Andrew Doran and Jared D. McNeill have announced in a mailing list post that they are starting a NetBSD Desktop Project with the goal of: “Given a NetBSD CD and a reasonably modern x86 computer, make it possible to install a useful desktop system in under 15 minutes, responding to only a few prompts in the process.” Initial plans are being formulated on the project wiki page.
I would love OpenBSD to do this too. It would probably require a lot of work if they wanted to keep everything as secure as the core (most likely).
OpenBSD isn’t a desktop OS.
Neither is FreeBSD or NetBSD, but they can both be made into very nice desktops if you wish. It’s foolish to classify one os as “desktop” or “server.” OpenBSD, as it stands now, is more suited to servers. Put a DE on top of it, write some configuration utilities, and work on ACPI and guess what? You’ve got a desktop-oriented os. OpenBSD is not currently oriented to the desktop. That does not mean it’s “not a desktop os” inherently.
I thought one of FreeBSD’s goals is the desktop.
Here is how I remember it:
– FreeBSD: Desktop, more compromising (e.g. binary driver)
– NetBSD: Simplicity, clean code, portable
– OpenBSD: Security, Theo de Raadt’s ego
Yes… but it can be used for a desktop just like any other OS with a UI.
OpenBSD, like all the BSD systems, is a multi-purpose OS. As an example from reality, I’m using FreeBSD exclusively on my desktop since 4.0 without any problems. NetBSD has the same possibilities. Agreeing to this fact, “starting a desktop project” would be a bit misleading, because it’s already usable as a desktop. Some kind of preconfiguration and adjustment (like DesktopBSD and PC-BSD did with FreeBSD OS) could of course make things easier. The NetBSD Live 2007 live system CD, for example, already has some stuff to be considered belonging to a desktop system (KDE environment and accessories).
As cool as that would be, I don’t see OpenBSD doing it. Say what you might about Theo, but he knows what he wants and sticks to it, and in OpenBSD’s case, that’s a minimal core with the user tweaking the system how they wish. Naturally there’s nothing to stop someone else from doing it, but I don’t expect the OpenBSD team themselves to make it a priority.
You can use OpenBSD as a primary desktop OS. Just tell it you want to use X during the install and then install Gnome, KDE, Flux/Openbox, or almost any window manager from packages or ports. Is it as slick as Ubuntu or Fedora for new users? No, but those aren’t the primary users of OpenBSD.
The thing is, if the OpenBSD project wanted to make a desktop version of their OS, they would most probably want to keep their ‘security’ focus so they couldn’t just slap KDE or Gnome on top because as shown here:
http://lwn.net/Articles/319072/ those desktop have a serious security issue..
I’ve read a bit through the mailing list and looked at http://wiki.netbsd.se/Desktop_Branding
They just want to do an Ubuntu clone with a NetBSD core. They’ll use GNOME because it’s supposed to be the “industry standard”. And they’ll sacrifice the famous cross-architecture idea of NetBSD to go x86 only.
Here’s my take:
I think they want to go mainstream at all costs. Becoming an Ubuntu clone will not make Desktop NetBSD successful. People will go for the better supported “original” instead.
NetBSD hasn’t even a recent GNOME build in its repository if http://distrowatch.com/table.php?distribution=netbsd is to believed, let alone KDE 4.x (they re still stuck with 3.5.9).
NetBSD shouldn’t care what they think the “industry standard” is, because in reality it’s Windows XP. NetBSD’s strength is that it runs almost everywhere, including low-end hardware. Desktop NetBSD should use a really lightweight DE, like Antico. And it should support more that just one CPU architecture.
Desktop NetBSD team, please don’t spit on NetBSD’s strengths.
Who says they’re going to drop everything else? You can’t blame them for wanting to support X86 and X86_64 first. Those are, *gasp*, the most likely platforms the majority of users will be using and especially the most likely platforms for desktops at the moment. I didn’t see where it said they were dropping support for all other architectures, or replacing the core NetBSD with this desktop flavor. That’s like saying because Ubuntu only supports x86(64) that Debian will drop support for all other architectures too.
I never got the impression from NetBSD that they want to go mainstream at all costs. Come on, what’re you ticked off about? That they’re not using your favorite DE, or don’t support one of your machines, or something? Or do you just hate *BSD, as quite a few on here do?
“Desktop NetBSD” (I have to put quotes around the name, because you obviously confused something I wrote) will be x86 only, which is not in line with NetBSD’s spirit.
http://mail-index.netbsd.org/current-users/2009/02/08/msg007927.htm… says “x86 only”, not “beginning with x86”.
I never said that “plain NetBSD” drops architectures.
Read my post again. I never said that “plain NetBSD” wants to be mainstream at all costs. However, “Desktop NetBSD” wants to be mainstream. “Desktop NetBSD” copies Ubuntu so much that it even uses the same theme, just the Ubuntu logo is replaced with NetBSD’s flag.
Being an Ubuntu clone is not NetBSD’s philosophy and I don’t understand why “Desktop NetBSD” wants to be one.
I apologize, I must have confused what you were saying. The way it appeared to me, you seemed to be saying NetBSD in general, and not specifically the desktop variant. Perhaps they’re going with the Ubuntu look and feel thinking it’s a good place to start. Ubuntu, love it or leave it, is the buzz of the Linux world right now. While it may not be NetBSD’s philosophy to copy Ubuntu, honestly it’s as good a place to start, or at least to put together a prototype, as any other. NetBSD hasn’t exactly ever been huge on interface design, so perhaps they want functionality first and will worry about the looks later. More likely they’ll implement the functionality, and let the community worry about the looks.
If the project takes off, it will be ported to other architectures. But it seems to me this is a solid, general starting foundation.
[q]They just want to do an Ubuntu clone with a NetBSD core.[…] I think they want to go mainstream at all costs.[/quote]
No, we don’t. (Speaking on my own behalf:) Keep two things in mind:
– First off, NetBSD has traditionally been a general purpose operating system that works on any thing from handheld embedded devices to high-end servers. It takes time and work to keep it strong in all areas. This subproject is a good effort to stay sharp on the desktop side.
– Second, NetBSD is used a lot for embedded device development. If we don’t keep up with the expectations of developers desktop-wise, people will build NetBSD appliances on, say Ubuntu. Of course, it would be a lot more comfortable to do such development on NetBSD, provided that it gives a good desktop environment.
[quote]NetBSD’s strength is that it runs almost everywhere, including low-end hardware.[/quote]
And it still will.
[quote]Desktop NetBSD should use a really lightweight DE, like Antico. And it should support more that just one CPU architecture.[/quote]
You can still use a lightweight desktop. Part of the project is IIRC to make it easier to install packages at installation time so you won’t be dropped in shell-land with very little installed. Both users of GNOME/KDE and lightweight environments profit from these changes.
And what’s the talk about making GNOME the default. GNOME as an option is fine, even prominently placed. However defaulting to GNOME works against low-end devices. The very least the “Desktop NetBSD” project should do (IMHO) is a DE selection screen similar to this one: http://en.opensuse.org/Image:11_1-install-006.png
Of course, to be in line with NetBSD’s philosophies there should be other options, like IceWM and Antico added to that list.
(To make things clear: I also don’t want NetBSD to become an openSUSE clone, I just like the wording of that screen.)
If they released a desktop NetBSD with KDE 3.5 as the default, there would be a mass of stubborn KDE 4-haters switching to NetBSD 🙂
Those people would at least expect to get updates for KDE 3.5. NetBSD is still on 3.5.9.
Oh I know the OpenBSD team would never do that. They would probably just laugh at my request. They will always be about security and thats fine. No need to waste resources on other things. Hey, I can dream can’t I?
> “I can dream”
You aren’t the only one. Rather frightened by ease with which I put my financial details into browsers on mainstream operating systems …. I tried using OpenBSD as a secure desktop.
It did work. Getting the details sorted out was time consuming – like recognition of my unAmerican keyboard. Flash and Java just weren’t going to happen.
I still dream and slightly wish Mark Shuttleworth had started with OpenBSD rather than Debian; but that’s life.
Reading at the NetBSD mailing lists about this topic, I saw it is creating a lot of controversy between NetBSD users, devs, fans, etc.
Anyway, it is a good idea [though I disagree with the choice of Gnome instead of KDE] if the NetBSD “personality” is not removed.
And its “personality” includes:
* Multiplatform
– They want to implement support for x86, and though
its not a bad idea doing this “AT FIRST”, I could
assert that the “desktop support” for the other
architectures will never occur [NetBSD is not an
OS, let’s say, rich in development resources].
* pkgsrc
– pkgsrc is IMHO one of the best package manager
[next to the other BSDs ports and Gentoo’s Portage];
shipping NetBSD with no pkgsrc goes nowhere, I
think.
Well, not worth much I know, but I’m chiming in to say I’m glad they’re using GNOME instead of KDE.
Me too, by far!
I’d love to see it too – from what I have also seen, BWI (Broadcom wireless driver) has been merged and part of the NetBSD 6. 0 changes:
http://www.netbsd.org/changes/changes-6.0.html
I’d love to see BWI ported to OpenSolaris – I’ve emailed the address associated with the Bradcom bug report on defect.opensolaris.org – it has been over a week and no reply.
This is awesome news and a great way to make NetBSD stand out from the crowd even more.
But…I think this would still be too difficult for normal people to use, no matter what. And it would hardly make any difference to current NetBSD users who, I suspect, will still use the traditional method of installing the DE and desktop applications because they can install exactly what they want, and because they may think the traditional way to install software is the right way, albeit more painful. So…Who will this desktop project benefit? I don’t know.
Some of them on the mailing list seem worried about fragmentation. Makes me wonder if BSD in general might not have been better off if it hadn’t split off in three directions. Maybe have special versions of for, desktop security and networking, but still just be one BSD.
Great idea! Let’s get Theo, the FreeBSD guys, and the NetBSD guys all to sit down in the same room and work out a merger plan. I wonder how powerful the explosion would be?
Edited 2009-02-17 06:59 UTC
Why not? The project is very good.
DesktopQNX.
And a DesktopOS/390. 🙂
For the people whining about how this will change NetBSD or how this is not the direction NetBSD needs to be going in…what are you people smoking? This is a good thing.
NetBSD is a great platform, but it has a very small user base. An effort to make a variant of the traditional installer that installs a pre-configured desktop system would help to increase this user base.
Not only will this help to increase the user base, but the traditional installer will still be available!
I see no reason to complain about this. I think its a great idea and I would even like to help with what I can.
One more time, news relevant to the OS world gets on Page 2, even when there are a lot of non-OS related news in Front Page, like hardware reviews, Pirate Bay, iPhone, just to cite a few.
Edited 2009-02-17 16:50 UTC
Would love to see my Quadra 650 running GNOME.
OK, just kidding…
As I commented out in the mailing list, I once rejected the idea of a Desktop NetBSD, arguing the project would not lose the focus it already has.
On the other side, maybe this could bring users to NetBSD. Not that this fact isolated is a good thing, but more users can mean more developers in the future.
I think it’s very hard for someone to become a developer of the system if it isn’t already a user. This may sounds obvious, but it’s not.
Just give me 1 app per task (like Zenwalk) + wireless connectivity and I will take care of the rest.
I can not wait!!!!
-2501