Actually, lemme amend that: it does displace the atmosphere and lower the oxygen to where it's unbreathable. This it does to greater effect where it hits the flame, it for sure doesn't "suck all the atmosphere out". If you get exposed to a "halon rich" atmosphere, you've got enough residual oxygen from holding your breath to be just fine in order to get out. The smoke would probably kill you right good anyway if not for the halon. It's definitely not like stepping into a vacuum.
In order to be effective of course, it has to be released in a closed environment, and combined with the confusion that comes with a fire, makes it hard to escape whilst the oxygen is so low.
Still, it was banned because of the ozone depletion problem. There might be a few safety hazards associated with it, but fire is kind of a bigger hazard...
Honestly, do python apologists cut and paste this argument every time they see it? I use python, I hated the whitespace thing, I got used to it. But this still irks the hell out of me. It might help to get out of the damn bunker and stop "defending" this Pure And Beautiful Thing Of Absolute Perfection That Is Beyond All Question against the unwashed hordes that don't appreciate its wonderful elegance and all that rot. "you're one of the people who blahblahblah can't write clean code blahblahblah protects you from yourself blahblahblah"... patronizing gets you nowhere.
Trying to understand the people you're evangalizing to can go a long way.
> How exactly does that "eliminate the anonymous cash transaction?"
I'm not actually with the conspiracy theorists here, but:
1. Walk past RFID sensors scanning for currency. 2. Camera snaps your picture, complete with RFID #'s of bills you're carrying.
There doesn't need to be a step #3 now, does there? Know what happens to people carrying large amounts of cash in the USA? They get stopped and searched and treated as drug dealing suspects. Their cash is usually confiscated, and they then have to prove that it's clean in order to get it back.
I suppose I'll have to work on a fanny pack with a faraday cage in it now...
> You got it pretty close for somebody who (I'm assuming) doesn't use it and hasn't studied it much.
Written a bit of C# for kicks, wrote a vb.net mode in emacs (alas it's gone so don't ask me for it), played with webmatrix, but i can't say i've done anything serious with it. The whitepapers that I never got to reading much of gave me the impression that CLR was more of a calling convention, linker format, and type conventions that worked across languages, but that it could be in the realm of either IL or native code. OLE3 I guess, and that the IL was actually a subsystem of it, so that the CLR would underpin everything, the VM ran on top of that, and web services (the most dubious of "technologies" as opposed to marketing hype) would be implemented on top of that.
Maybe.NET is like the Tao. The.NET that can be named is not the true.NET...
Anyway, I learned what I did largely because I was very excited about the possibility of languages like haskell and ocaml running on it. What we're getting now is Mondrian and F#, which are so stripped down as to be useless compared to their capable parent languages. At one time I was defending.NET against uninformed detractors, but I'm afraid that it's just going to end up as web services, glorified heavyweight RPC, and that all sign of innovation that really makes a difference to any other programming approach will be ruthlessly crushed. And that goes for Java too.
I used to tell people who put on airs about computers at the same time they were asking for help "you need to upgrade your LRF subsystem", or "how many LRF's do you have?" or "looks like the LRF needs adjustment". And they'd nod accordingly.
This is the second time I've seen the suggestion of printing a receipt for the voter that allows access to their vote, and it belies some really basic misunderstanding about the nature of secret ballots. In a proper secret ballot, you cannot prove how you voted. Only that you voted, and then only until the polls close. This is necessary to deter voter manipulation through incentive or intimidation. It can't eliminate it, but it does remove the important element of provability.
Not even a ballot with a serial number. Same idea as numbered ballots. A hash is meaningless, a decent machine can zip through millions of them in minutes.
The local precinct doesn't need a third copy, as long as they can't unlock the box that the second copy goes into. If they need an informal total, they can use the electronic report -- hopefully they'd secure the damn machine a little better than those at least...
I was looking for the familiar site, but the project appears to be dead. It was never that far along anyway. I think PHP is more likely to choose interoperability using SOAP and DIME than in compiling to MSIL.
Call it what you want, but have you used it? There is nothing in deployment descriptors to state a JAAS context. Know where it's kept? server.xml. That's tomcat boys and girls, not J2EE.
Well, not quite so fast, you can put it in the deployment descriptor in Sun ONE. It goes in "sun-application.xml". Gee that looks kind of vendor-specific, doesn't it? Hope you weren't planning on running that.ear on anything but Sun ONE, unless you wrote the dd's for all the vendors you might deploy on.
Maybe I should resurrect my old.sig about the word "FUD"...
.NET is the new ActiveX. ActiveX by itself was this nebulous definition, but what it boiled down to was nothing more than COM..NET boils down to three things behind the marketing umbrella name:
* the.NET Virtual Machine: Basically the same idea as other bytecode compiled languages, like UCSD Pascal (ooh you thought I was going to say JVM, well sun didn't invent the idea). Write once, deploy anywhere where windows (or mono) is. It has some features not seen in JVM's, like cached JIT code, so it doesn't have to rerun the JIT every single time you run the app.
* The.NET Common Language Runtime, including the system library: This is intended to replace the Win32 API with something as easy to use as most Visual BASIC libs, getting rid of HWNDS and HRESULTS and __farcall lpzsFoobletch and so on.
* Web services: Really just the first application of the first two, but Microsoft is plugging this SOAP-based stuff like the second coming. I somehow don't see it replacing RPC for communication with system services, but there it is.
> I'm not sure how many times I've seen this single point refutued, but your not tied to a single language to use the the JVM. Want proof, here you go. That's COBOL to Eifel with all the good bits in the middle.
Sure, as long as you don't mind the the language either lacking some basic features, or having to use slow kludgy workarounds to get said features, because all the JVM can understand is the java object model. Try implementing a language with tailcall elimination in the JVM. Or multiple dispatch.
.NET and php are orthogonal. There's one effort underway to port PHP to.NET, for one. I recommend getting some understanding of the ASP.NET architecture before making statements like this, because it's like saying "php is better than fastcgi" (considering you can run php as a fastcgi).
The main problem I have with PHP is that it's not OO. Objects are syntactic sugar for grouping functions, but objects are by default copied by value, and worse yet, always compared by value, not by identity (so when $a === $b at one point, it might not later, even with the same objects, because they got the implementation so wrong). PHP5 is supposed to fix that, though things like its error handling still leave much to be desired (try eval'ing code with a syntax error -- your script will die, and you can NOT stop it. sort of defeats the purpose of eval, don't it?)
But that's all off the topic of.NET, which is a platform, whereas PHP is simply a language.
Assuming you don't mind mechanisms that aren't standardized and thus vendor-specific across all implementations, like oh, authentication. There's roles. There's role mappings. There's even a standard form name and fields for passwords. But there's nothing in J2EE for something as simple as mapping a password to a user. Let alone something like PAM or JSSE. Not even an XML file. Not even a function. The reference implementation? Tomcat-specific.
I just picked up a largish complex perl program I wrote two years ago, and even though I'd written very little perl since (I wrote python and java) and it was largely uncommented, I still understood all of it just fine.
Granted, perl gives people all kinds of ways to be "clever", so reading someone else's code can be a nightmare. Given the ease of operator overloading and creating metaclasses in python, it's quite possible to create python code that looks perfectly readable on the surface, that no one can actually really understand. Then there's Ruby, with continuations... those let you create control structures that twist your brain in knots. Let's not even get into the evil twistedness you can perform with templates in C++, or lazy patterns in haskell.
If you're overcome by mere syntax decoration on variables, you're far too much of a lightweight to be programming.
You are not an affiliate. Slashdot is. I'll give them this, slashdot wears the conflict of interest on its sleeve, as they've stated since they began doing reviews how most reviews were going to be glowing because of the affiliation.
One might imagine that a little integrity would spur more buying of books that were well-reviewed, because the review would mean something, but apparently for now it's worth just getting mentioned on slashdot.
Bleh, I can't get into the habit of rewriting my terms when my language instincts (and I am writing in a computer language, meant for humans) tell me to put the thing being checked on the left.
any decent compiler will warn you about unintended assignment, and will give you an option to turn warnings into errors. Frankly I thought using = instead of:= was just silly, but there it is, so at just let the compiler get at it.
Also, one should be using const wherever they can. It helps the compiler out too, which means potentially faster code.
Know what'd be nice, is if whatever program they used could automatically generate asserts and insert them into the source.
> My 11 year old's computer routinely spawns 58 popups until the whole thing dies while I exercise the futility of trying to keep her from reinfecting her computer with this garbage.
You can run ad-aware on bootup or on a schedule, from the commandline, and automatically wipe said crap off the system. I imagine Spybot Search and Destroy also has similar functionality. Given the usage pattern, you probably want to invest in a full-blown virus scanner while you're at it.
If your 11yo won't let you, then screw it, it's her computer, just make sure your own computer is protected...
The "ASP Loophole" has nothing to do with generated code. It has to do with the notion that someone could, for example, take the SourceForge code, fork it, run an entire enterprise on it that the public uses, make bank from this software, and not have to open up a single line because technically they never distributed it (Sourceforge.net doesn't technically count since they're the originators, I'm just using sourceforge as a well-known ASP app).
GPL3 would consider the program being deployed in some significant "public facing" way as being the same as distribution. This is of course not perfectly defineable, just as "distribution" isn't, so it's still going to be rich debate fodder for the amateur lawyers on slashdot and usenet tho, I guarantee that:)
Slashdot Mirror
Actually, lemme amend that: it does displace the atmosphere and lower the oxygen to where it's unbreathable. This it does to greater effect where it hits the flame, it for sure doesn't "suck all the atmosphere out". If you get exposed to a "halon rich" atmosphere, you've got enough residual oxygen from holding your breath to be just fine in order to get out. The smoke would probably kill you right good anyway if not for the halon. It's definitely not like stepping into a vacuum.
In order to be effective of course, it has to be released in a closed environment, and combined with the confusion that comes with a fire, makes it hard to escape whilst the oxygen is so low.
Still, it was banned because of the ozone depletion problem. There might be a few safety hazards associated with it, but fire is kind of a bigger hazard...
> Halon displaces the oxygen in the environment it is released into.
No it doesn't. What it does do is decompose into some truly nasty chemicals that will irritate the hell out of your lungs and possibly kill you.
What it also does is eat the ozone layer like mad. It's like opening up a thousand old-fashioned air conditioners at once. That is why it was banned.
> Ok, have you actually tried python?
... patronizing gets you nowhere.
Honestly, do python apologists cut and paste this argument every time they see it? I use python, I hated the whitespace thing, I got used to it. But this still irks the hell out of me. It might help to get out of the damn bunker and stop "defending" this Pure And Beautiful Thing Of Absolute Perfection That Is Beyond All Question against the unwashed hordes that don't appreciate its wonderful elegance and all that rot. "you're one of the people who blahblahblah can't write clean code blahblahblah protects you from yourself blahblahblah"
Trying to understand the people you're evangalizing to can go a long way.
> I have never heard of this beach before. Where can I find it? Is it a GNUde beach?
I'm picturing RMS on the GNUde beach. Thanks, I'm gonna sleep lousy tonight...
> How exactly does that "eliminate the anonymous cash transaction?"
I'm not actually with the conspiracy theorists here, but:
1. Walk past RFID sensors scanning for currency.
2. Camera snaps your picture, complete with RFID #'s of bills you're carrying.
There doesn't need to be a step #3 now, does there? Know what happens to people carrying large amounts of cash in the USA? They get stopped and searched and treated as drug dealing suspects. Their cash is usually confiscated, and they then have to prove that it's clean in order to get it back.
I suppose I'll have to work on a fanny pack with a faraday cage in it now...
> You got it pretty close for somebody who (I'm assuming) doesn't use it and hasn't studied it much.
.NET is like the Tao. The .NET that can be named is not the true .NET...
.NET against uninformed detractors, but I'm afraid that it's just going to end up as web services, glorified heavyweight RPC, and that all sign of innovation that really makes a difference to any other programming approach will be ruthlessly crushed. And that goes for Java too.
Written a bit of C# for kicks, wrote a vb.net mode in emacs (alas it's gone so don't ask me for it), played with webmatrix, but i can't say i've done anything serious with it. The whitepapers that I never got to reading much of gave me the impression that CLR was more of a calling convention, linker format, and type conventions that worked across languages, but that it could be in the realm of either IL or native code. OLE3 I guess, and that the IL was actually a subsystem of it, so that the CLR would underpin everything, the VM ran on top of that, and web services (the most dubious of "technologies" as opposed to marketing hype) would be implemented on top of that.
Maybe
Anyway, I learned what I did largely because I was very excited about the possibility of languages like haskell and ocaml running on it. What we're getting now is Mondrian and F#, which are so stripped down as to be useless compared to their capable parent languages. At one time I was defending
One of my favorite acronyms comes from an IBM mainframe manual.
FAMD - Forced Air Movement Device
leave it to IBM to come up with a four-letter acronym for a three-letter word.
I used to tell people who put on airs about computers at the same time they were asking for help "you need to upgrade your LRF subsystem", or "how many LRF's do you have?" or "looks like the LRF needs adjustment". And they'd nod accordingly.
LRF stands for "Little Rubber Feet".
This is the second time I've seen the suggestion of printing a receipt for the voter that allows access to their vote, and it belies some really basic misunderstanding about the nature of secret ballots. In a proper secret ballot, you cannot prove how you voted. Only that you voted, and then only until the polls close. This is necessary to deter voter manipulation through incentive or intimidation. It can't eliminate it, but it does remove the important element of provability.
Not even a ballot with a serial number. Same idea as numbered ballots. A hash is meaningless, a decent machine can zip through millions of them in minutes.
The local precinct doesn't need a third copy, as long as they can't unlock the box that the second copy goes into. If they need an informal total, they can use the electronic report -- hopefully they'd secure the damn machine a little better than those at least...
I was looking for the familiar site, but the project appears to be dead. It was never that far along anyway. I think PHP is more likely to choose interoperability using SOAP and DIME than in compiling to MSIL.
I call FUD!!!
.ear on anything but Sun ONE, unless you wrote the dd's for all the vendors you might deploy on.
.sig about the word "FUD"...
JAAS has been a standard in J2EE for ages now.
Call it what you want, but have you used it? There is nothing in deployment descriptors to state a JAAS context. Know where it's kept? server.xml. That's tomcat boys and girls, not J2EE.
Well, not quite so fast, you can put it in the deployment descriptor in Sun ONE. It goes in "sun-application.xml". Gee that looks kind of vendor-specific, doesn't it? Hope you weren't planning on running that
Maybe I should resurrect my old
.NET is the new ActiveX. ActiveX by itself was this nebulous definition, but what it boiled down to was nothing more than COM. .NET boils down to three things behind the marketing umbrella name:
.NET Virtual Machine: Basically the same idea as other bytecode compiled languages, like UCSD Pascal (ooh you thought I was going to say JVM, well sun didn't invent the idea). Write once, deploy anywhere where windows (or mono) is. It has some features not seen in JVM's, like cached JIT code, so it doesn't have to rerun the JIT every single time you run the app.
.NET Common Language Runtime, including the system library: This is intended to replace the Win32 API with something as easy to use as most Visual BASIC libs, getting rid of HWNDS and HRESULTS and __farcall lpzsFoobletch and so on.
* the
* The
* Web services: Really just the first application of the first two, but Microsoft is plugging this SOAP-based stuff like the second coming. I somehow don't see it replacing RPC for communication with system services, but there it is.
> I'm not sure how many times I've seen this single point refutued, but your not tied to a single language to use the the JVM. Want proof, here you go. That's COBOL to Eifel with all the good bits in the middle.
Sure, as long as you don't mind the the language either lacking some basic features, or having to use slow kludgy workarounds to get said features, because all the JVM can understand is the java object model. Try implementing a language with tailcall elimination in the JVM. Or multiple dispatch.
.NET and php are orthogonal. There's one effort underway to port PHP to .NET, for one. I recommend getting some understanding of the ASP.NET architecture before making statements like this, because it's like saying "php is better than fastcgi" (considering you can run php as a fastcgi).
.NET, which is a platform, whereas PHP is simply a language.
The main problem I have with PHP is that it's not OO. Objects are syntactic sugar for grouping functions, but objects are by default copied by value, and worse yet, always compared by value, not by identity (so when $a === $b at one point, it might not later, even with the same objects, because they got the implementation so wrong). PHP5 is supposed to fix that, though things like its error handling still leave much to be desired (try eval'ing code with a syntax error -- your script will die, and you can NOT stop it. sort of defeats the purpose of eval, don't it?)
But that's all off the topic of
> Has anybody worked out what it is yet?
Have you bothered reading any documentation? They let you read the books in the store at Barnes & Noble and Borders, yunno...
> J2EE is a standard
Assuming you don't mind mechanisms that aren't standardized and thus vendor-specific across all implementations, like oh, authentication. There's roles. There's role mappings. There's even a standard form name and fields for passwords. But there's nothing in J2EE for something as simple as mapping a password to a user. Let alone something like PAM or JSSE. Not even an XML file. Not even a function. The reference implementation? Tomcat-specific.
What a bloody joke.
"Insightful". Huh.
... those let you create control structures that twist your brain in knots. Let's not even get into the evil twistedness you can perform with templates in C++, or lazy patterns in haskell.
I just picked up a largish complex perl program I wrote two years ago, and even though I'd written very little perl since (I wrote python and java) and it was largely uncommented, I still understood all of it just fine.
Granted, perl gives people all kinds of ways to be "clever", so reading someone else's code can be a nightmare. Given the ease of operator overloading and creating metaclasses in python, it's quite possible to create python code that looks perfectly readable on the surface, that no one can actually really understand. Then there's Ruby, with continuations
If you're overcome by mere syntax decoration on variables, you're far too much of a lightweight to be programming.
> Or even better yet: Good book.
Book is implied, it's a book review. So the review can boil down to: Good.
You are not an affiliate. Slashdot is. I'll give them this, slashdot wears the conflict of interest on its sleeve, as they've stated since they began doing reviews how most reviews were going to be glowing because of the affiliation.
One might imagine that a little integrity would spur more buying of books that were well-reviewed, because the review would mean something, but apparently for now it's worth just getting mentioned on slashdot.
Slashdot used you.
Bleh, I can't get into the habit of rewriting my terms when my language instincts (and I am writing in a computer language, meant for humans) tell me to put the thing being checked on the left.
:= was just silly, but there it is, so at just let the compiler get at it.
any decent compiler will warn you about unintended assignment, and will give you an option to turn warnings into errors. Frankly I thought using = instead of
Also, one should be using const wherever they can. It helps the compiler out too, which means potentially faster code.
Know what'd be nice, is if whatever program they used could automatically generate asserts and insert them into the source.
> My 11 year old's computer routinely spawns 58 popups until the whole thing dies while I exercise the futility of trying to keep her from reinfecting her computer with this garbage.
You can run ad-aware on bootup or on a schedule, from the commandline, and automatically wipe said crap off the system. I imagine Spybot Search and Destroy also has similar functionality. Given the usage pattern, you probably want to invest in a full-blown virus scanner while you're at it.
If your 11yo won't let you, then screw it, it's her computer, just make sure your own computer is protected...
I dunno, I feel as though I got something useful out of this article. I learned a new word, "smelging" for one.
The "ASP Loophole" has nothing to do with generated code. It has to do with the notion that someone could, for example, take the SourceForge code, fork it, run an entire enterprise on it that the public uses, make bank from this software, and not have to open up a single line because technically they never distributed it (Sourceforge.net doesn't technically count since they're the originators, I'm just using sourceforge as a well-known ASP app).
:)
GPL3 would consider the program being deployed in some significant "public facing" way as being the same as distribution. This is of course not perfectly defineable, just as "distribution" isn't, so it's still going to be rich debate fodder for the amateur lawyers on slashdot and usenet tho, I guarantee that
Lord in heaven... Do you ever get laid?
(Many Highways were required by law to have a certain amount of space that could be used as a landing strip)
<trebek>Ohh, I'm sorry, that's incorrect</trebek>. Check here for the answer