I fail to see how you could pick out loops as a unifying concept in CS. Honestly, if loops are what you spend your time thinking about when programming, I would hate to have to clean up after you.
The assertion that operating systems are irrelevant is very typical of someone who has no understanding of them. I accept that most programs can be written for most oses - I never said that it couldn't. But to beleive that operating systems are not even a part of computer science, should not be taught, is so breathtakingly ignorant its beyond description.
Computer Science has nothing to do with operating systems
A good CS course has quite a few operating systems courses. Someone who has only ever used Windows is likely to be utterly clueless as to how operating systems work : The workings are hidden from you, and it is riddled with frankly horrible backwards compatibility hacks. There are around 5000 syscalls, compared to less than 250 for any *nix. Most CS labs now are run on Linux (some are BSD), because it is possible to understand what is going on. Components are replaceable, boundaries are sharp. Experimentation is easy. Minix is still used when even greater simplicity is required : hacking on the VM of BSD or Linux is kind of wearing for an undergrad.
Maybe you like the idea of signing NDAs just to get an education. Most people don't - and that is why almost no serious CS in universities is done using MS technology - the only ones I know of are funded by MS (eg Cambridge). "Our way is the only way" is not really a rallying cry in universities the way it is among some corporate boardrooms.
CS is a state of mind which bases itself in logic, especially loops.
Fuck are you clueless. Please be honest, you have never been near a decent CS department, have you? Loops, for crying out loud.
Lets take it as given that the entire purpose of the SCO lawsuits is to discredit Linux/Free Software and spread FUD, bankrolled by...erm... someone.
Enderle has constantly harped on about "these zealots", "death threats", "die for what you believe in" etc, etc.
Enderle is setting himself up to either:
Get killed by goons hired by... someone
Kill himself ( some people really are zealots)
Most likely, fake his own death
in order to discredit Linux "zealots"
He'll be found in a Reno hotel room with the full text of the GPL version 2 (or later at your discretion) inscribed on his chest using a slide rule. A single penguin feather will be protruding from between his mutilated butt cheeks, and the smell of herring will linger.
Who would use software that causes people to kill in such a biblical fashion? Thats how it is going to play - Just wait until the mainstream media condenses down John Katz' old columbine mooning into a blanket condemnation of all geekhood.
I think it could happen just when... someone feels that they are getting almost no return from the SCO lawsuit in terms of FUD, and they need a real boost. Its desparate, but the potential pay off is huge - better than patents!
Do I really beleive this? No. But its worth putting it out there, just so if it did happen we can all pat each other on the back in Gitmo, and say "Knew it was coming! Why didn't we listen?
So use the slow casting method if the vm ever notices a boxed instance being put in the collection, ie either a) The boxed instance was not created by code immediately before insertion (therefore it was not auto boxing in the compiler) b) The boxed instance had another reference in the same frame ( this really is pretty simple).
Not really a show stopper... given that I doubt very many people are going to be doing that.
Of course assembler is useful. It is shocking that a large number of CS courses don't teach it at all. I'm even more surprised at the number of CS students who have no clue about any area of the subject they have a degree in.
I do use inline assembler... as rarely as I can. Only when the compiler is being stupid. C has its place too - operating systems courses.
But to believe that it is a good idea to be teaching algorithms or fundamental concepts in assembler is utterly braindead. Functional languages are just so much better for learning this kind of stuff. And this is the stuff that computer *SCIENCE* is about.
What I find amazing is the number of fools who think that writing assembler makes them a programming god. The fact is that most of the time, you are far better off, optimisation wise, looking at your algorithms and data structures than messing at that level. Only if you know a small bit of code is messing you over, AND the compiler is doing something wrong, should you resort to assembler.
The other area where it is necessary is when you are working on a compiler (which I've done), a virtual machine or a dynamically translating emulator( which I've done.)
My post was a reaction to the grandparents assertion that all code in CS labs should be in "cross platform C". Thats totally fucking wrongheaded.
Could you please explain why a JVM could not notice that you are autoboxing a primitive type into a parameterised collection, and then just store the unboxed primitive efficiently, with a type specialised piece of code, as Pizza or.Net would?
I can't really see why this would be impossible or even hard... and the Java way does keep the class format the same.
I thought monotone, codeville, and darcs all used the distributed repository model as well as arch & bk. They may be a little further behind in terms of features or surrounding tools, but each one does have some interesting theory/philosophy of version control behind it.
And darcs is written in haskell, so it wins points for enjoying the soundness and showing once again that pure FP can be and is used in the "real world"...
I wouldn't discount any of them yet, but I agree that the subversion fanboys are pretty damn irritating, trying to get every project to switch away from CVS now, when it would clearly be better to wait and see how some of the more revolutionary free systems evolve.
This is of course totally non-binding and non-enforceable. When you create something ( the source of your program), you hold the copyright o f that something and can licence it as you wish.
I want a new law, which puts the same legal penalties on a lawyer for writing false legal notices ( known to be bullshit or unenforceable) as false legal advice : possible disbarment and malpractice lawsuits.
That is the exact same feature. Its really pretty standard on everything but IE.
Re:What, no editorial?
on
Red Hat Recap
·
· Score: 5, Informative
No, the updates that you were referring to are *still* derivative works of a GPL'ed work. Thus, Redhat cannot provide them to you with additional restrictions. Once you have downloaded the RPMS, you can do what you like with them.
What Redhat can and do restrict is the access to their download servers. They are not required to let you use their bandwidth.... and that is what you are paying for, along with support.
Sun was able to kick Microsoft completely out of the Java business.
Wow, that worked really well for Sun didn't it? Have you not noticed that all of the Microsofties are going around screaming about.Net and C# all the time now? Its hard to convince these people that Java has any advantage once you realise that
Giving limitless amounts of money to MS is fine with them.
They honestly beleive that visual studio.Net is the best IDE ever ( because they only compare it to the last version..)
They feel that it is good that their code only works on Windows.
Sun prompting MS to produce a Java clone/alternative was a bad idea.... you know, the whole keep your enemies closer thing.
You have absolutely no fucking clue if you think that Konsole or gnome-terminal embed bash or sh. They are terminal programs! They talk to shells via a pty! "bash mode" indeed....
It almost always seems to be an AGP chipset issue with nvidias binary crapware. Since I changed mother boards from an oldish VIA chipset to a newish one, I've not had any lockups from the binary drivers.
Doesn't it ever get just a tiny bit boring trolling for Microsoft? KDE do *NOT* release binaries.
Lets put it this way - a patch was released that allowed any user to fix their system, in 90 minutes, by KDE. The competitive market place of Linux distros led to some vendors offering this fix in less than a day.
The same was only done by Microsoft in two weeks.
How the *fuck* can you try to spin that Microsofts way?
Or maybe you could ask them why the fuck they can't write a permissable binary kernel module like every body else, but actually patch the kernel instead?
Trolltech are mistaken in that statement.
They have released QT/E under the GPL, which means an aggregate product can be commercial, as long as it is distributed under the GPL as well.
IE, you can't make proprietary software without paying Trolltech. Commercial software is fine.
But when will they actually release an updated SVG Viewer?
The currently released version is just wbout 2 years old. The preview of version 6 is better, but won't get installed anywhere until its actually released officialy and bundled with Acrobat reader.
And corporate muppets won't roll out preview releases.
I have to wonder how commited Adobe is to SVG. Their preview release of ASV6 is good enough to discourage competitors (it would take quite an effort to match it in a ~2 year timeframe), but won't get installed anywhere until release. Are they scared that SVG will eat PDF as well as Flash?
The licencing of your cell project is inconsistent:
doc/licence says LGPL. But it also has a vague preamble that adds additional terms to the LGPL, namely that derivative works authors must distribute the work back to you ( and thereby release the source to you too.) The plain LGPL does not require this.
readme says that it is a "commercial open-source" project that can be licenced for $99. This is confusing.
It would be good if this could be cleared up. I would recommend that you make it LGPL. Ask a lawyer how to add your extra restriction if you really want it. Be aware that this will make it incompatible with the GPL, and the plain LGPL.
If you also want to offer commercial licencing under different terms, make the dual licence nature clear. However, going this route, the GPL is a better bet as the free licence to get commercial users to cough up. There is very little that a commercial project can't do with an LGPL'ed library.
Slashdot Mirror
I fail to see how you could pick out loops as a unifying concept in CS. Honestly, if loops are what you spend your time thinking about when programming, I would hate to have to clean up after you.
The assertion that operating systems are irrelevant is very typical of someone who has no understanding of them. I accept that most programs can be written for most oses - I never said that it couldn't. But to beleive that operating systems are not even a part of computer science, should not be taught, is so breathtakingly ignorant its beyond description.
Are you talking about BBC america?
US != the world
Computer Science has nothing to do with operating systems
A good CS course has quite a few operating systems courses. Someone who has only ever used Windows is likely to be utterly clueless as to how operating systems work : The workings are hidden from you, and it is riddled with frankly horrible backwards compatibility hacks. There are around 5000 syscalls, compared to less than 250 for any *nix. Most CS labs now are run on Linux (some are BSD), because it is possible to understand what is going on. Components are replaceable, boundaries are sharp. Experimentation is easy. Minix is still used when even greater simplicity is required : hacking on the VM of BSD or Linux is kind of wearing for an undergrad.
Maybe you like the idea of signing NDAs just to get an education. Most people don't - and that is why almost no serious CS in universities is done using MS technology - the only ones I know of are funded by MS (eg Cambridge). "Our way is the only way" is not really a rallying cry in universities the way it is among some corporate boardrooms.
CS is a state of mind which bases itself in logic, especially loops.
Fuck are you clueless. Please be honest, you have never been near a decent CS department, have you? Loops, for crying out loud.
I've got a really good conspiracy theory.
Lets take it as given that the entire purpose of the SCO lawsuits is to discredit Linux/Free Software and spread FUD, bankrolled by ...erm... someone.
Enderle has constantly harped on about "these zealots", "death threats", "die for what you believe in" etc, etc.
Enderle is setting himself up to either:
He'll be found in a Reno hotel room with the full text of the GPL version 2 (or later at your discretion) inscribed on his chest using a slide rule. A single penguin feather will be protruding from between his mutilated butt cheeks, and the smell of herring will linger.
Who would use software that causes people to kill in such a biblical fashion? Thats how it is going to play - Just wait until the mainstream media condenses down John Katz' old columbine mooning into a blanket condemnation of all geekhood.
I think it could happen just when ... someone feels that they are getting almost no return from the SCO lawsuit in terms of FUD, and they need a real boost. Its desparate, but the potential pay off is huge - better than patents!
Do I really beleive this? No. But its worth putting it out there, just so if it did happen we can all pat each other on the back in Gitmo, and say "Knew it was coming! Why didn't we listen?
So use the slow casting method if the vm ever notices a boxed instance being put in the collection, ie either
a) The boxed instance was not created by code immediately before insertion (therefore it was not auto boxing in the compiler)
b) The boxed instance had another reference in the same frame ( this really is pretty simple).
Not really a show stopper... given that I doubt very many people are going to be doing that.
Of course assembler is useful. It is shocking that a large number of CS courses don't teach it at all. I'm even more surprised at the number of CS students who have no clue about any area of the subject they have a degree in.
I do use inline assembler... as rarely as I can. Only when the compiler is being stupid. C has its place too - operating systems courses.
But to believe that it is a good idea to be teaching algorithms or fundamental concepts in assembler is utterly braindead. Functional languages are just so much better for learning this kind of stuff. And this is the stuff that computer *SCIENCE* is about.
What I find amazing is the number of fools who think that writing assembler makes them a programming god. The fact is that most of the time, you are far better off, optimisation wise, looking at your algorithms and data structures than messing at that level. Only if you know a small bit of code is messing you over, AND the compiler is doing something wrong, should you resort to assembler.
The other area where it is necessary is when you are working on a compiler (which I've done), a virtual machine or a dynamically translating emulator( which I've done.)
My post was a reaction to the grandparents assertion that all code in CS labs should be in "cross platform C". Thats totally fucking wrongheaded.
Could you please explain why a JVM could not notice that you are autoboxing a primitive type into a parameterised collection, and then just store the unboxed primitive efficiently, with a type specialised piece of code, as Pizza or .Net would?
I can't really see why this would be impossible or even hard... and the Java way does keep the class format the same.
If your CS lab thinks that C is a good language for doing computer science in, they are doing something seriously wrong.
Seriously, only operating systems courses can really justify mandating C in this day and age.
Thats what I said. Maybe learn how to read?
Refcounts? I thought the problem with refcount-based systems was that they were not able to deal with circular references. What have I missed?
Cycle breaking. See python.
Basically, ref counting is used as an optimisation : for things which this works on, its a great boost. For the rest, mark-and-sweep is used anyway.
I thought monotone, codeville, and darcs all used the distributed repository model as well as arch & bk. They may be a little further behind in terms of features or surrounding tools, but each one does have some interesting theory/philosophy of version control behind it.
And darcs is written in haskell, so it wins points for enjoying the soundness and showing once again that pure FP can be and is used in the "real world"...
I wouldn't discount any of them yet, but I agree that the subversion fanboys are pretty damn irritating, trying to get every project to switch away from CVS now, when it would clearly be better to wait and see how some of the more revolutionary free systems evolve.
However, anything is better than clearcase...
The EULA of VS.Net tries to forbid it.
This is of course totally non-binding and non-enforceable. When you create something ( the source of your program), you hold the copyright o f that something and can licence it as you wish.
I want a new law, which puts the same legal penalties on a lawyer for writing false legal notices ( known to be bullshit or unenforceable) as false legal advice : possible disbarment and malpractice lawsuits.
That is the exact same feature.
Its really pretty standard on everything but IE.
No, the updates that you were referring to are *still* derivative works of a GPL'ed work. Thus, Redhat cannot provide them to you with additional restrictions. Once you have downloaded the RPMS, you can do what you like with them.
What Redhat can and do restrict is the access to their download servers. They are not required to let you use their bandwidth.... and that is what you are paying for, along with support.
Eclipse, IntelliJ, Visual Age for Java (old), Visual Works - any small talk IDE ...
if you don't have refactoring capabilities, wtf is the IDE for?
Wow, that worked really well for Sun didn't it? Have you not noticed that all of the Microsofties are going around screaming about
Sun prompting MS to produce a Java clone/alternative was a bad idea.... you know, the whole keep your enemies closer thing.
You have absolutely no fucking clue if you think that Konsole or gnome-terminal embed bash or sh. They are terminal programs! They talk to shells via a pty! "bash mode" indeed....
Except people who actually want to swim.
It almost always seems to be an AGP chipset issue with nvidias binary crapware. Since I changed mother boards from an oldish VIA chipset to a newish one, I've not had any lockups from the binary drivers.
But I still resent them...
Doesn't it ever get just a tiny bit boring trolling for Microsoft?
KDE do *NOT* release binaries.
Lets put it this way - a patch was released that allowed any user to fix their system, in 90 minutes, by KDE. The competitive market place of Linux distros led to some vendors offering this fix in less than a day.
The same was only done by Microsoft in two weeks.
How the *fuck* can you try to spin that Microsofts way?
Saying something doesn't make it true.
Or maybe you could ask them why the fuck they can't write a permissable binary kernel module like every body else, but actually patch the kernel instead?
Trolltech are mistaken in that statement. They have released QT/E under the GPL, which means an aggregate product can be commercial, as long as it is distributed under the GPL as well.
IE, you can't make proprietary software without paying Trolltech. Commercial software is fine.
It works, kinda, but its shite. Not even as good as ASVG6.
But when will they actually release an updated SVG Viewer?
The currently released version is just wbout 2 years old. The preview of version 6 is better, but won't get installed anywhere until its actually released officialy and bundled with Acrobat reader.
And corporate muppets won't roll out preview releases.
I have to wonder how commited Adobe is to SVG. Their preview release of ASV6 is good enough to discourage competitors (it would take quite an effort to match it in a ~2 year timeframe), but won't get installed anywhere until release. Are they scared that SVG will eat PDF as well as Flash?
The licencing of your cell project is inconsistent:
doc/licence says LGPL.
But it also has a vague preamble that adds additional terms to the LGPL, namely that derivative works authors must distribute the work back to you ( and thereby release the source to you too.) The plain LGPL does not require this.
readme says that it is a "commercial open-source" project that can be licenced for $99. This is confusing.
It would be good if this could be cleared up. I would recommend that you make it LGPL. Ask a lawyer how to add your extra restriction if you really want it. Be aware that this will make it incompatible with the GPL, and the plain LGPL.
If you also want to offer commercial licencing under different terms, make the dual licence nature clear. However, going this route, the GPL is a better bet as the free licence to get commercial users to cough up. There is very little that a commercial project can't do with an LGPL'ed library.