Reading through the posts, I was hoping I'd eventually stumble along one that pointed out that "the.NET framework" is a different thing entirely to "the.NET strategy". The first is a very nice common runtime framework... the second is just a bundle of marketing, licencing and privacy-issues. Nobody suggests we do the latter.
The bottom-line principle of the framework is that you can write in whatever language you want, using a single unified class library, and it'll run on any.NET runtime. This is even more ambitious than Java... and if.NET works out, and is ported everywhere, it even means the end of ports - no need to port that snazzy program you just wrote, because it already runs everywhere. I can't see how anyone could argue against this! You write a C# program, compile, and it runs on any CLI runtime.
C# is a good language, and it is to be standardised. Nobody will "own" the language - remember, despite everything, Sun still own Java and can still pretty much do what they want with it - if they didn't, they wouldn't have been able to sue MS over J++. CLI is also a good system, and is also to be standardised. The class library is a well-organised class library, and large chunks of it are to be standardised by the same ECMA taskgroup that is working on CLI. I can see why Microsoft have gone down the standardisation route - it does provide a solidarity behind the language that Java sometimes lacks - but I think there is an opportunity here. In some respects, Microsoft may have erred by going for standardisation - it levels the playing field a lot. If Microsoft sit around and allow the standardisation process to finish, and then alter their implementation in a way that breaks standards-compliant code for non-Microsoft runtimes, then they are really going to have some antitrust issues that I don't think even they can afford to be blase about.
I have to admit also to being concerned about how so many people here have slammed.NET when they blatantly haven't got the foggiest what it is. You can bet your life that Microsoft have got a team of people analysing everything the OSS community does, but the OSS community seems to be avoiding everything MS does in case it 'taints' them. MS do occasionally produce good ideas, C# and CLI are two of them, and this may be an opportunity that we can't afford to let slip by.
Question time: If.NET does succeed in a big way (and I fail to see how it can't), and is ported to non-MS OS's other than Linux... where does that leave Linux? Stuck sadly on the sidelines, as corporations and developers move to a standardised language on a standardised runtime structure whereby they can write programs that run anywhere... anywhere, that is, except Linux, because it's developers didn't want to touch anything that had been anywhere near MS. I don't think Microsoft are going to produce a Linux port... on the contrary, they're probably hoping nobody does. Our own ignorance and contempt of the.NET offering could be our undoing.
.NET isn't rocket science. People have written compilers before, and people have written class libraries before, and people have written runtime environments before. It's fairly simple stuff in that respect. But it's simple stuff done better than most other simple stuff, including our simple stuff, and it's standardised simple stuff.
Slashdot Mirror
Reading through the posts, I was hoping I'd eventually stumble along one that pointed out that "the .NET framework" is a different thing entirely to "the .NET strategy". The first is a very nice common runtime framework... the second is just a bundle of marketing, licencing and privacy-issues. Nobody suggests we do the latter.
The bottom-line principle of the framework is that you can write in whatever language you want, using a single unified class library, and it'll run on any .NET runtime. This is even more ambitious than Java... and if .NET works out, and is ported everywhere, it even means the end of ports - no need to port that snazzy program you just wrote, because it already runs everywhere. I can't see how anyone could argue against this! You write a C# program, compile, and it runs on any CLI runtime.
C# is a good language, and it is to be standardised. Nobody will "own" the language - remember, despite everything, Sun still own Java and can still pretty much do what they want with it - if they didn't, they wouldn't have been able to sue MS over J++. CLI is also a good system, and is also to be standardised. The class library is a well-organised class library, and large chunks of it are to be standardised by the same ECMA taskgroup that is working on CLI. I can see why Microsoft have gone down the standardisation route - it does provide a solidarity behind the language that Java sometimes lacks - but I think there is an opportunity here. In some respects, Microsoft may have erred by going for standardisation - it levels the playing field a lot. If Microsoft sit around and allow the standardisation process to finish, and then alter their implementation in a way that breaks standards-compliant code for non-Microsoft runtimes, then they are really going to have some antitrust issues that I don't think even they can afford to be blase about.
I have to admit also to being concerned about how so many people here have slammed .NET when they blatantly haven't got the foggiest what it is. You can bet your life that Microsoft have got a team of people analysing everything the OSS community does, but the OSS community seems to be avoiding everything MS does in case it 'taints' them. MS do occasionally produce good ideas, C# and CLI are two of them, and this may be an opportunity that we can't afford to let slip by.
Question time: If .NET does succeed in a big way (and I fail to see how it can't), and is ported to non-MS OS's other than Linux... where does that leave Linux? Stuck sadly on the sidelines, as corporations and developers move to a standardised language on a standardised runtime structure whereby they can write programs that run anywhere... anywhere, that is, except Linux, because it's developers didn't want to touch anything that had been anywhere near MS. I don't think Microsoft are going to produce a Linux port... on the contrary, they're probably hoping nobody does. Our own ignorance and contempt of the .NET offering could be our undoing.
.NET isn't rocket science. People have written compilers before, and people have written class libraries before, and people have written runtime environments before. It's fairly simple stuff in that respect. But it's simple stuff done better than most other simple stuff, including our simple stuff, and it's standardised simple stuff.