On 27/10/06 11:23 +0530, Baishampayan Ghose wrote: <snip> > I really fail to understand the reason behind your fascination for BSD / > MIT style licenses. Do you really think Linux (the kernel) would have > been as powerful as t is now had it been released under, say the BSD
Yes. Linux happened at the right time. In case you didn't know your history, the original BSD group was sued by AT&T for releasing BSD in the late 80s/early 90s. The suit was eventually won by the BSD hackers, but they lost crucial momentum in the early 90s (till ~ 1994 or so). After ths, the BSD project forked, with FreeBSD focussing on x86, wile NetBSD focussed on portability. > license? Exactly why do you think FreeBSD doesn't support half the > hardware that Linux (the kernel) supports today? Even 5-6 years back Because Linux ran with the PC, while BSD ran on far more servers. Until 2.6, the BSD kernel was far superior to Linux. With 2.6, Linus had resources from IBM and the NSA thrown in to help, making it take a slight lead over FreeBSD 5.x. Also, FreeBSD 5.x was the first BSD version which had kernel threads, and was basically an experimental release (think Linux 2.5 quality). Today, more developers use Linux and are happy if their code works there, rather than writing portable code. Earlier, developers would write on *BSD at home, and test on Solaris at work, with the resultant benefits of stability and performance. > FreeBSD was considered far more superior than Linux (the kernel), so > exactly what happened to the Linux kernel project in the recent times > and how did FreeBSD lose the race? IBM happened. > Now don't talk about the licenses of Python, PostgreSQL, etc. They are > in BSD style licenses because those projects are relatively smaller in > scope and size as compared to say gcc or the Linux kernel. None can take What does size have to do with it? <snip> > kernel or gcc, you need to give up some freedoms to make sure the > essential freedoms are maintained no matter what. Otherwise you may Do you understand the meaning of irony? > suffer as FreeBSD is suffering these days. Theo de Raadt (hacker > extraordinaire) has absolutely no way to make sure people who use > FreeBSD source contribute back in some way, and thus the only thing he Theo De Raadt is the lead developer for OpenBSD, not FreeBSD. His goal is to ensure that _all_ the code out there is good, regardless of whther it is closed source or not. This is a different goal from RMS, whose goal is to ensure that hackers can always modify the code and make their systems do what they want done. > can do is cry out loud and beg people for code and or money. A lot of GPLed projects also ask for donations. Keep in mind that OpenBSD has avoided a lot of security exploits because of their insistence on source, not binary blobs. <snip> > in any sense and yet they are not so simple. What you need to understand > is that the GPLv3 text _is_ legalese, and legalese is never simple. > The problem is that legalese looks like English, but isn't. I am sure that the lawyers will actually understand the GPLv3, and the preamble will explain the intent to the non-lawyers out there (for those who actually read licenses). Devdas Bhagat -- http://mm.glug-bom.org/mailman/listinfo/linuxers
