DesktopBSD 1.0 Final Released

Don Church writes "DesktopBSD is reporting that the 1.0 Final of DesktopBSD was released today for both 32-bit and 64-bit x86 architectures. This cutting edge FreeBSD derivative now includes KDE 3.5.1 and a host of tools designed to make the BSD experience more palatable to novices. The DVD release even includes Amarok, Firefox and other popular software ready to go. They are offering downloads via several mirrors or the official torrent."

Comments

Re:Ready for the desktop?

[squiggleslash]

You think this is bad? Microsoft says Vista will need a "modern" CPU. That means it should run on a Power Macintosh G5 right? Well, if you click on that link you get to this, which in turn gives you links to Intel, AMD, and VIA CPU thingies. And what are these CPUs that, say, Intel (I think it says "Intel inside" on my Dell, but doesn't that mean I have a Dell CPU?) has? Well, on "Desktop" platforms (another link) it says I need a "Intel® Pentium® 4 Processor 600 sequence with HT Technology and Intel® Extended Memory 64 Technology."

I don't know about you but all this stuff about HT Technology and stuff is very confusing. Do I have that?

This just proves that Vista is unready for the desktop. I guess that's why they cancelled it. Har har! Har har. Har, har. *sigh*

Seriously, what exactly is DesktopBSD's website supposed to say? The thing you quote seems reasonable to me, anyone who doesn't understand it is unlikely to find any way of wording it useful anyway, unless it was worded in such a way that'd make it useless to an actual computer professional.

It's not like they'll be installing it. They'll be asking us to do it, as usual.

BSD could beat Linux to the desktop

[teslatug]

FreeBSD could still beat Linux to the desktop just because it's standardised on what comes with it, and you could release packages for it a lot more easily. What's lacking is hardware support (which is even more miserable than linux), and desktop performance. If they worked on desktop performance, I think they could easily get drivers by porting them from Linux. I wouldn't mind running FreeBSD on my laptop if only they'd get the performance right. I have actually dual booted FreeBSD and Linux on the same box, both running the same version of KDE, and FreeBSD is just dog slow compared to Linux, which isn't that fast to begin with. Sure KDE can be a hog, but it's either more of a hog on FreeBSD or FreeBSD just doesn't pay attention to a desktop user's needs.

Re:BSD could beat Linux to the desktop

[Billly Gates]

Are you sure that applies to FBSD 5.x and above?

I used to be a FBSD fanatic. 5.0 turned me off and 6.0 made me leave. Its I/O and threading is slower than 4.x and its much less stable not to mention my hardware worked fine with 4.x but has issued with 5.x and higher. Strange indeed?

Linux has improved with low latency timers in the kernel which make it alot faster than earlier versions. Version 2.4 and FBSD 4.x it was a no brainer on which was faster and that was FBSD. But times are changing.

I still have hope in the dragonflyBSD project.

Re:BSD could beat Linux to the desktop

[MichaelSmith]

However, the FreeBSD system seemed to stand up better to high loads than Linux

Not sure about FreeBSD, but I have noticed that NetBSD performs better than Linux when resources are scarce. I tried redhat, and then NetBSD on a very old laptop which I got second hand. NetBSD was more responsive under heavy load.

I put this down to history. BSD had to function on very slow computers in the 80's before linux was written, so the kernel is written with different assumptions about resources.

Re:BSD could beat Linux to the desktop

[the chao goes mu]

Regarding throughput, I know in terms of network this is true. BSD reaps closed sockets almost immediately while Linux (and Solaris, the other OS with which I am familiar enough to comment) tend to wait a long-ish time before cleaning up old sockets. This can lead to real problems in machines handling messaging and other, similar tasks, where there are huge number of rapid, small data packets coming from a large number of sources. Guess it isn't exaclty a "throughput" optimiziation after all, more that BSD is better as a server. Which means this is completely off-topic. Nevermind.

Re:BSD could beat Linux to the desktop

[Eivind Eklund]

Your assumption is correct - we've generally optimized for throughput on large loads. Traditionally, our #1 benchmark has been "How fast can this make world?", meaning a full build of the entire system. Benchmarking systems was able to measure this at an accuracy of better than +- 0.1%. That's pretty damn good: To get that kind of accuracy required GPS clocks (accurate to better than 1ppm), cache preloading, and extreme temperature control of the machine.

I can't say how things are going right at the moment (I've been mostly offline from the FreeBSD development process for a couple of years), but when I "left" there was a number of things going on that should improve this: Preemtable kernel, pluggable schedulers, etc.

I think the ULE scheduler would give you more of the same performance curve as Linux, and it might be more suitable for desktop work. In general, I've found FreeBSD's performance fine for my personal desktop needs - it's been snappy enough, and my only issues have been when web browsers or similar eat enormous amounts of RAM and I/O capacity. I've not tested Linux for desktops for a good many years, so I don't have direct experience to say how it compares in practice.

Eivind (FreeBSD developer "in exile").