Slashdot Mirror
FreeBSD 5.4 Review
gammelgul writes "Jem Matzan has written a review of the new FreeBSD 5.4 release on NewsForge. He writes about enhancements and the 64bit edition of the OS."
← Back to Stories (view on slashdot.org)
← Back to Stories (view on slashdot.org)
Well, I wouldn't normally respond to this but there is one bit of truth here. I have witnessesed the 5.* series corrupt its own file system several times. Usually it tends to fail to write correctly to the master.passwd file thus corrupting it. But I have also seen times where the binaries have been corrupted as well.
That being said, FreeBSD is not all bad. The 4.* series clearly is a top notch OS. I have seen the 4.* take slashdot-type spikes of traffic as a server and not blink an eye when similiar linux based OS have crumbled.
Quality Hosting e3 Servers
This is sad. I too remember fondly the 4.x days. FreeBSD hasn't made the transition to these "enterprise" features like the ULE scheduler, and getting out from under the "big lock" SMP.
The 4.x series is still alive and well, but the writing is on the wall. If the FreeBSD 5.x series doesn't start fixing some of those show-stoppers, it risks becoming irrelevant.
NetBSD and OpenBSD seem to have found their niche to a certain extent. I suspect that some network equipment vendor like HP will start putting OpenBSD in switches and routers. It seems like Cisco and IOS have the most to lose from OpenBSD gaining ground.
NetBSD seems to chug away at its own pace making solid incremental gains. They tend toward evolutionary rather than revolutionary changes. When they do make more revolutionary changes, they tend to include them in small numbers, and only after a long period of vetting in the current branch.
The laundry list of improvements to FreeBSD 5.x makes me wonder if that project didn't bite off more than it could chew. That the BSD faithful are starting to raise questions about the long road to stability with FreeBSD 5.x should be a warning to other Open Source projects to stick to regular release cycles with clearly defined and narrowly scoped improvements.
I suspect that FreeBSD development may have slowed somewhat due to the "fun factor" waning. Announcing Big Gigantic Changes can be good to generate enthusiasm in a user base, but it can be oppressive to the poor developers caught doing the work. Lots of small, discrete tasks can be fun for experienced developers, and a good way to snag novices.
Despite these problems, FreeBSD has very recently been a very vibrant project. They have traditionally had a level of coordination rarely seen in any other Open Source project. I think this can work, but FreeBSD 5.x may fall into the "lessons learned" category.
Or, as I mention in my blog, Darwin may see a surge in popularity following Apple's Intel announcement.
-Peter
. Penguins Surely Ca
I can tell you that I'm running FreeBSD-amd64 since 5.2 and it IS stable on this platform.
Perhaps Apache 2.0 is not as clean as one might think? That's why I'm still using Apache 1.x.
On a sidenote: I just finished compiling openoffice1.9m107 a few days ago (now that the port is updated to m109, I'm there compiling again) - with KDE support. Running it under KDE with native widget support is simply amazing speedwise. Startup times don't change much, however, opening various dialogues (options for instance), the help, etc. is instantaneous - just like konqi when preloaded. Not that it was very slow before, but still, once oo starts up, it is lightning fast using the native widgets!
I uploaded the m107 package to my ftp server (see link in my signature) - you may try pkg_adding it, but read the README.txt before that. And while we are at it, one majore advantage of FreeBSD (that a review should include) is its excellent software support. I can't think of a linux distro with such great support for oo.o as FreeBSD (not by default anyway). The latest binaries distributed by openoffice.org are at m104, and I couldn't find (on google) the source downloads for m107, much less m109 today, except at good-day.net of course, which is pretty much overloaded recently, because ports uses it for source downloads, and it also distributes precompiled binaries: m107 builds are ready with all language packs! So, if you want only the english (or KDE) version, use this link to get it. I just began compiling m109 an hour ago, so in a day it will also be uploaded.
yes, technically EM64T is the equivalent of x86-64.
it might not be as microarchitecturally efficient, but it is 100% compatible (as far as i know).
hm... what you mean by referencing sources?
i had the behavior i described (syscalls burst then panic) repeatedly, which was corrected by rebuilding-reinstalling the i386 world and kernel.
no problems since then.
Apache (2.0.54) wasn't even doing any heavy-duty php/perl, just static content, SSI and some proxying.
alongside apache there's a lightweight httpd serving exclusively static content ( called nginx. it's not threaded, but rather based on kqueue and is extremely efficient).
the box was underloaded most of the time, cpu usage not exceeding 25%.
those massive amounts of syscalls must originate from somewhere in the libpthread, though i haven't had time to investigate - it tended to happen in peak daytime hours and i needed to put the server back up asap.