That shouldn't be necessary. Just rewrite some of it so that it has a proper backend and frontend like xine. Then we can write a gnome and kde frontend without so much effort, like totem and kaffeine.
And yes, I am willing to do it myself if no-one else will. But if that's what you want, expect a full port of OOo 1.3 some time in 2008. Seriously, is anyone organising a project anything like this? Because I will join and submit anything I can.
Well, KDE integration is less important, since KDE has a proper office suite of its own. As far as I can see, "gnome office" consists of Abiword and Gnumeric pretending they're an office suite.
Yes, it will be possible, just like there's a no-drm patch for xpdf. However, just like that patch, anyone publishing this player probably won't want to include the nodrm bit for fear of looking like a piracy tool, and I doubt studios will be worried about the few people who will put the effort into getting it. There's a cracker for wmv drm available anyway, that doesn't even use the keys.
How big is this WMV which the dll is too slow on? My 800mhz duron, hardly a speed demon, can play every wmv9 I've thrown at it without a hint of slowdown except while loading it, and doesn't seem to be going anywhere near 100% cpu.
No one really knows how the brain works, and so far I've only really seen one way to solve a general quadratic equation. So I do think our brains actually solve quadratic equations in the same way we do, only much faster. If you can think of another way the brain would do it I'm prepared to be convinced.
The problem is you can't burn "bit for bit". You have to burn at least a sector at a time, and you can't (completely) control the control bits which get written by the burner. So at the very least you need to do something to the burner firmware.
The book "Anvil of Stars", by Greg Bear, has an idea of something like this in it, and is a pretty good book anyway. Think about it - the uncouncious trig you do to work out how far away something is, and the quadratic you solve every time you catch a ball, are much faster than you could do "manually".
That interface was still faster and more stable than the code you were supposed to use, which is why the cd tool still used it. It was deprecated too hastily and dropped too soon, and the only justification I have ever seen has been that linus didn't like it.
No, that just means 2.4.27 is not 100% stable everywhere. Maybe 2.4 hasn't fully stabilised yet, after all there are still a few new features being added (an equally brief changelog scan found a new crypto algorithm in 2.4.28-rc3), but what about 2.2? It is still maintained, yet there have been no new releases since February.
I think it's sad that they decided to go all "corporate". You won't find the real name anywhere on that site, not even a buried faq entry saying where the acronym comes from. I realise that this sort of thing is probably necessary if linux is going to become mainstream, but it strikes me as the equivalent of local shops turning into faceless chain stores.
I still use it, but I miss telling people I have a pimp-ass newsreader. At least I still gimp my images, although from a recent thread it seems there are people who want to get rid of that too.
If you're sure safe mode doesn't work, F8 on startup and choosing step by step confirmation will. Of course you need to know which drivers are real ones, but you learn that.
You overestimate people. The average windows user wouldn't know what the registry was if it hit them on the head. And why should they? To them, the computer is just an appliance, they don't care about what's underneath.
You don't like a language because you have to write eight lines for hello world? Hmmm, sounds like a serious metric to me!
Well, it's not the best measure, but it's an indication of how useless the language is for doing anything other than OO. A language with support for OO is good, but a language which only supports OO is not so good. Historically languages designed around a specific paradigm (Haskell and Lisp for example) have usually been pretty specialised and used only within a small area, not for general purpose programming. Until it has better support for non-oo programming, java should be the same, a niche language for certain types of programs. It's used far more than it should be, probably because of the "name appeal". I would only use java for a project where I knew almost everything would be OO, and I can't imagine a big project where that is the case.
But he will make bad decisions. Everyone does. And one person's bad decision shouldn't cause problems for everyone, which is the problem with a cathedral model.
IMO we've already paid a price for this with the cd-writing fiasco. Linus chucked out the only decent cd-burning interface because he didn't like it, and it made things worse for everyone.
EJB are bloated etc: J2EE is does NOT equal Enterprise Javabeans. J2EE contains classes for lots of things. XML processing, messages, web servers, database connectivity, etc. You don't have to use EJB. Lots of Java developers don't like EJB because they are too cumbersome, and there are plenty of alternatives. Check out for instance O'Reillys recent book Better, Faster, Lighter Java.
However, java itself is bloated. The freaking runtime is 20mb. If you want to be able to develop in it, don't expect much change out of 100mb. And that's without documentation, and compressed. Installed compiler and documentation, you're pushing 200mb. I can get a whole operating system, with full office suite, video playback, email, usenet, instant messaging, the works, in less than that, here.
Java is slow: Startup time for the JVM is still slow yes. This rarely matters for a web/application server. When it comes to running, it is plenty enough.
Most of the running stuff is slow too, especially with gui. You will say this is because of swing. This is true. However, swing is touted as a *main advantage of java*, because it means the uis are consistent. So either java is hopelessly slow for anything that needs a gui, or it's just as inconsistent as c or anything else wrt guis and all those megabytes of runtime and compiler are not helpful libraries but just useless bloat. Take your pick.
It isn't open source: So what. It's close enough.
Not for all of us. Some people are devoted debian users. Even for those of us who aren't, I personally think it's unreasonable to expect me to use a language that I am not allowed to distribute the *runtime* for, yet alone the compiler. Yes, I'm allowed to distribute a jre with a java application, but this is a special exception in the license which, to my eyes, means the license is very wrong.
OO is appropriate for many cases, but nowhere near all of them. Anyone who says otherwise is a zealot. Take a look at the java hello world some time. It is EIGHT LINES. You have to crate a new class and reimpliment the main method just to be able to print text. Compare that to python, where a hello world is simply "print "Hello, World!\n"". That's why I don't like java.
if you prefer using c++ to c, then use gtkmm, the c++ api for gtk, which follows the conventions used in the c++ stl so will be far more familiar than qt for someone who hasn't used either.
I don't. I prefer C to C++, and I prefer python to either. (Side note: PyQt rules, PyGtk is fairly nice but still not as easy to use). But even without actually knowing c++ (I guess based on my knowledge of c and java) and knowing c, I still find programming Qt apps in c++ to be easier than programming gtk apps in c.
As for gtkmm, http://www.telegraph-road.org/writings/gtkmm_vs_qt.html is a good summary of the advantages of qt by someone who was actually just trying to find a toolkit to use for their c++ program, probably better than any arguments I can make here. I'll just say that, despite being "more standardised", gtk is basically harder to use. I have written programs, basic ones but still programs for things I needed to do, in both gtk and qt, and actually learnt gtk first, but I find qt to be far more intuitive.
The tools are a black-box, just like function calls, libraries, and lots of other things. And we need them so that we can concentrate on things without having to worry about the big picture, because in a modern programming project the big picture is simply too big.
The reason the linux kernel works is that people who know the hardware can write drivers without worrying about the rest of the kernel. The same is true for almost any multi-person project, where perhaps the gui designer can design the gui without having to understand the rest of the project.
A good tool is one that means you don't have to understand how it works, just what it does. If not you might as well be doing whatever the tool does by hand.
Huh? I don't know about you, but when I moved from machine language to assembler, I used a *tool* called an assembler, and when I moved to C I used a *tool* called a compiler. Tools are what make languages possible, and IDE templates are just as much a language as c++ is.
Slashdot Mirror
And yes, I am willing to do it myself if no-one else will. But if that's what you want, expect a full port of OOo 1.3 some time in 2008. Seriously, is anyone organising a project anything like this? Because I will join and submit anything I can.
Well, KDE integration is less important, since KDE has a proper office suite of its own. As far as I can see, "gnome office" consists of Abiword and Gnumeric pretending they're an office suite.
Yes, it will be possible, just like there's a no-drm patch for xpdf. However, just like that patch, anyone publishing this player probably won't want to include the nodrm bit for fear of looking like a piracy tool, and I doubt studios will be worried about the few people who will put the effort into getting it. There's a cracker for wmv drm available anyway, that doesn't even use the keys.
How big is this WMV which the dll is too slow on? My 800mhz duron, hardly a speed demon, can play every wmv9 I've thrown at it without a hint of slowdown except while loading it, and doesn't seem to be going anywhere near 100% cpu.
No one really knows how the brain works, and so far I've only really seen one way to solve a general quadratic equation. So I do think our brains actually solve quadratic equations in the same way we do, only much faster. If you can think of another way the brain would do it I'm prepared to be convinced.
The problem is you can't burn "bit for bit". You have to burn at least a sector at a time, and you can't (completely) control the control bits which get written by the burner. So at the very least you need to do something to the burner firmware.
The book "Anvil of Stars", by Greg Bear, has an idea of something like this in it, and is a pretty good book anyway. Think about it - the uncouncious trig you do to work out how far away something is, and the quadratic you solve every time you catch a ball, are much faster than you could do "manually".
That interface was still faster and more stable than the code you were supposed to use, which is why the cd tool still used it. It was deprecated too hastily and dropped too soon, and the only justification I have ever seen has been that linus didn't like it.
No, that just means 2.4.27 is not 100% stable everywhere. Maybe 2.4 hasn't fully stabilised yet, after all there are still a few new features being added (an equally brief changelog scan found a new crypto algorithm in 2.4.28-rc3), but what about 2.2? It is still maintained, yet there have been no new releases since February.
I still use it, but I miss telling people I have a pimp-ass newsreader. At least I still gimp my images, although from a recent thread it seems there are people who want to get rid of that too.
I think you're talking about The Avengers.
If you're sure safe mode doesn't work, F8 on startup and choosing step by step confirmation will. Of course you need to know which drivers are real ones, but you learn that.
You overestimate people. The average windows user wouldn't know what the registry was if it hit them on the head. And why should they? To them, the computer is just an appliance, they don't care about what's underneath.
IIRC the story must be exaggerated at least, as the heights etc. given are implausible.
Well, fine then, but my point still stands. Why is it that using a compiler is ok but using an ide to generate code is not?
Well, it's not the best measure, but it's an indication of how useless the language is for doing anything other than OO. A language with support for OO is good, but a language which only supports OO is not so good. Historically languages designed around a specific paradigm (Haskell and Lisp for example) have usually been pretty specialised and used only within a small area, not for general purpose programming. Until it has better support for non-oo programming, java should be the same, a niche language for certain types of programs. It's used far more than it should be, probably because of the "name appeal". I would only use java for a project where I knew almost everything would be OO, and I can't imagine a big project where that is the case.
Then don't claim "no kernel is stable everywhere", if you haven't tried using a mature kernel.
The FSF is a charity. Just how would MS "forcefully take ownership"?
Will this version become comptiable with licenses with patent clauses, like the apache, IBM and CPL licenses?
IMO we've already paid a price for this with the cd-writing fiasco. Linus chucked out the only decent cd-burning interface because he didn't like it, and it made things worse for everyone.
J2EE is does NOT equal Enterprise Javabeans. J2EE contains classes for lots of things. XML processing, messages, web servers, database connectivity, etc. You don't have to use EJB. Lots of Java developers don't like EJB because they are too cumbersome, and there are plenty of alternatives. Check out for instance O'Reillys recent book Better, Faster, Lighter Java.
However, java itself is bloated. The freaking runtime is 20mb. If you want to be able to develop in it, don't expect much change out of 100mb. And that's without documentation, and compressed. Installed compiler and documentation, you're pushing 200mb. I can get a whole operating system, with full office suite, video playback, email, usenet, instant messaging, the works, in less than that, here.
Java is slow:
Startup time for the JVM is still slow yes. This rarely matters for a web/application server. When it comes to running, it is plenty enough.
Most of the running stuff is slow too, especially with gui. You will say this is because of swing. This is true. However, swing is touted as a *main advantage of java*, because it means the uis are consistent. So either java is hopelessly slow for anything that needs a gui, or it's just as inconsistent as c or anything else wrt guis and all those megabytes of runtime and compiler are not helpful libraries but just useless bloat. Take your pick.
It isn't open source:
So what. It's close enough.
Not for all of us. Some people are devoted debian users. Even for those of us who aren't, I personally think it's unreasonable to expect me to use a language that I am not allowed to distribute the *runtime* for, yet alone the compiler. Yes, I'm allowed to distribute a jre with a java application, but this is a special exception in the license which, to my eyes, means the license is very wrong.
OO is appropriate for many cases, but nowhere near all of them. Anyone who says otherwise is a zealot. Take a look at the java hello world some time. It is EIGHT LINES. You have to crate a new class and reimpliment the main method just to be able to print text. Compare that to python, where a hello world is simply "print "Hello, World!\n"". That's why I don't like java.
I don't. I prefer C to C++, and I prefer python to either. (Side note: PyQt rules, PyGtk is fairly nice but still not as easy to use). But even without actually knowing c++ (I guess based on my knowledge of c and java) and knowing c, I still find programming Qt apps in c++ to be easier than programming gtk apps in c.
As for gtkmm, http://www.telegraph-road.org/writings/gtkmm_vs_qt .html is a good summary of the advantages of qt by someone who was actually just trying to find a toolkit to use for their c++ program, probably better than any arguments I can make here. I'll just say that, despite being "more standardised", gtk is basically harder to use. I have written programs, basic ones but still programs for things I needed to do, in both gtk and qt, and actually learnt gtk first, but I find qt to be far more intuitive.
The reason the linux kernel works is that people who know the hardware can write drivers without worrying about the rest of the kernel. The same is true for almost any multi-person project, where perhaps the gui designer can design the gui without having to understand the rest of the project.
A good tool is one that means you don't have to understand how it works, just what it does. If not you might as well be doing whatever the tool does by hand.
Huh? I don't know about you, but when I moved from machine language to assembler, I used a *tool* called an assembler, and when I moved to C I used a *tool* called a compiler. Tools are what make languages possible, and IDE templates are just as much a language as c++ is.