The FreeBSD team has started the beta cycle for FreeBSD 8.0 by releasing this new version’s first beta. It was intended as an evolutionary release with few groundbreaking features, but as time went by, more and more interesting features crept in. As always, the What’s cooking for FreeBSD? page has a good overview.
I’ll be looking forward to installing 8. DTrace on FreeBSD is by far the most promising version of it outside of Solaris (At least, more so than the Linux version. Can’t personally comment on the OSX version).
If LLVM really replaces GCC as the base system compiler, I’m curious how that will affect buildkernel and buildworld times as well. Apparently this isn’t scheduled for 8.x though
If the LLVM benchmarks are anything to go by, it should improve build times considerably. LLVM 2.6 is scheduled for September 21 release, so I am pretty confident that by at least 2.8/2.9 that we’ll see it having the ability to compile all the ports cleanly. Right now with some small modifications it can compile quite a number of the ports already.
That’s going to be a really exciting release, although I don’t like freebsd particularly.
By the way, perhaps the ‘front page’ and ‘page2’ have been mistakenly swapped ?
Cheers
Please first read somewhere explaining what front page and page 2 means. Those aren’t what you think they are.
Forgive my ignorance, or lack of a knowledge in this particular area [FreeBSD], but which version is newer: 8-current, or 8-beta1? and I’m talking about ISO ones.
8-current is always “newer” because it’s the development branch.
While I haven’t followed FreeBSD development that closely in recent years, the beta stage of the 8.0 release — in all odds a major release — surely caught my attention.
The summary was okay so here goes my top #7 from it:
* CLANG / LLVM. It is always a pleasure to see a big project opt for a different compiler (suite). However, as a personal disclaimer: I hoped that the BSD community would have had a wider interest in PCC, but sadly this seemed to be just high hopes.
* Jails improvements. Needs no explanations; one core of the “power to serve”.
* Xen DomU support. Some would say: finally!
* New USB stack. While rewriting is more of a Linux thing, I look forward to this. How was it that Augustsson’s once so elegant and leading USB implementation lead to so many problems across BSD systems in recent years?
* DTtrace. I do not hide my love for Sun’s brilliant technical work, and why should I? Like with ZFS, others are still years behind.
* SSP. I have no idea why FreeBSD hold back modern security technologies for so long, but this one hopefully leads to acceptance of other solutions as well (namely ASLR).
* Virtual IEEE 802.11. While real production usage may be questionable, this sure sounds cool for home use!
What else? Is ZFS is ready for production?
Edit: typos. Odd sentence structures; leave it be.
Edited 2009-07-11 19:25 UTC
Why do you have interest with PCC? It lacks C++ and it looks like PCC isn’t going to add C++? FreeBSD base system has a few C++, so I think PCC is not useful for FreeBSD.
“the GPLv3 is a big dissapointment for some users of BSD systems (mostly commercial users who have no-gplv3-beyond-company-doors policy) […] GCC 4.3 will not be imported into the base system (the same goes for other GPLv3 code)”
So a community-based project rejects FOSS code, because some corporate di*kheads say so? Let’s go through the implications:
* Worse performance, because newer GCC versions produce better machine code.
* No Exchange support at all (through libraries provided by OpenChange)
* No even somewhat recent Samba releases.
* No OpenOffice (it’s LGPLv3)
… and so on.
That decision probably hurts FreeBSD.
This sounds pretty insane. Not that I’m not excited about LLVM gaining mindshare, but – what does the license of the compiler have to do with the license of produced binaries? Do the corporate bigwigs actually plan to ship *compilers* in their tivo-ized end user devices?
All of those will never and never be in FreeBSD base system, even if those are BSD or public domain license. The ports tree are fine to have GPLv3 and we already have those in ports tree. I think someone is working on license framework for ports tree.
Don’t spread FUD, it’s sleazy. Those softwares you speak of will be available through the ports system.