One of the reasons why.NET is doomed to fail hopelessly is that it is windows-only. The other is that.NET costs money.
The success of internet is it's possibility to run on any platform that supports an open standardized networking implementation and all this being free.
.NET will be windows-centered. Especially with macOS X and the coming gnome-2.0 (with a fast , stable mozilla and nautilus), nobody will ever pay for something like.NET. People will use linux rather than paying $1000 for windows. And would you pay $1 per hour for using word (.NET strategy)? Or rather use AbiWord? I know you'll use AbiWord (or StarOffice, KWord, whatever) if it comes to choosing and that is *exactly* why.NET will fail.
Nice question for the slashdot poll: is more than 50% of your software legal and what OS do you use? Hypothesis: 50% no/windows, other 50% yes/linux.
In 20 years, I wonder if the web will still exist at all. Maybe we'll have found something much better, like a world of virtual reality contacts, where we don't need internet anymore.
C# is a proprietary non-standard patented and copyrighted format by MS. The standard of C# is, although the name suggest differently (C...), to be compiled for cross-platform interpretation, like java bytecode.
C/C++/C-NextGen(tm) is supposed to be a free, simple, well-thought-over platform-independent standard programming library which you compile on systems to get system-specific executable binary code.
This is all so opposite. I cannot see how anyone can even think of C# when you're thinking of C... Java in a linux kernel, how odd can we get? (I know it's possible, I just don't want it - the performance decrease is simply to high)
*Ugh*
I don't see why we should make UI extensions, graphical extensions, system-specific extensions in a beautiful yet difficult-to-understand code such as C. Programming is not for weenies or newbies, _real_ programming is for people who know what they're doing. So let the newbies use C# when it's ready because it says M$-something, let the professional newbies use java since their company wants it and let the _real_ programmers use C, C++ or C-NextGen(tm).
I am actually hoping for something more than a basic layer for connecting to other languages like SQL. The problem would then be that companies still blame the layer to be too "plain" and not specific enough and would still make an own implementation which would be incompatible with that of others. I know the C-language developers will know better than this. And they will surely know better than to include UIs, that is something for third party libraries. They are usually system- or at least environment-dependent anyway.
You don't seem to see the whole point of BSD. Linux might be better in a lot of respects than BSD. But by far not in all.
What Windows is for Linux fanatics is what Linux is for BSD-fanatics. And partly, they're right - linux is getting more and more an instable (compare mdk 7.2 stability with openBSD, for example) and desktop-oriented OS, as a competitor for Redmond's crap (btw, I'm talking about full linux/Gnu-distro packages here, not the kernel as such).
BSD is not really a desktop-oriented OS, though freeBSD wouldn't do bad on geek desktops, and I've found openBSD much more stable than any linux distro could ever be. BSD gives you even more the choice, freedom (also compare the licenses) etc. than linux does and this is what makes BSD so nice to use. Linux distros don't go as far as BSDs in choice, configurability, stability etc. See the linux-BSD and windows-linux similarities? It's actually kind of the same discussion.
So for companies that want extremely stable and configurable (unix) servers for free, BSD is THE choice! Unless they wanna make their own linux distro.
It's short-sighted to say that BSD is dying. Just that windows users have never heard of it doesn't mean it's dead. BSD is getting more and more users, just not as quickly as linux but that is probably because it's by far not as user/desktop-oriented as linux, so not really for the desktop, and you just gotta know of it to put it up as a server and that's a process that we're right in now.
Source is smaller? Wrong, compile something, tgz it, tgz source and compare. You'll see that binarie-tgz's are much bigger.
Why source? Because what should the binary be like?
Should it support athlon? P-III? 386? All?
Every option is bad because it either doesn't support them all or it doesn't use the features being able to be used by that hardware or it gets too big and thus slow which is really the last thing we want. And now consider all hardware linux supports and think again...
Compiling a binary for each separate hardware is not the task of the kernel.org-people, that's what distro-makers should do.
Which is why you probably use mandrake, redhat or suse or similar and are happy with it. Those who are not happy compile their new kernel and hope to be happy then...
A friend of mine has been able to see some of the source code of M$' products (somertimes, M$ provides small bits of source code to students). According to him, the amount of reused OO-code is nada (z-e-r-o).
Probably only the GUI-toolkit (buttons, boxes, etc), but is that what we really care about for a good OS?
And for anyone yawning about us bashing IE/M$ when new bugs are discovered - far more bugs have been discovered for IE than for, for example, mozilla. Also, outlook is far more buggy than evolution. Although evolution and mozilla are not even final (non-beta) yet! coincidence? I don't think so...
Slashdot Mirror
One of the reasons why .NET is doomed to fail hopelessly is that it is windows-only. The other is that .NET costs money.
.NET. People will use linux rather than paying $1000 for windows. And would you pay $1 per hour for using word (.NET strategy)? Or rather use AbiWord? I know you'll use AbiWord (or StarOffice, KWord, whatever) if it comes to choosing and that is *exactly* why .NET will fail.
The success of internet is it's possibility to run on any platform that supports an open standardized networking implementation and all this being free.
.NET will be windows-centered. Especially with macOS X and the coming gnome-2.0 (with a fast , stable mozilla and nautilus), nobody will ever pay for something like
Nice question for the slashdot poll: is more than 50% of your software legal and what OS do you use? Hypothesis: 50% no/windows, other 50% yes/linux.
In 20 years, I wonder if the web will still exist at all. Maybe we'll have found something much better, like a world of virtual reality contacts, where we don't need internet anymore.
Imagine, online porn will be much nicer then ;-)
*Ugh*
C# is a proprietary non-standard patented and copyrighted format by MS. The standard of C# is, although the name suggest differently (C...), to be compiled for cross-platform interpretation, like java bytecode.
C/C++/C-NextGen(tm) is supposed to be a free, simple, well-thought-over platform-independent standard programming library which you compile on systems to get system-specific executable binary code.
This is all so opposite. I cannot see how anyone can even think of C# when you're thinking of C... Java in a linux kernel, how odd can we get? (I know it's possible, I just don't want it - the performance decrease is simply to high)
*Ugh*
I don't see why we should make UI extensions, graphical extensions, system-specific extensions in a beautiful yet difficult-to-understand code such as C. Programming is not for weenies or newbies, _real_ programming is for people who know what they're doing. So let the newbies use C# when it's ready because it says M$-something, let the professional newbies use java since their company wants it and let the _real_ programmers use C, C++ or C-NextGen(tm).
I am actually hoping for something more than a basic layer for connecting to other languages like SQL. The problem would then be that companies still blame the layer to be too "plain" and not specific enough and would still make an own implementation which would be incompatible with that of others. I know the C-language developers will know better than this. And they will surely know better than to include UIs, that is something for third party libraries. They are usually system- or at least environment-dependent anyway.
Go C!
You don't seem to see the whole point of BSD. Linux might be better in a lot of respects than BSD. But by far not in all.
What Windows is for Linux fanatics is what Linux is for BSD-fanatics. And partly, they're right - linux is getting more and more an instable (compare mdk 7.2 stability with openBSD, for example) and desktop-oriented OS, as a competitor for Redmond's crap (btw, I'm talking about full linux/Gnu-distro packages here, not the kernel as such).
BSD is not really a desktop-oriented OS, though freeBSD wouldn't do bad on geek desktops, and I've found openBSD much more stable than any linux distro could ever be. BSD gives you even more the choice, freedom (also compare the licenses) etc. than linux does and this is what makes BSD so nice to use. Linux distros don't go as far as BSDs in choice, configurability, stability etc. See the linux-BSD and windows-linux similarities? It's actually kind of the same discussion.
So for companies that want extremely stable and configurable (unix) servers for free, BSD is THE choice! Unless they wanna make their own linux distro.
It's short-sighted to say that BSD is dying. Just that windows users have never heard of it doesn't mean it's dead. BSD is getting more and more users, just not as quickly as linux but that is probably because it's by far not as user/desktop-oriented as linux, so not really for the desktop, and you just gotta know of it to put it up as a server and that's a process that we're right in now.
For BeOS: go BeOS! Everything unix-based rocks!
Source is smaller? Wrong, compile something, tgz it, tgz source and compare. You'll see that binarie-tgz's are much bigger.
Why source? Because what should the binary be like?
Should it support athlon? P-III? 386? All?
Every option is bad because it either doesn't support them all or it doesn't use the features being able to be used by that hardware or it gets too big and thus slow which is really the last thing we want. And now consider all hardware linux supports and think again...
Compiling a binary for each separate hardware is not the task of the kernel.org-people, that's what distro-makers should do.
Which is why you probably use mandrake, redhat or suse or similar and are happy with it. Those who are not happy compile their new kernel and hope to be happy then...
-- Ronald
In your dreams.
A friend of mine has been able to see some of the source code of M$' products (somertimes, M$ provides small bits of source code to students). According to him, the amount of reused OO-code is nada (z-e-r-o).
Probably only the GUI-toolkit (buttons, boxes, etc), but is that what we really care about for a good OS?
And for anyone yawning about us bashing IE/M$ when new bugs are discovered - far more bugs have been discovered for IE than for, for example, mozilla. Also, outlook is far more buggy than evolution. Although evolution and mozilla are not even final (non-beta) yet! coincidence? I don't think so...