Submitted by “The first RC build for the FreeBSD-9.0 release cycle is now available. ISO images for the architectures amd64, i386, ia64, powerpc, powerpc64, and sparc64 are available on most of our FreeBSD mirror sites. One of the many new features in 9.0 we would like to be tested is the new installer, so we encourage our users to do fresh installation on test systems. Alternatively, users upgrading existing systems may now do so using the freebsd-update(8) utility.”
http://www.freebsd.org/relnotes/CURRENT/relnotes/new.html#DISKS
Is it just me, or do the recent releases of FreeBSD (especially the .0 releases) since 7.0 seem fairly thin on new features? I am a DragonFly BSD committer and the general consensus in the past has been that where we are behind FreeBSD, we are only behind due to a manpower shortage. Yet lately it seems as if our release notes are as full as theirs are, is the FreeBSD labor pool drying up, or are they just doing a poor job making sure everything interesting makes it into the release notes?
The more advanced a product is the fewer changes are necessary. It then become mainly a matter of polishing the existing features.
That’s a clever response, but it is not the case here. How about getting serious?
Capsicum, DTrace.. it’s just you.
See http://ivoras.net/freebsd/freebsd9.html
That is what I and probably the parent was looking for. It seems odd that its that difficutl to find, and not hosted on Freebsd.org. Compare that to the sparse official release notes.
@DOSguy, that sounds plausible. I was also thinking it may be the case that since the kernel API’s have stabilized since 7.0 that the team is being more diligent about MFC’s and the features are simply more spread out between point and major releases now. It is hard to be certain since it is purely subjective.
@jefro/@celt, you both missed the point. New features do not necessarily a great product make, but they have traditionally been a pretty good indicator as to how rich the labor and intellectual economy of a given open source project is.
@laffer1, knowing how much work it is simply to keep up with the modern age wrt drivers and third party software support, I do not envy you guys. That said, I would love to see more collaboration between our respective projects, though I am not sure how exactly one goes about putting a foot forward in that regard. There are a number of us who are effectively red-headed stepchildren of the BSD community after all. It couldn’t hurt to establish a dialogue to see if there are ways in which our work can overlap beyond periodically porting and sync’ing code.
@joekiser, I know a large number of FreeBSD developers and users that left the project for greener pastures (Linux, typically) during the SMPng/5.x development cycle. That timeframe I can certainly comment on. I haven’t really kept a well-tuned ear to the FreeBSD community since sometime in the 6.x development cycle, which is why I raise the question. I am really genuinely curious if there has been some exodus of developers (did Apple hire them all?) or if things have otherwise slowed down, or if it is simply my mistaken impression. Despite what a couple have said in this thread, I am certain that at least _ideas_ for new features and improvements are not lacking.
New features in the latest 8-STABLE / 9.0-RC1, that I know of off the top of my head:
USB3 support.
GEOM-based fake-RAID support (all those software RAID features on motherboards)
ZFSv28 support.
bunch of stuff around the syscons driver
802.11n wireless support
This page shows more:
http://ivoras.net/freebsd/freebsd9.html
That is a much better list.
I’m not sure what I was down-moderated even though I gave the list he was looking for.
And I still assert that I’m quite impressed by the FreeBSD9 features (DTrace, Capsicum, etc).
Let’s look at Linux 3.1, what’s are the interesting new feature?
The support of OpenRISC, a RISC CPU implemented currently only in FPGA? Bleah.
It’s not the same project, not the same release process, but still..
Although I don’t think the last few FreeBSD releases where light on new features, you’re probably not wrong either.
I remember reading an interview with the new FreeBSD project lead years ago. In the interview he said they where going to put out major releases more frequently. By looking at the FBSD timeline ( http://en.wikipedia.org/wiki/FreeBSD#Timeline ) it seems they did just that. ( now 2 years between .0 releases versus ~3 years between releases before 7.0 )
Considering that the number of contributors probably didn’t increase a lot, it would only seem logical that new releases contain less new ( big ) features.
Edited 2011-10-24 13:41 UTC
Yeah, it’s just you…
Why should the quality of code be measured by changes, features and new bells and whistles?
What has that have to do with anything?
FreeBSD isn’t ahead or behind anyone or anything. I deploy many FreeBSD servers as a professional, and it has never failed me like many “feature of the day” nonsense that the rest of you are judging your success by.
Leave FreeBSD alone, you want the latest nonsense and all the instability that comes with it – go run Windows or Linux for that matter.
Somewhat concerned a DragonFly BSD committer would post such a comment.
troll…
It’s just you.
DragonFly isn’t short on manpower compared to MirBSD or MidnightBSD. You might know us as 5th and 6th place in this little group. Don’t complain about manpower again. It’s weak.
It’s not just you.
There are several current and former FreeBSD users who believe that the project has stagnated, some going so far as to say that if you took out all the code that is being imported from OpenSolaris, the project looks like NetBSD did a few years ago.
It’s also no small surprise that a DragonFly developer would point this out, as this project and OpenBSD seem to be where all the innovation is happening these days.
Not sure the people who depend on this need new features. The entire bsd thought seems to have been quality over speed and features. It is more intended for server work. The people who have been working on this for maybe decades have done some great work.
Well, speed is a feature, so is quality. 🙂
Honestly, I’m always impressed about the “hidden work” that you’ll find when working with FreeBSD. A famous example are the documented and tidy OS sources. But you’ll also find that manpages are kept in a very good state – which you’ll be thankful for when needing to research something offline, be it a system binary’s command options, a kernel interface, a library call, a file format or a maintenance procedure; this kind of feature traditionally appeals to developers.
For “serious production use”, FreeBSD doesn’t offer a trade-off between speed and stability. You get both. (Note that this statement depends on the kind of your workloads!)
Maintaining features, no matter which ones, deserves honor for those who actually do it.
FreeBSD is a multi-purpose OS. I agree that its main use is servers, but you can use it as a versatile desktop OS too (as I’m doing exclusively since 4.0). It depends on the hardware you use and the software you want to run. In this regards, FreeBSD provides a powerful foundation for ported applications.
Well, installer’s part of disk partitioning is terrible and it doesn’t do its thing right. I had to go down knee-deep in fdisk and label … [NOTE: I am using manual layout, not the automatic one]