Speaking as a Solaris admin of nine years, this is the best news Dell and Red Hat could ever get.
Yeah, Oracle has been so unkind to customers since the Sun acquisition that at this point, it's less like Oracle is a doctor trying to bring Sun back to life, and more like Oracle is a drug addled psychotic who filled the rotted corpse of Sun with a bunch of knives and used needles and has decided to rape it continuously until sunrise. At this point, so many would-be Sun customers have been hearing this steady drumbeat of "Oracle are acting like jackass" stories that even if they became the perfect vendor tomorrow, almost nobody would touch them with a ten foot pole.
This is just conjecture. On a 'big' war day we are going to want to disable enemy satellites. We have ground based interceptors -- but there can be delays in launch windows, plus the 'bad' guys are going to be on guard and can take some evasive actions.
How about our little X-37 with a cargo bay and manipulator arm goes and pays those 'nasty' satellites a visit right now and attaches a few pounds of high explosive with a radio detonator. When the war starts you push a button and they all disappear!
Just in case they send a maintenance flight up, our little bomblets can also be equipped with a radio controlled 'spring' that detaches them from the satellite. No one is the wiser.
Possible?
Technically, yes it is possible. However, it would require a massive amount of delta-v to rendezvous with a significant number of enemy sats. And, that doesn't provide any reason to make something that can return to earth. You could launch a vehicle capable of your proposed mission on a normal rocket. Also, the enemy would notice that you launched something that is visiting all of their sats. At least, they would know you were doing close-pass espionage runs. They will be monitoring the space near their sats to see if anything is on an intersecting orbit that they need to maneuver away from in order to avoid a collision.
I imagine that there are a bunch of draft proposals floating around the Pentagon with ideas like what you propose. Some probably involve lasers (ground based or bolted onto ISS or something), others involve launching anti-sat weapons as a first strike, some involve predeploying assets in orbit for improved "preparednness." All are probably impractical with current opponents, budgets, and levels of technology.
Seems like this could be an effective plug for the analogue hole.
Only if you misunderstand the analog hole. You can push it out as far as you like, but there will always be one. For audio, you eventually have to run analog voltages to a speaker element. If you can't tap that for some reason, you always wind up with actual *sound* that you can record with a microphone.
When the analog hole is pushed that far out, it can be damned inconvenient to make use of it, but it will always be there. Just having a protected stream of bits doesn't matter if you never present the media. Sure, it's secure, but nobody wants it if they can never play it. Once you try and present it, all your digital security stops and the media spills out into the real, analog world.
What a completely useless metric. It makes sense to examine the context and meaning of speech in order to accurately transcribe words, but the number of possible sentences doesn't seem to accurately describe the problem here...
Yeah, my first reaction was to wonder how many valid C language programs are possible. Not that it effects the complexity of making a compiler in any way! If you take it just as a sort of random factoid that adds flavor to your understanding of the problem, I guess it works.
Personally, I've never understood why there hasn't been more work on teaching people to speak in a way that computers understand. I mean, we don't type plain English at the BASH prompt. We learn rules for quoting. We learn special characters. We learn to type command in an English-inspired language that the computer already knows.
Similarly, I think that if speech recognition ever becomes genuinely useful, it'll be because somebody gave up trying to make a computer understand English (which not even a human can do 100% correctly!), and instead tried to make computer-understandable-english. I'm imagining special characters stated with whistles and clicks. All commands defined in terms of phonemes instead of defining them in terms of written statements and attempting to recognise indirectly. Lots of short, 1-2 syllable non-words used for commands.
How much more performance do we need before we all say: "enough."?
Never. For many people, it happened years ago. My smartphone is plenty fast for checking email, and it has all sorts of un-necessary eye-candy transition effects in the UI. Dialog boxes blur out the background, etc. For something meant to be purely functional, half the performance of my phone would be perfectly adequate for many tasks. Even the lowest end current mainstream Atom is a much faster processor than what's in my phone. So for my dad running apps he is used to, enough already is "enough." Yay, that probably happened over a decade ago.
But, there will always be some of us for whom fast is never fast enough. We'll always find uses for more power. Look at how brilliantly Adobe manages to make even the fastest systems seem wildly inadequate for light web browsing (thanks to flash) and document viewing (their PDF reader). There will always be applications that make us want faster computers. Some of just poor implementations, like flash. Some are inherently hard problems, like detailed fluid simulation. Some are just for the entertainment value, like the latest game engine.
For me, film res compositing, and 3D rendering are the applications that primarily drive my interest in high performance systems. In my end of the world, we can always throw more performance at the problem. More RAM, more/faster storage, more CPU. We'll take whatever we can get. We always will.
I know Kubuntu is the redheaded stepchild of Ubuntu, but you should try out Kubuntu 10.04. I don't know how I lived without tabbed windows.
Without tabbed windows, you lived like some sort of animal. Probably sitting in a pile of your own sick and excrement, as likely to use your computer as to try to eat it or hump it.
Simply put: why is it that we assume an "advanced" civilization means that it is militarily advanced and not ethically advanced?
If they have the technology to make their ships go fast enough to get here, then they have the technology to make things go fast enough to be world killers. They may have no military ambition, but compared to us, it's impossible for an alien civilisation to show up and be incapable of killing us.
That said, I figure that alien attack is pretty far down on my personal list of things to worry about.
His first name is Jae Yoon, and his surname is Ma. Typically written in Korean with the surname first, as Ma Jae Yoon. Ja Mae Yoon is wrong.
Stop using the term "first name" to refer to a second name. It isn't helping. "Given name," works pretty much universally, without implying a correct order.
Unbelievably, most of the C++ coders miss this point! Some try to rewrite it with cin and cout - as if that would fix the problem! Still others mention some gobblygook about stacking dishes and overflowing the sink, or something like that. What they don't realize is that C++ - using information hiding - can hide the passwords in your code from prying eyes. They even have a keyword for it - private - which means that hackers can't see it.
Of course, the gobblygook answer about overflowing the stack is a good starting point, it assumes more than you are actually guaranteed. For example, on an embedded system, the "password" text may wind up in ROM. When you try to scanf into a pointer than is pointing to an address in ROM, you can rest assured that you won't actually overflow the stack. (Especially if the password you input is adequately short.) Still, even with a short password, the results might not be quite what you had been hoping.
I have to say, I've gotten some pretty (ahem) creative responses, too. And for all you job hunters out there, if you put "C/C++" on your resume, I guarantee my first technical question is going to be, "What's the difference between C and C++?" All the while knowing that there is a >50% chance I'm about to get a "creative" answer.
The difference is that C++ complains less when I put functions in a struct. But, I prefer C anyways because I can name a variable "class." For some reason, C++ complains when I try and do simple stuff like that, so I guess it just isn't as complete of a language as C yet, since it has such weird restrictions on variable names.
One of the first people on board with the project is Trent Nelson; he's all about Python.
I think we're gonna cover that.
I think I may be in your target audience. I'm a Linux guy. I am pretty comfortable developing and deploying in my home environment. I'm not an expert, but I can get by. If you can make it easy for somebody like me to make a.msi package that ordinary Windows users can easily install with a few clicks, then I would probably pay a lot more attention to that half of the world.
So wait - artificial physical contact generated by a computer can desensitize people from their real fear of being touched? But artificial mayhem in video games generated by a computer does not desensitize them against real violence?
Somewhere, a behavioral psychologist is quietly crying...
If playing Quake actually felt like getting shot or shooting, then yes, I expect that people who played lots of Quake would get desensitized to real violence. Basic Training puts a soldier through a vague simulation of some warlike actions, and it certainly desensitizes soldiers to warlike situations.
I remember doing hacks like this to Mac applications -- back when they had resource + data forks. The resource forks contained all their sounds, graphics, icons, forms, etc. With ResEdit you could simply open up (most) applications and tweak them to your hearts content:)
These days, if you want to customize an OS X app, you just open up the.app folder, and play with the.nib files. It's actually kind of fun to open up the.nib's in Interface Builder and change around all the buttons for iTunes or whatever. It als doesn't require super fancy research...
You don't have to use 3d studio max to generate an image with a computer. I would suggest that this image is in fact generated by a computer. It's just generated from sonar data instead of an artists interpretation.
Yeah, that's kind of a canonical example of a computer generated image. They had a bunch of sonar data which was put through an algorithm which resulted in a picture. People don't really seem to care what words mean anymore. It's a shame. Or, maybe it's a pancake. Doesn't make any difference to most people.
I've tried EVERYTHING under windows, and none of it can hold a candle to the workflow and speed of quality production as the FCP suite. Even AVID. I'd utterly kill for something that was 1/2 as effective as FCP for linux. but sadly nothing exists except toys that crash all the time or are for making really low quality home movies.
Unfortunately, the only real answer for NLE on Linux that can thrash Final Cut is something like Smoke. If you don't like what it costs to get a MacPro ready with Final Cut, you *really* won't like the cost of moving to a platform like Smoke. Also, because it's very high end, it's almost impossible to find information on how to do stuff if you run into a problem. (Though, by all accounts, the commercial support is very good.)
I'm hard pressed to believe that the Space Shuttle is the best idea we've come up with in this industry. When I see people shedding tears over the canceled program, I see the same old heavy client programmers who couldn't adapt to web programming.
There certainly is some sort of analogy that can be made. The Shuttle is kind of like a hacked together old code base made out of Cobol and MUMPS that requires special mainframe hardware to stay running. In theory, we could replace the whole think while some Python or Java running on a single Linux box. We could even use the new box to run Apache and do all sorts of new web stuff. So, we have decided to throw away the old mainframe, and delete the old code base because it is so out of date and expensive.
Oh, and the requisition for the Linux box and a Java developer? Well, that's held up by the accounting department. We may or may not ever get it.
And, a car is involved somehow because this is an analogy.
Cybermen and Sontarans - I agree. Very boring. But the Daleks and their successive returns from oblivion never get old.
The 9th Doctor episode "Dalek" kicked ass.
Every appearance of the Daleks in the new run since has been worthless. Doubly so fr the appearances since the 9th Doctor left. Too much reliance on the fact that CGI allows the show to use far more Daleks on screen than they could in the old series. Absolutely no thought given to whether or not showing a million CGI Daleks actually substitutes for a plot.
Bring back the unstoppable sense of ruthlessness that they pulled off with that one lone Dalek in the underground bunker. *That* never gets old.
Ahh, but this is compile-once, right!? That makes it so much better!...but Java already had this. Sure, you have to detect the platform and adjust accordingly - but it's been happening for a huge part of Java's existence.
In all honesty, you pulling it off in C/C++ makes you a "good" programmer, because your code doesn't actively make the codebase harder to maintain or port. If your code retains the possibility of targeting different architectures and devices, you're doing it right - at least for games.
C++/qt is compile-once-per-platform, but is that really that big of a deal? It honestly only takes me a few seconds to make a change and rebuild my current C++/qt project. It would take more time to rebuild from scratch, but 3/4 of the bject files are already built whenever I need to build my current changes. And, if you really are that compilation averse, why go with a compile-once toolchain anyway, when you can use a compiler-never toolchain like python.
The world would be a better place if somebody would just stab Adobe in the teeth.
And, it's kind of sad if I actually count as a good programmer these days. In my last job, I wrote and maintained a suite of in-house tools for doing a variety of tasks. Mostly python and C++. Everything worked on Linux and Windows, and would have been trivial to bring to OS-X if needed. I ran some of it on my phone when I had to do sutff from inside the datacenter.
My Job was running Network Operations. I wasn't even a developer. Seriously, a little interest in the subject, and a willingness to take a little effort to do things properly and make even a half-assed evaluation of the available tools, and this stuff just isn't hard to do these days. It was strictly a spare-time thing at that job.
Maybe you should read some dynamic language code before making a comment like this. It's not variable and function names that are terse and compact: it's the amount (or lack thereof) of boiler plate code that accomplishes the same job. Compact syntax makes code more readable, not less, because you don't have to figure out which parts of the complexity are accidental, and which parts are essential. And of course you know that you can write inexplicably named, compact variable and function names in any language.
Sorry, that shouldn't have been anonymous.
Exactly! The reason I love Python so much is because I don't have to write very much of it for the sorts of things that I use it for. In some cases, a 100 line python script does as much work as a codebase in c++ of a few headers, a few implementation files, and a make file which add up to like 10x the writing I have to do. (Sure, a Makefile isn't actually written in c++. But, controlling the build process is one extra thing I have to worry about in c++ vs. python, and more stuff that I need to write.)
OTOH, with some types of code, the amount of headaches that I'd have to deal with in python just aren't worth it. As much as I love python, if I have to write a lot of it, I go looking for a better solution.
Java? It's really sluggish and non-elegant compared to AIR. I'm still not sure how the Java interpreter performs so badly on UI elements. While technically probably OK (it's been losing to.NET though), the mere fact that all Java apps feel sluggish creates a bad image for Java. That combined with the fact that.NET has been constantly instructing new features and tens of languages support it (C#, Basic, Object Pascal, ASP..)
Java does bad on GUI's. It's true. OTOH, it doesn't actually do that badly on graphics. It has OpenGL bindings, so for things like making games, it actually does just fine. I'm not personally a big fan of Java, but normal "businessy" GUI apps are really the only are where Java really falls down on performance. And, even then, it's nowhere near as bad as the Java reputation would suggest.
Guy creates functionality I've been using in Java for 8 years; film at 11.
Off the top of my head... Java, Python, C using GLUT-OpenGL-OpenAL, C++ using Qt... There are tons of platforms that I've personally worked with where I had no problem supporting multiple platforms with GUI and audio while being able to maintain a single code base. Does this make me some sort of portability super-developer? Should I be doing speaking tours to brag about my amazing accomplishments?
No, I'm just some dumb fuck mediocre part-time programmer who has his head stuck slightly less far up his ass that the people who think AIR is interesting.
Uhuh... so you determine language quality by the terseness of it's text.
Interesting.
It's certainly not a complete measure of a language, but they are certainly worse ways to look at it. Java has some inconveniences, like famously deep class hierarchies which lead to very long lines of code. OTOH, Java has support for generics, and some inherent safety mechanisms which would need to be handled manually in C. In C, it's normal to have a set of similarly named functions that operate on floats, doubles, int, shorts, etc., which are all basically copy-pasta, which is a bad thing about C. Like most bad things in a language, it really does directly correlate with the fact that you have to write more C.
Nobody ever sold a language feature on the fact that it was hard to use. Every language prides itself on the fact that you don't have to write more of it.
Slashdot Mirror
Yeah, Oracle has been so unkind to customers since the Sun acquisition that at this point, it's less like Oracle is a doctor trying to bring Sun back to life, and more like Oracle is a drug addled psychotic who filled the rotted corpse of Sun with a bunch of knives and used needles and has decided to rape it continuously until sunrise. At this point, so many would-be Sun customers have been hearing this steady drumbeat of "Oracle are acting like jackass" stories that even if they became the perfect vendor tomorrow, almost nobody would touch them with a ten foot pole.
Technically, yes it is possible. However, it would require a massive amount of delta-v to rendezvous with a significant number of enemy sats. And, that doesn't provide any reason to make something that can return to earth. You could launch a vehicle capable of your proposed mission on a normal rocket. Also, the enemy would notice that you launched something that is visiting all of their sats. At least, they would know you were doing close-pass espionage runs. They will be monitoring the space near their sats to see if anything is on an intersecting orbit that they need to maneuver away from in order to avoid a collision.
I imagine that there are a bunch of draft proposals floating around the Pentagon with ideas like what you propose. Some probably involve lasers (ground based or bolted onto ISS or something), others involve launching anti-sat weapons as a first strike, some involve predeploying assets in orbit for improved "preparednness." All are probably impractical with current opponents, budgets, and levels of technology.
Only if you misunderstand the analog hole. You can push it out as far as you like, but there will always be one. For audio, you eventually have to run analog voltages to a speaker element. If you can't tap that for some reason, you always wind up with actual *sound* that you can record with a microphone.
When the analog hole is pushed that far out, it can be damned inconvenient to make use of it, but it will always be there. Just having a protected stream of bits doesn't matter if you never present the media. Sure, it's secure, but nobody wants it if they can never play it. Once you try and present it, all your digital security stops and the media spills out into the real, analog world.
Yeah, my first reaction was to wonder how many valid C language programs are possible. Not that it effects the complexity of making a compiler in any way! If you take it just as a sort of random factoid that adds flavor to your understanding of the problem, I guess it works.
Personally, I've never understood why there hasn't been more work on teaching people to speak in a way that computers understand. I mean, we don't type plain English at the BASH prompt. We learn rules for quoting. We learn special characters. We learn to type command in an English-inspired language that the computer already knows.
Similarly, I think that if speech recognition ever becomes genuinely useful, it'll be because somebody gave up trying to make a computer understand English (which not even a human can do 100% correctly!), and instead tried to make computer-understandable-english. I'm imagining special characters stated with whistles and clicks. All commands defined in terms of phonemes instead of defining them in terms of written statements and attempting to recognise indirectly. Lots of short, 1-2 syllable non-words used for commands.
Never. For many people, it happened years ago. My smartphone is plenty fast for checking email, and it has all sorts of un-necessary eye-candy transition effects in the UI. Dialog boxes blur out the background, etc. For something meant to be purely functional, half the performance of my phone would be perfectly adequate for many tasks. Even the lowest end current mainstream Atom is a much faster processor than what's in my phone. So for my dad running apps he is used to, enough already is "enough." Yay, that probably happened over a decade ago.
But, there will always be some of us for whom fast is never fast enough. We'll always find uses for more power. Look at how brilliantly Adobe manages to make even the fastest systems seem wildly inadequate for light web browsing (thanks to flash) and document viewing (their PDF reader). There will always be applications that make us want faster computers. Some of just poor implementations, like flash. Some are inherently hard problems, like detailed fluid simulation. Some are just for the entertainment value, like the latest game engine.
For me, film res compositing, and 3D rendering are the applications that primarily drive my interest in high performance systems. In my end of the world, we can always throw more performance at the problem. More RAM, more/faster storage, more CPU. We'll take whatever we can get. We always will.
D'oh. Well, it was almost a good joke.
Maybe I should get outside more. Like most slashdotters, I can't actual tell the difference between the months!
No, just another ten years of November.
Without tabbed windows, you lived like some sort of animal. Probably sitting in a pile of your own sick and excrement, as likely to use your computer as to try to eat it or hump it.
We all did. Thank god we upgraded.
If they have the technology to make their ships go fast enough to get here, then they have the technology to make things go fast enough to be world killers. They may have no military ambition, but compared to us, it's impossible for an alien civilisation to show up and be incapable of killing us.
That said, I figure that alien attack is pretty far down on my personal list of things to worry about.
Stop using the term "first name" to refer to a second name. It isn't helping. "Given name," works pretty much universally, without implying a correct order.
And Windows is only the retail price under the exact same conditions.
Of course, the gobblygook answer about overflowing the stack is a good starting point, it assumes more than you are actually guaranteed. For example, on an embedded system, the "password" text may wind up in ROM. When you try to scanf into a pointer than is pointing to an address in ROM, you can rest assured that you won't actually overflow the stack. (Especially if the password you input is adequately short.) Still, even with a short password, the results might not be quite what you had been hoping.
The difference is that C++ complains less when I put functions in a struct. But, I prefer C anyways because I can name a variable "class." For some reason, C++ complains when I try and do simple stuff like that, so I guess it just isn't as complete of a language as C yet, since it has such weird restrictions on variable names.
What do I win?
I think I may be in your target audience. I'm a Linux guy. I am pretty comfortable developing and deploying in my home environment. I'm not an expert, but I can get by. If you can make it easy for somebody like me to make a .msi package that ordinary Windows users can easily install with a few clicks, then I would probably pay a lot more attention to that half of the world.
If playing Quake actually felt like getting shot or shooting, then yes, I expect that people who played lots of Quake would get desensitized to real violence. Basic Training puts a soldier through a vague simulation of some warlike actions, and it certainly desensitizes soldiers to warlike situations.
These days, if you want to customize an OS X app, you just open up the .app folder, and play with the .nib files. It's actually kind of fun to open up the .nib's in Interface Builder and change around all the buttons for iTunes or whatever. It als doesn't require super fancy research...
Yeah, that's kind of a canonical example of a computer generated image. They had a bunch of sonar data which was put through an algorithm which resulted in a picture. People don't really seem to care what words mean anymore. It's a shame. Or, maybe it's a pancake. Doesn't make any difference to most people.
Unfortunately, the only real answer for NLE on Linux that can thrash Final Cut is something like Smoke. If you don't like what it costs to get a MacPro ready with Final Cut, you *really* won't like the cost of moving to a platform like Smoke. Also, because it's very high end, it's almost impossible to find information on how to do stuff if you run into a problem. (Though, by all accounts, the commercial support is very good.)
There certainly is some sort of analogy that can be made. The Shuttle is kind of like a hacked together old code base made out of Cobol and MUMPS that requires special mainframe hardware to stay running. In theory, we could replace the whole think while some Python or Java running on a single Linux box. We could even use the new box to run Apache and do all sorts of new web stuff. So, we have decided to throw away the old mainframe, and delete the old code base because it is so out of date and expensive.
Oh, and the requisition for the Linux box and a Java developer? Well, that's held up by the accounting department. We may or may not ever get it.
And, a car is involved somehow because this is an analogy.
The 9th Doctor episode "Dalek" kicked ass.
Every appearance of the Daleks in the new run since has been worthless. Doubly so fr the appearances since the 9th Doctor left. Too much reliance on the fact that CGI allows the show to use far more Daleks on screen than they could in the old series. Absolutely no thought given to whether or not showing a million CGI Daleks actually substitutes for a plot.
Bring back the unstoppable sense of ruthlessness that they pulled off with that one lone Dalek in the underground bunker. *That* never gets old.
C++/qt is compile-once-per-platform, but is that really that big of a deal? It honestly only takes me a few seconds to make a change and rebuild my current C++/qt project. It would take more time to rebuild from scratch, but 3/4 of the bject files are already built whenever I need to build my current changes. And, if you really are that compilation averse, why go with a compile-once toolchain anyway, when you can use a compiler-never toolchain like python.
The world would be a better place if somebody would just stab Adobe in the teeth.
And, it's kind of sad if I actually count as a good programmer these days. In my last job, I wrote and maintained a suite of in-house tools for doing a variety of tasks. Mostly python and C++. Everything worked on Linux and Windows, and would have been trivial to bring to OS-X if needed. I ran some of it on my phone when I had to do sutff from inside the datacenter.
My Job was running Network Operations. I wasn't even a developer. Seriously, a little interest in the subject, and a willingness to take a little effort to do things properly and make even a half-assed evaluation of the available tools, and this stuff just isn't hard to do these days. It was strictly a spare-time thing at that job.
Exactly! The reason I love Python so much is because I don't have to write very much of it for the sorts of things that I use it for. In some cases, a 100 line python script does as much work as a codebase in c++ of a few headers, a few implementation files, and a make file which add up to like 10x the writing I have to do. (Sure, a Makefile isn't actually written in c++. But, controlling the build process is one extra thing I have to worry about in c++ vs. python, and more stuff that I need to write.)
OTOH, with some types of code, the amount of headaches that I'd have to deal with in python just aren't worth it. As much as I love python, if I have to write a lot of it, I go looking for a better solution.
Java does bad on GUI's. It's true. OTOH, it doesn't actually do that badly on graphics. It has OpenGL bindings, so for things like making games, it actually does just fine. I'm not personally a big fan of Java, but normal "businessy" GUI apps are really the only are where Java really falls down on performance. And, even then, it's nowhere near as bad as the Java reputation would suggest.
Off the top of my head... Java, Python, C using GLUT-OpenGL-OpenAL, C++ using Qt... There are tons of platforms that I've personally worked with where I had no problem supporting multiple platforms with GUI and audio while being able to maintain a single code base. Does this make me some sort of portability super-developer? Should I be doing speaking tours to brag about my amazing accomplishments?
No, I'm just some dumb fuck mediocre part-time programmer who has his head stuck slightly less far up his ass that the people who think AIR is interesting.
It's certainly not a complete measure of a language, but they are certainly worse ways to look at it. Java has some inconveniences, like famously deep class hierarchies which lead to very long lines of code. OTOH, Java has support for generics, and some inherent safety mechanisms which would need to be handled manually in C. In C, it's normal to have a set of similarly named functions that operate on floats, doubles, int, shorts, etc., which are all basically copy-pasta, which is a bad thing about C. Like most bad things in a language, it really does directly correlate with the fact that you have to write more C.
Nobody ever sold a language feature on the fact that it was hard to use. Every language prides itself on the fact that you don't have to write more of it.