as a developer of my own private little OS I have to agree. Problems when developing an OS (or at least problems I found out to be problematic:) are...
Device Documentations: Writing your own 32-Bit OS means that you've got to go away from the BIOS and write your own device drivers. However its quite hard to obtain documentation about your hardware. Sure, there is the VGA standard and Intel/Amd still release their documents for free. But if you try to go deeper into the mud then you discover lots of dirt. Some of the docs are only avaiable for money (Like the PCI docs), while others are not avaiable at all (Nvidia NEVER releases any information about how to program their chips!)
Having a plan: After all, an OS still is a quite complex piece of software. While you are able to hack a small one with few features and few device drivers quickly, its getting more and more complicated when adding features. So you have to step back and think, probably for weeks, planning most details of your OS. Lots of people start coding right away only to realise some weeks later that most of the code has to be rewritten.
Compatibility: I absolutely agree with the poster above me. Implementing a standard like POSIX cannot be done in a couple of days, not even weeks. Its easier when trying to write a POSIX OS right away, but implementing standards after the design of your OS is complete results in a lot of work.
10 Processors for 1000 Users, that makes 1/100th processor time for each user. I don't know how fast they are, but if we compare them with, say, 2GHz Athlons, and take 1/100 then we've got a 20MHz Athlon for each one currently busy on the box. OK, not everyone logged in uses his processing time completely, say every 3rd does. It is compareable to developing on a 60MHz Athlon over the internet - not very fast I guess. Then there certainly are limitatios of what you are allowed to change on your configuration. Developing is only possible when connected to the iNet. And don't forget that Big Blue is watching you all the time...well, all in all I prefer my 500MHz Box right here. Faster, more flexible, lower latency and surely not observed by anyone.
They'll have to start writing good software then..
on
Shared Source?
·
· Score: 1
If they don't like the GPL, well, noone forces them to use code based on it. But if I, as a volunteer, write code in my free time, I have the goddamn right to do with it whatever I want.
Writing code and releasing it to the public doesn't give any benefits in the form of money or sth. to the coder. But if others improve it and help him, or write some other free software, he gets his reward in form of good and free (of both pay and source) software. Thats what the GPL is all about. Its a giving and a taking.
Now if companies were allowed to take the code, improve it but keep their changes the original coder will get nothing. No improved code for sure. He might even see his very own program, written by himself, with only few minor changes, sold by a company, and he would have to pay for it! This would be a "giving" by the coder and a "taking" by the company. Why should anyone keep on writing "free" software then?
If MS is unhappy with the GPL then its their good right. If I was MS, I wouldn't like it either. But saing that the should be allowed to use the software withour crediting the original, unpaid author is painful. If they dont like the GPL then they should finally start writing good software by themselves and stop copying from others.
Slashdot Mirror
10 Processors for 1000 Users, that makes 1/100th processor time for each user. I don't know how fast they are, but if we compare them with, say, 2GHz Athlons, and take 1/100 then we've got a 20MHz Athlon for each one currently busy on the box. OK, not everyone logged in uses his processing time completely, say every 3rd does. It is compareable to developing on a 60MHz Athlon over the internet - not very fast I guess. Then there certainly are limitatios of what you are allowed to change on your configuration. Developing is only possible when connected to the iNet. And don't forget that Big Blue is watching you all the time...well, all in all I prefer my 500MHz Box right here. Faster, more flexible, lower latency and surely not observed by anyone.
If they don't like the GPL, well, noone forces them to use code based on it. But if I, as a volunteer, write code in my free time, I have the goddamn right to do with it whatever I want. Writing code and releasing it to the public doesn't give any benefits in the form of money or sth. to the coder. But if others improve it and help him, or write some other free software, he gets his reward in form of good and free (of both pay and source) software. Thats what the GPL is all about. Its a giving and a taking. Now if companies were allowed to take the code, improve it but keep their changes the original coder will get nothing. No improved code for sure. He might even see his very own program, written by himself, with only few minor changes, sold by a company, and he would have to pay for it! This would be a "giving" by the coder and a "taking" by the company. Why should anyone keep on writing "free" software then? If MS is unhappy with the GPL then its their good right. If I was MS, I wouldn't like it either. But saing that the should be allowed to use the software withour crediting the original, unpaid author is painful. If they dont like the GPL then they should finally start writing good software by themselves and stop copying from others.