application, it is just as good and we solved the firewall and init issues.
In other words the thread per connection problem is not a big issue for us because we don't have "permanent" connection to our clients.
Another advantage is since we are using servlets, thread pooling is done by the servlet engine (in this case Jigsaw) which takes another big head ache out of our hands. We just code "business" logic and it works fine. This is how I wish development was in my future projects:)
Going back to the servlet/thread issue , I remember a talk at JavaOne about this with a huge game server. You might want to take a look over those slides.
I (and I'm guessing most developers) have no interest in yet another "we implement whatever we want" JVM. It fairly stupid (to put it mildly) to use an VM that is not fully compliant, because it marries you to that implementation and limits you your ability to run in other platforms.
I'm also left wondering about the use of a Java 1.2 capable compiler with a hybrid 1.1/1.2 VM. I must be missing something here...
> I'm wondering if anyone out there uses Java for > something useful. Perhaps I sound naive, but > I've never seen it deployed beyond annoying > applets or slow as molasses games in COS246 > (Java Programming) And java support in Netscape > seems to be lacking to put it nicely.
You need to go out more often:)
We are using Java for rewritting a server from C++ to Java. Why ? Because it's easier to run it in other platforms now (like Linux, and a PC), it's more flexible and has even better performance than the C++ version due to some design changes that are easier to implement and maintain in Java.
> I've done some reading about servlets and > Jserver pages which seem quite interesting. Does > anyone have any experience with them or > experience with gnu-jsp?
Is gnu-jsp still being developed, I tought that project was left in limbo. For servlets I've been doing a lot of that at work too using the Jigsaw webserver, a good little server that is written in Java and it's easy to embed it in your application. I might get stoned for saying this here but, we've also replaced a lot of perl code with java/servlet code and the results are really really interesting. Give it a try...
Any stats. How can I do fast image processing code in perl ? Have any graphic engines written in perl you like to show me ? How about a web server written in perl to compare against the java ones.
Perl has much better text handling functionality(remember what the web is made of ? yeah! text!)
Sure, perl does, perl is a scripting language !!! Scripting languages are made for that type of stuff. Full blown programming languages will always suck at text manglin (Ada, C, C++, Smalltalk, Pascal) compared to scripting languages (perl, awk , etc)
Perl has a much greater base...
Nobody is ramming anything down my throat. I use whatever tool fits the job. For what I do , perl doesn't cut it. For what you do it does. Great !
Please *blah* *blah* Sun stats
Whatever. I think there are more J programmers than what you think, but the user base for a language doesn't mean anything. If it did, then you should be using Visual Basic instead of perl.
It's called an example. I can tell you havent used configure on non Linux/x86 platforms. You know , there *are* other operating systems and architectures out there. I would love for you to show me how configure works in Windows and Mac (and lots of other Unix flavors).
BTW , I didn't say I didn't like C. C is great for a lot of things, but don't tell me how portable C is when I have to deal with the pain of porting C code all the time at work !!!
As for perl, it's nice for "SCRIPTING" but not for writing applications. But that's just my pig headed OO programing opinion, and I can agree that people will have different language tastes.
So many things to pick apart, so little time../configure is beautiful when it works but you are obviously oversimplifying "C" portability../configure or no./configure,compiling accross different platforms is just a mess.
As an example, we tried compiling the TAO ORB for AIX 4.1.5 last weekend. The compile took almost 2 days, failed many times, we modified makefiles, removed features (templates, etc), UPGRADED THE COMPILER AND RUNTIME LIBRARIES, and what did we get at the end ??? nothing. After so much time wasted, the final step of linking everything together just did not work at all.
Java is not magical, but I do develop faster with it not only because I don't have to worry about damned memory pointer magic, core dumps, fake OO (C++), #ifndef #define #endif blocks, etc, but that it really works in the platforms I use at home and at work ***WITHOUT*** having to recompile (AIX, Solaris, Linux, Win98, WinNT).
He mentioned in an interview that he was definetly not happy about George using somebody else for the "ghost" Anakin. I found some comments (not a very a very reliable source IMHO) here :
http://red.affiliation.com/jedicouncil/editorials/ 012399.shtml Again, I'd rather quote the magazine article I read, but this is as close as I get with a reference:)
Sorry, I don't remember the name of the magazine, just that is one of the main Star Wars fan mags (maybe SW Insider?).
> The actor says he's mystified by the movie's >"obsessive fans, and he throws all fan letters > the trash
I have never sent a "fan letter", but I'd imagine that most people who do so are not "obsessive fans".
Like it or not SW is a series that has touched and continues to touch many generations. Obviously, people will want to express their admiration. If he doesn't care or doesn't read them, he doesn't have to tell fans that their admiration is so meaningless to him that it belongs in the trash.
If he felt the dialog was so bad and beneath him, why did he accept the role in the first place ?!? Of course, as usual, money comes before principle. Now that we have a new (and better) Obi-One and he's not needed, he has no problem expressing his real feelings.
What a jerk...
And don't get me started on Prowse. The guy didn't understand that no, he's not considered an actor, just a guy in a suit !!! He could never accept that Darth Vader is the voice, James Earl Jones. He also got pissed because he was replaced. The funny think is, that he's been telling magazines that he would like to play Darth Vader on EP3 when he goes cyborg.
... dubious feature of sometimes being able to have clients and servers on different machines...
*gasp* This statement really surprises me from somebody that has used Un*x for such a long time !
In my University I used this dubious feature to do all my assignments, at work I use it to remotely connect to Linux, Solaris and AIX boxes, at home I use this feature again to connect to work !!!
What are you talking about !?!?!?!?
Sure, running Quake is slower, that's why I have a dual partition and boot up '98 ***to play games***. When I want do real work , I switch back to Linux running good old X.
> I'm not sure what "poor minority" you were > including yourself in above but I know the kids > in my neighborhood.
I'm Hispanic. Born in Panama.
> A lot of the people I live around are 1st > generation immigrants from Central America and > Mexico. In the world the parents grew up in > there was no attitude towards education at all. > It was so far out of the possible that it never > entered their minds. The economies in many of > these countries is pretty close to Medieval > Feudalism.
Hey, I'm from Central America, the Medieval Feudal society !!! Education in Central and South America is *very* important no matter what social class you belong to !!!!!!
Yes, access to the best education is an issue, but I don't remember any poor parent in my country saying "I don't want my kids to have education, forget about that crap!". When you graduate from the University , you are called a "licensiado" and people respect that. I would say their attitude about education is *VERY* good. Some just don't have the means to go to school , or the time (have to work to eat).
I just ***hope*** you're not trying to imply us poor minorities are less interested in technology as "white America" (whatever that is).
> It's also about attitudes toward education and > learning, which frankly is very poor in most > inner city environments, and among certain > cultures within America.
Humm... what are you really saying here ? And what cultures are you talking about ??? I'd like to know what "minority" culture does not have a good attitude toward education and learning. Specially, when in my University, our CS grad program had over 50%-75% total minority students in it. Please be more clear !
Source code isn't a machine. It is a series of instructions. A machine does some thing.
A series of instructions do something, if not, what are the instructions for ????
Source ( or any executable instructions) doesn't do any thing per say it tells the machine (computer) what to do.
I'll say it yet again See my previous definition of machine, or read up on compiler and or automata theory for the proper definition of a machine (ie state machines, etc)
BTW - I'm really getting more and more convinced that we need to start voting "aye" for that CS certification...
A program (source code) is not a machine; it is at best a description of a machine.
You're limiting your scope to a physical "machine" (calculator, etc) to a running program , but the description of a machine is a machine according to CS theory. If not, then how come a finate state automaton is a machine, have you ever seen one of those built (you have in way, but you know what I mean)?.
Don't tell me you never had an exam like this
Build a finate state machine that functions as the coin return mechanism of a soda machine ? Obviously you didn't build a soda machine nor a circuit, but you wrote the machine either in a little diagram or pseudocode or a regular expression, right ? That's the type of machine we're talking about here.
You can give names to machine code instructions (they do have names you know, STORE, LOAD , etc), or you can write it directly with an editor. Also, what do you say about hex code , isn't that another representation of binary numbers more easily understood by humans.
Free speech is meant to safeguard communication between people
Agreed, but even tough many argue that code is for communicating with others, I have yet to see a valid argument against the fact that computer languages are mainly for easier human->computer communication ! (And no, I'm not including pseudocode here, that's another argument)
Anyways, the anonymous response is better than mine, make sure you read it.:P
The only inherit limit to predicting a human response to normal speech is the due to quantum uncertainity. _Theoretically_ we could predict a human response that would be 99.99999999999999% accurate if we just knew enough about the system (brain chemistry, neural structure, etc...) (ie. "given all variables").
I would say it's a much more difficult problem even if you knew "all the variables". I would even argue that this problem is NP complete and thus has no "solution" (like many other problem). Of course, for me to say that would require a mathematical proof, that I don't dare try:)
Likewise, our ability to predict a computers response to speech is also bound by this same uncertainity. Maybe the RAM flipped a bit, a drive sector demagnetized, transistor wouldn't switch, or some other little unexpected, highly unlikely quantum disturbance, and suddenly the computer is no longer predictable. Sure it doesn't happen often, but it's basically the same problem as with humans.
No, when I said computers before I meant computers in the general term, not computers as we know them know. A computer could be anything, made out of silicon, a biological machine (DNA strands ! Has been done:), etc. But in the theoretical sense of a computer, it can be predicted, and it can be done mathematically. Not so with humans, since there are no theoretical humans.:)
Regardless, a computer language, like a human language, is often manipulative, but never deterministic.
Ouch ! This statement is waaaay wrong in some many ways it's not even funny. You're confusing concepts here. Computer languages are deterministic by definition, the failures that you talk about have nothing to do with determinism(if a then b) or non-determinism (if a it is be b or c), those are implementation details irrelevant to this type of discussion at this level. For more information on this refer to a book like "The Dragon Book" (Compilers, Principles, Techniques, and Tools - by Aho, Sethi and Ullman) or any Automata Theory book available in your school.
... does not make natural language a "computer language". Computer languages as usually defined need to be expressed as a non deterministic finate state automata (machine) or a set of regular expressions (which can then build a NDFA).
In addition, and I'll say it again, the main purpose of a computer language is allow humans to build a machine (or set of instructions) for a computer to execute. That's not the main purpose of natural languages, even if the computer understands you (which will require Star Trek levels of AI), the main purpose of natural languages is and will forever be to communicate with humans.
The main purpose of source code is to give the compiler a recipe about how to assemble a binary file...
No, I have to disagree. The main purpose of a computer language (hence source code) is to allow a human to give a set of instructions or build a machine/program to be executed by a computer (whatever a computer is). Having it compiled and converted to binary is implementation dependent, and obviously what happens today, but not really the "main purpose".
You could build a braind dead piece of hardware that read ascii text of some language (let's say BASIC, to keep it simple) and use that language as it's machine code. It'd be stupid, but possible.
As far as the "recipe" angle, I don't know what that buys you. You could call a program a recipe (a set of steps to... / an algorithm) but I call it a (state) machine. How does your recipe idea differ from technical blueprints ??? Are these also free speech ?
Slashdot Mirror
Please show some courtesy.
If you a decent level of tecnical ability and are familiar with the man, you know Bill Joy is not stupid.
As for a liar, I don't know him personally nor do you, so I don't think you should accuse him of such a character defect. You might disagree with his views, and he might be incorrect (or correct), but that doesn't mean that he is a liar. I don't know of anything publicacly that he's done for you to accuse him like that.
Yes, I remember a discussion about that issue in www.javalobby.org . The C++ server connected to clients which were in turn applets using CORBA. This worked kind of good but had two major flaws.
:)
1) Firewalls
2) ORB init time (at least orbix)
3) Applet/client code was bloated due to CORBA libraries
The new server still uses an ORB to connect to another server, but to talk to clients it uses serialization through servlets and HTTP. This works really well, but when you think about it, the cost of HTTP socket creation all the time could be bad. It turns out that for our application, it is just as good and we solved the firewall and init issues.
In other words the thread per connection problem is not a big issue for us because we don't have "permanent" connection to our clients.
Another advantage is since we are using servlets, thread pooling is done by the servlet engine (in this case Jigsaw) which takes another big head ache out of our hands. We just code "business" logic and it works fine. This is how I wish development was in my future projects
Going back to the servlet/thread issue , I remember a talk at JavaOne about this with a huge game server. You might want to take a look over those slides.
I (and I'm guessing most developers) have no interest in yet another "we implement whatever we want" JVM. It fairly stupid (to put it mildly) to use an VM that is not fully compliant, because it marries you to that implementation and limits you your ability to run in other platforms.
...
I'm also left wondering about the use of a Java 1.2 capable compiler with a hybrid 1.1/1.2 VM. I must be missing something here
Hmmm ... maybe like get a clue.
Boy, are you glad you posted anonymously or what ?
> I'm wondering if anyone out there uses Java for
> something useful. Perhaps I sound naive, but
> I've never seen it deployed beyond annoying
> applets or slow as molasses games in COS246
> (Java Programming) And java support in Netscape
> seems to be lacking to put it nicely.
You need to go out more often
We are using Java for rewritting a server from C++ to Java. Why ? Because it's easier to run it in other platforms now (like Linux, and a PC), it's more flexible and has even better performance than the C++ version due to some design changes that are easier to implement and maintain in Java.
> I've done some reading about servlets and
> Jserver pages which seem quite interesting. Does
> anyone have any experience with them or
> experience with gnu-jsp?
Is gnu-jsp still being developed, I tought that project was left in limbo.
For servlets I've been doing a lot of that at work too using the Jigsaw webserver, a good little server that is written in Java and it's easy to embed it in your application.
I might get stoned for saying this here but, we've also replaced a lot of perl code with java/servlet code and the results are really really interesting. Give it a try
Perl is faster than Java on Unix
...
Any stats. How can I do fast image processing code in perl ? Have any graphic engines written in perl you like to show me ? How about a web server written in perl to compare against the java ones.
Perl has much better text handling functionality(remember what the web is made of ? yeah! text!)
Sure, perl does, perl is a scripting language !!! Scripting languages are made for that type of stuff. Full blown programming languages will always suck at text manglin (Ada, C, C++, Smalltalk, Pascal) compared to scripting languages (perl, awk , etc)
Perl has a much greater base
Nobody is ramming anything down my throat. I use whatever tool fits the job. For what I do , perl doesn't cut it. For what you do it does. Great !
Please *blah* *blah* Sun stats
Whatever. I think there are more J programmers than what you think, but the user base for a language doesn't mean anything. If it did, then you should be using Visual Basic instead of perl.
It's called an example. I can tell you havent used configure on non Linux/x86 platforms. You know , there *are* other operating systems and architectures out there. I would love for you to show me how configure works in Windows and Mac (and lots of other Unix flavors).
BTW , I didn't say I didn't like C. C is great for a lot of things, but don't tell me how portable C is when I have to deal with the pain of porting C code all the time at work !!!
As for perl, it's nice for "SCRIPTING" but not for writing applications. But that's just my pig headed OO programing opinion, and I can agree that people will have different language tastes.
Kestrel (1.3) has a Hotspot optimized for client side performance. The last bottleneck to fix is problems with the Java2D rendering pipeline.
Mr. masochist, who writes GUIs and vector graphic rendering engines in Java that perform great !
So many things to pick apart, so little time. ./configure is beautiful when it works but you are obviously oversimplifying "C" portability. ./configure or no ./configure ,compiling accross different platforms is just a mess.
As an example, we tried compiling the TAO ORB for AIX 4.1.5 last weekend. The compile took almost 2 days, failed many times, we modified makefiles, removed features (templates, etc), UPGRADED THE COMPILER AND RUNTIME LIBRARIES, and what did we get at the end ??? nothing. After so much time wasted, the final step of linking everything together just did not work at all.
Java is not magical, but I do develop faster with it not only because I don't have to worry about damned memory pointer magic, core dumps, fake OO (C++), #ifndef #define #endif blocks, etc, but that it really works in the platforms I use at home and at work ***WITHOUT*** having to recompile (AIX, Solaris, Linux, Win98, WinNT).
Any questions ?
He mentioned in an interview that he was definetly not happy about George using somebody else for the "ghost" Anakin.
/ 012399.shtml :)
I found some comments (not a very a very reliable source IMHO) here :
http://red.affiliation.com/jedicouncil/editorials
Again, I'd rather quote the magazine article I read, but this is as close as I get with a reference
Sorry, I don't remember the name of the magazine, just that is one of the main Star Wars fan mags (maybe SW Insider?).
> The actor says he's mystified by the movie's
...
... what a loser.
>"obsessive fans, and he throws all fan letters
> the trash
I have never sent a "fan letter", but I'd imagine that most people who do so are not "obsessive fans".
Like it or not SW is a series that has touched and continues to touch many generations. Obviously, people will want to express their admiration. If he doesn't care or doesn't read them, he doesn't have to tell fans that their admiration is so meaningless to him that it belongs in the trash.
If he felt the dialog was so bad and beneath him, why did he accept the role in the first place ?!? Of course, as usual, money comes before principle. Now that we have a new (and better) Obi-One and he's not needed, he has no problem expressing his real feelings.
What a jerk
And don't get me started on Prowse. The guy didn't understand that no, he's not considered an actor, just a guy in a suit !!! He could never accept that Darth Vader is the voice, James Earl Jones. He also got pissed because he was replaced. The funny think is, that he's been telling magazines that he would like to play Darth Vader on EP3 when he goes cyborg.
Yeah right
*gasp* This statement really surprises me from somebody that has used Un*x for such a long time !
In my University I used this dubious feature to do all my assignments, at work I use it to remotely connect to Linux, Solaris and AIX boxes, at home I use this feature again to connect to work !!!
What are you talking about !?!?!?!?
Sure, running Quake is slower, that's why I have a dual partition and boot up '98 ***to play games***. When I want do real work , I switch back to Linux running good old X.
> I'm not sure what "poor minority" you were
> including yourself in above but I know the kids
> in my neighborhood.
I'm Hispanic. Born in Panama.
> A lot of the people I live around are 1st
> generation immigrants from Central America and
> Mexico. In the world the parents grew up in
> there was no attitude towards education at all.
> It was so far out of the possible that it never
> entered their minds. The economies in many of
> these countries is pretty close to Medieval
> Feudalism.
Hey, I'm from Central America, the Medieval Feudal society !!! Education in Central and South America is *very* important no matter what social class you belong to !!!!!!
Yes, access to the best education is an issue, but I don't remember any poor parent in my country saying "I don't want my kids to have education, forget about that crap!". When you graduate from the University , you are called a "licensiado" and people respect that. I would say their attitude about education is *VERY* good. Some just don't have the means to go to school , or the time (have to work to eat).
I just ***hope*** you're not trying to imply us poor minorities are less interested in technology as "white America" (whatever that is).
... what are you really saying here ? And what cultures are you talking about ??? I'd like to know what "minority" culture does not have a good attitude toward education and learning. Specially, when in my University, our CS grad program had over 50%-75% total minority students in it. Please be more clear !
> It's also about attitudes toward education and
> learning, which frankly is very poor in most
> inner city environments, and among certain
> cultures within America.
Humm
1)I'm not talking about a legal definition, where did I say anything about that??? read the posts.
2)It would be nice if you made your arguments with maybe some references or technical detail.
BTW - A Turing machine is not a machine ???
See my previous reply to the other AC.
...
Source code isn't a machine. It is a series of instructions. A machine does some thing.
A series of instructions do something, if not, what are the instructions for ????
Source ( or any executable instructions) doesn't do any thing per say it tells the machine (computer) what to do.
I'll say it yet again
See my previous definition of machine, or read up on compiler and or automata theory for the proper definition of a machine (ie state machines, etc)
BTW - I'm really getting more and more convinced that we need to start voting "aye" for that CS certification
A program (source code) is not a machine; it is at best a description of a machine.
You're limiting your scope to a physical "machine" (calculator, etc) to a running program , but the description of a machine is a machine according to CS theory. If not, then how come a finate state automaton is a machine, have you ever seen one of those built (you have in way, but you know what I mean)?.
Don't tell me you never had an exam like this
Build a finate state machine that functions as the coin return mechanism of a soda machine ?
Obviously you didn't build a soda machine nor a circuit, but you wrote the machine either in a little diagram or pseudocode or a regular expression, right ? That's the type of machine we're talking about here.
Seems to be on topic. Yet another case of moderators putting their opinions before principle !
You can give names to machine code instructions (they do have names you know, STORE, LOAD , etc), or you can write it directly with an editor. Also, what do you say about hex code , isn't that another representation of binary numbers more easily understood by humans.
:P
Free speech is meant to safeguard communication between people
Agreed, but even tough many argue that code is for communicating with others, I have yet to see a valid argument against the fact that computer languages are mainly for easier human->computer communication ! (And no, I'm not including pseudocode here, that's another argument)
Anyways, the anonymous response is better than mine, make sure you read it.
*LOL*. You forgot to type in the sound of the police beating you up before taking you away !
BTW I love that signature "Free the Mallocs" , he hehe
The only inherit limit to predicting a human response to normal speech is the due to quantum uncertainity. _Theoretically_ we could predict a human response that would be 99.99999999999999% accurate if we just knew enough about the system (brain chemistry, neural structure, etc...) (ie. "given all variables").
:)
:), etc. But in the theoretical sense of a computer, it can be predicted, and it can be done mathematically. Not so with humans, since there are no theoretical humans. :)
I would say it's a much more difficult problem even if you knew "all the variables". I would even argue that this problem is NP complete and thus has no "solution" (like many other problem). Of course, for me to say that would require a mathematical proof, that I don't dare try
Likewise, our ability to predict a computers response to speech is also bound by this same uncertainity. Maybe the RAM flipped a bit, a drive sector demagnetized, transistor wouldn't switch, or some other little unexpected, highly unlikely quantum disturbance, and suddenly the computer is no longer predictable. Sure it doesn't happen often, but it's basically the same problem as with humans.
No, when I said computers before I meant computers in the general term, not computers as we know them know. A computer could be anything, made out of silicon, a biological machine (DNA strands ! Has been done
Regardless, a computer language, like a human language, is often manipulative, but never deterministic.
Ouch ! This statement is waaaay wrong in some many ways it's not even funny. You're confusing concepts here. Computer languages are deterministic by definition, the failures that you talk about have nothing to do with determinism(if a then b) or non-determinism (if a it is be b or c), those are implementation details irrelevant to this type of discussion at this level. For more information on this refer to a book like "The Dragon Book" (Compilers, Principles, Techniques, and Tools - by Aho, Sethi and Ullman) or any Automata Theory book available in your school.
... does not make natural language a "computer language". Computer languages as usually defined need to be expressed as a non deterministic finate state automata (machine) or a set of regular expressions (which can then build a NDFA).
In addition, and I'll say it again, the main purpose of a computer language is allow humans to build a machine (or set of instructions) for a computer to execute. That's not the main purpose of natural languages, even if the computer understands you (which will require Star Trek levels of AI), the main purpose of natural languages is and will forever be to communicate with humans.
Great post. It seems some of us are debating the same thing the judges are. This is a very complicated issue.
The main purpose of source code is to give the compiler a recipe about how to assemble a binary file ...
... / an algorithm) but I call it a (state) machine. How does your recipe idea differ from technical blueprints ??? Are these also free speech ?
No, I have to disagree. The main purpose of a computer language (hence source code) is to allow a human to give a set of instructions or build a machine/program to be executed by a computer (whatever a computer is). Having it compiled and converted to binary is implementation dependent, and obviously what happens today, but not really the "main purpose".
You could build a braind dead piece of hardware that read ascii text of some language (let's say BASIC, to keep it simple) and use that language as it's machine code. It'd be stupid, but possible.
As far as the "recipe" angle, I don't know what that buys you. You could call a program a recipe (a set of steps to
A program is a machine, it has nothing to do with a computer nor wheter it's compiled or not. See my response to "Feh!".