Quote:
Originally posted by Yves M:
Occasionally, I was thinking that it might be great if Linux buck up it's efforts and provide a good API so that more programmers will program and create programs for the Linux OS.
More importantly, make development as smooth as in the Windows OS. I mean, just pay $$$ and u get the development tools, like in the case of Visual Studio. But in Linux, if I'm not wrong, u have to do some complex license agreement or something. Aye.........
I really hate the idea of those new Microsoft technologies, starting from COM to Managed C++.
These are just very artificial add-ons, the truth is.
Before we even know of such stuff, we don't even need them!
It's like : if a tree fells in a forest and u wasn't there, will u even hear a sound?
No.
Someone please sponsor me and Mick and some coders here with tickets and hotel accomodation to the Microsoft HQ in Redmond. We plan to protest in all our glory of being naked outside the dreaded company.
The slogans are as follows :
---Down with VB.net!!!!
---Rid Artificial Coding!!!!
---We want C++! We want C++! We want C++!
---Give us back the old days!!!!
---God destroy Microsoft!!!!!
---The Lord is with us! Microsoft DOWN! Microsoft DOWN!
Have these glorious slogans written down on a very large cardboard with white background and black markers.(black fonts)
I hear you. COM is an a pathetic attempt at making an OS object aware. You want the problem fixed? Write the OS in C++, do that by writing a self sustaining assembler/compiler that relies on no OS to run. Specificially designed to write an OS. It could be done, to install Linux you have to use a bootstrapped GCC to compile it. You'll never see it from Microsoft, but I can almost guarantee it will happen for another OS in the future. (If noone else does it, 10 - 15 years down the road I will.) But I mean a whole editing environment (the IDE, etc...) and not just compiler/assembler. Only disadvantage to that, is the whole thing would have to be written to the particular CPU it's going on. And that would require writing a compiler/assembler for ANOTHER OS that would 'cross compile' for that purpose. (I wouldn't do it in ML - NFW. The fact you'd have to know the architecture and it's ML and the definition of the assembly instructions (in order to map these to the correct ML output) it would be much easier to break it apart and just write that cross compiler/assembler. Then it could 'logically' written.) Devil's advocate - if it were today - write the 'cross compiler' on some type of X86 (on a Linux or Windows OS, so you have an easier job on the development environment) - learn the architecture of 'pick a CPU' and write the code necessary to implement the compiler/assembler. Then, after probably 3 years or so after all the bugs are worked out, you have an independent compiler/assembler for a particular CPU type. Plus, that would give hardware manufacturers some incentive to create ML compatible machines. ;) I'm only turning 22 next Wednesday, and have 2 1/2 years of C/C++ and some assembly experience (I ain't near as good at assembly as I am C/C++, but I am knowledgable) - I couldn't do it right now, but in 10 - 15 years I will be experienced enough to do so. Knowing how young I am, and already amassed as much as I have without a single course in my life - I know I have unlimited potential. I've been told that all my life, only just recently began tapping it. I ain't gonna waste the opportunity.