“After using PC-BSD several days, I was impressed with how easy it is to use. It’s a good desktop OS, and a great way to introduce BSD to new users. The 1.0 release has a few rough edges, but nothing that should scare off prospective users. For the future, I’d like to see something like Synaptic to manage PBI packages and allow users to browse for software without having to visit the PC-BSD Web site, and it would be nice if the site had a little more documentation, but I expect such things will come along in due time as the project matures.”
PC-BSD is indeed an excellent FreeBSD distro. I’d like to see more software in the VERY easily installed PBI format, actually developer and server applications. As a desktop or development/testing server, PC-BSD looks very promising.
Absolutely. Best KDE-based desktop OS I have ever tried: snappy, rock solid and professionally designed. Actually, PCBSD is the first unix-based OS I can seriously consider as usable desktop OS not only for geeks but for casual users too.
Well said, I agree with your comment 100%. PC-BSD rocks!! It is the best UNIX based desktop OS with a real potential. I personally love the PBI system. I use both PBI and the ports system. I wander if it is possible to port DesktopBSD’s ports GUI to PC-BSD?
To the other reader that said put KDE on FreeBSD and you get the same thing. Yea but the thing with PC-BSD is KDE is already put up for you and the installation of new programs is piece of cake so I find your argument invalid. A normal user won’t be able to put KDE on a FreeBSD. PC-BSD is a “desktop” OS so I believe you are missing the point with that comment. That’s why it’s called a desktop OS so that you don’t have to put things up yourself. Another thing, even if you put up KDE yourself, it won’t be the same. In speed? Yes, it will be, but you won’t have PBI (PBI is a must for a “desktop”), the menu structure won’t be the same etc.
Edited 2006-02-15 11:46
I really like (Free)BSD a lot and will certainly install PC-BSD once it’s out of beta and RC stages. I don’t understand why it takes so many months to finish 1.0, considering it’s based on a current release of FreeBSD.
I guess the PC-BSD project is looking for a fine polish that would help gain it mass market appeal? How much extra value does the pc-bsd project add?
Here are the features they claim to add (?) to FreeBSD on their web site:
* Fully functional desktop operating system, running FreeBSD 6.0 under the hood.
* Graphical system installer, makes the process of installing your system take only minutes.
* Support for many native languages.
* Advanced stability & performance that FreeBSD users have come to expect.
* Easy to use package management system (PBI), but also works with standard FreeBSD ports/packages.
* Graphical tools for system administration and support.
* Friendly and helpful support community.
* PBI packages are self-contained, and do not rely on numerous system dependancies.
* Very clean minimal install (1 installation CD), no cluttered mess of programs and dependancies to untangle.
* Online Update Manager – Manually or automatically downloads and installs updates for your operating system, without touching your installed programs.
I’ve just got a magazine cover disk DVD with FreeBSD 6.0 on it. Worth installing? Or will it not give me the kind of thing PC-BSD might give me, meaning a good desktop with KDE? Either way, PC-BSD sounds a very interesting project.
Do yourself a favor and install FreeBSD proper. You can have the same desktop with FreeBSD+KDE. The pbi system that they’re so proud of is not a real option.
1. There are only a handful of packages in pbi format
2. when you run out of pbi’s and try to install from ports, you’ll be the proud owner of a broken system
Do yourself a favor and install FreeBSD proper. You can have the same desktop with FreeBSD+KDE.
Thanks for replying. Will give it a go. I’m a Linux person and this will be very interesting, with lots to learn.
I am a Linux person as well, and have used FreeBSD as my everyday desktop for 2 months in the past. It was great from a Slackware/Gentoo perspective, but ports were weak for upgrading the system. Too much recompiling (even more than Gentoo).
To be honest I haven’t tried PCBSD yet, I plan to one of these days. What I have tried is DesktopBSD, and it looks good to me, I think it deserves more public attention. What it gives you is a “proper” FreeBSD, with a quick & easy installation. Nice way to get started with FreeBSD (be sure to read the handbook – it’s a must). Binary upgrades are preferred to compiling from source (don’t know how well this works, haven’t used it enough).
Now I am rather sceptical about PCBSD. .pbi probably can’t scale much… but if it can and I am wrong, then it’s a step in the right direction for desktop Un*x.
The fact that (I think) PCBSD doesn’t, for example, install bash by default (DesktopBSD does) which is something most Linux users would do, and .pbi ignoring ports, prompt me that PCBSD are attempting to build a desktop OS despite being based on FreeBSD, rather than a preconfigured FreeBSD desktop (DesktopBSD’s approach).
But still, it brings FreeBSD desktop some publicity.
Now: my dream FreeBSD desktop would use apt-get-ish binary package management (handling versions and such) and still keep ports v2 around, somehow. I don’t think anyone is working on this though.
I am a Linux person as well, and have used FreeBSD as my everyday desktop for 2 months in the past. It was great from a Slackware/Gentoo perspective, but ports were weak for upgrading the system. Too much recompiling (even more than Gentoo).
I wander why you say that – I spent two months with gentoo and came back screaming to ports, which gives me the same lean system with 1/10 of the configuration headache. And you can’t have “more” recompiling – when you upgrade, you upgrade the packages you have installed. If you have more packages installed, there’ll be more recompiling during an upgrade.
Now: my dream FreeBSD desktop would use apt-get-ish binary package management (handling versions and such) and still keep ports v2 around, somehow.
Well, pkg_add -r pkgname is roughtly equivalent to apt-get install pkgname. In fact, the FreeBSD ports system (and this feature alone makes it better than any pkg management I tried) makes it ridiculously easy to create binary packages, even from software you installed from ports! a pkg_create -R -b ooo-build* will build an ooo-build binary with all its dependencies in your current directory. This binary will be similar to a .deb – it knows exactly what packages are needed to be installed as dependencies, the pkg_* tools are able to fetch them automatically, etc. This makes deployment of customized, precompiled, optimized (for instance, for i686) binary packages very easy. You can build on one machine, and put the binaries on an ftp, and install them via pkg_add on the rest. Oh, and creating binary packages during installation is as easy as adding a single letter to portupgrade/portinstall: portupgrade -ap. -p will put a binary package in /usr/ports/packages (in fact, it will automatically create the same directory layout you would find on the official ftp repositories for binary packages) if it exists, or in the port directory if it doesn’t.
.pbi probably can’t scale much
What do you mean? What is “scaling” in the context of package management?
I spent two months with gentoo and came back screaming to ports, which gives me the same lean system with 1/10 of the configuration headache. And you can’t have “more” recompiling – when you upgrade, you upgrade the packages you have installed.
FreeBSD ports miss information about versions (package abc requires def >=2.5). So, either you recompile all dependencies, even if you didn’t intend to, and be on the safe side or try your luck and skip it. Also, I think there isn’t a “stable” branch for ports, only “current” exists – that would be equivalent to running ~arch in Gentoo.
As for .pbi, it’s intention is probably to install a few desktop apps e.g. firefox, but the rest of the system e.g. Gnome is still using tradional packages? (or I am wrong). That doesn’t look good to me.
Re: CaptainFlint. portupgrade helps, but can’t solve the above problems. As I said I haven’t used binaries much though.
Edited 2006-02-15 13:52
For me it comes down to liking source compiled packages better. With faster machines now a days it is not that cumbersome. I did compile gnome 2.8 on a 450 mhz p3 with 128 megs of ram last year… took 2 days in total for everything. Did it again this year on a dual amd64s with 5 gigs of ram and it was < 2 hours.
FreeBSD ports miss information about versions (package abc requires def >=2.5). So, either you recompile all dependencies,..
No, not really – there are some occasions when you have to update related ports, but these are the exceptions, always notified in /usr/ports/UPDATING. So, for instance, there is an update to mysql. You compiled amarok with mysql support – which means, amarok now depends on mysql. When you portupgrade -a – which updates mysql to the newest version – you don’t have to recompile amarok. Also, if you take a look at the Makefiles, most ports depend on the presence of specific libraries, not package version. Everything is handled automatically by portugprade/install tools – you don’t have to skip anything, or have additional recompiles: most of the times (there are exceptions as I noted) when you update, you only update those packs that have newer version available in the ports tree. I might have misunderstood your post however, so please clarify if I missed your point.
Ports is a single version – supporting at least 4 branches of FreeBSD (4.x, 5.x, 6.x, 7-CURRENT) – I think that this is not a problem, in fact, it is a small miracle I believe
Ports are also very stable compared to ebuilds (except stable branch, which is slightly more outdated than ports). Yes, there are always broken ports (but currently, http://www.freshports.org doesn’t differenciate between broken on $arch or $version – but they are working on it, so half of the ports marked as broken will probably work for you, because they might be broken on Fbsd 4.x, or on the Alpha, or any combinations).
FreeBSD ports miss information about versions (package abc requires def >=2.5). So, either you recompile all dependencies,.FreeBSD ports miss information about versions (package abc requires def >=2.5). So, either you recompile all dependencies,,,
No, not really – there are some occasions when you have to update related ports, but these are the exceptions, always notified in /usr/ports/UPDATING.
It’s true I hadn’t picked up the habit of reading this file. Thanks for the tip. If I remember correctly though “portinstall” (not portupgrade) always insisted on upgrading dependencies – so you install “xtinyapp” and suddenly you have to upgrade x.org to its latest version.
I’ll try *BSD again though, this time using binaries and all the shiny tools (portsnap, portaudit). Maybe this time ports will work for me, because I was otherwise very pleased with it.
Glad to be of help
What you describe in installing “xtinyapp” is partly true. There is a ports freeze period that precedes each release, and once a Release is out of the door, pending updates are commited. Also, whether or not you have to upgrade most apps depends on how soon you install after a release is made: for instance, if you install FreeBSD 6.0 right now, probably all the installed packages will be updated when you install a new port. This is simply because the ports tree is tested as a whole. This is a tradeoff – a single ports tree supporting multiple releases, from very old ones like 4.x to 7-CURRENT, and the difference between 4.x and CURRENT is HUGE – most major FreeBSd subsystems has been rewritten. This is analogous to supporting Mandrake 8.0 and a Mandriva cooker from the same repository!
There is a solution howerer: someone wrote it here in this thread, that you can point your PKG_SITE variable to a repository of binary packages that contain the recent software, then use portupgrade -PP -a first (this will pull only updates that have binary packages available – “-P” will try binary first, and if not available, use ports). Then you can install your “xtinyapp” – with fewer upgrades. Even if you intend to use ports exclusively, you have to be aware that the massive upgrade you noticed is a one time job. Once your installed ports/packages are at the version of your current ports tree, you install software as normal. I would chose this latter route – read /usr/ports/UPDATING, than portupgrade -a (and go into a pub while it finishes the job, like someone suggested
. After that, you install ports like in gentoo. man ports, man portupgrade explain a lot of things, they are very well written, and – to me at least – a lot more easier to grasp than portage’s concepts.
You should read look up portinstall and portupgrade. They make upgrading and installing software from the ports easy. Also supported with portupgrade are binary upgrades to software. It is very comprehensive when it comes to software maintenance. Although it has been almost 6 months since I’ve used FreeBSD as my main desktop os.
2. when you run out of pbi’s and try to install from ports, you’ll be the proud owner of a broken system
Have you experienced this? I asked specifically this question on forums before I switched: would it be a problem if I update base system to 6-STABLE, or installed packages? I didn’t get a definitive answer, but I did install PC-BSD anyway. Since than, I upgraded everything – been running linux_base-rh-9 long before the “official” update, recent -STABLE (actually, 6.1-Prerelease now), KDE 3.5.1 before the update arrived, etc. I have skype (fixed the pbi myself), Linux Binary Compat 1.0, K3B and Midnight Commander installed from a pbi (just to keep track whether something breaks) – and everything works perfectly fine! No brokenness here, in any package. PBIs still install fine, even though there is very little of the original PC-BSD system left on my puter (I was very careful though during the mergemaster phase of buildworld/installworld not to overwrite or eliminate PC-BSD specific scripts or entries in /etc/).
and will be for a long time. I’ve been testing it since 0.5 and it sure has come a long way but for desktop usage it is still not as i’d like it to be.
It’s true that it’s base is rock solid, and from that point of view it could be considered mature but it still needs lots of work on gui tools and such.
Not to be considered as a troll, i am one of the most active PC-BSD users.
I’d say PC-BSD 2.0 will rock.
…it says “1.0” is a little rough around the edges, when it is actually 1.0rc2 (which the article mentions in the beginning, this was a quote from the closing paragraph).
Anyway, I agree. Great distro. Makes BSD on the desktop a sweet pleasure. I used it for awhile before giving it to my son… and while you should theoretically use PBIs for adding software, you CAN use ports quite easily. I used it to install a ton of stuff that I needed for work and even upgraded KDE (this was rc1 I was using).
What is great is that this can ONLY get better and more and more PBIs will be added as the community grows… great stuff. And FAST. FreeBSD 6 + KDE just ZIPped along on my old 1.8ghz p4.
I have tried beta BSD and the experiance was fantastic. I’m pretty excited to see PC BSD developing. Just wating for my local magazines to ship it with packages. It has a direct capacity to challange MS in terms of ease of use, and has direct capacity to outpass MS in security. PCBSD’s package manager is pretty easy.
Its good now we have so many choices on desktop. After all choices are the real power of life. Freedom to choose!
I love it… I consider it my fubuntu
Good work there guys.
For those who dislike PC-BSD because of disagreeing with the PBI technology, there is DesktopBSD.
DesktopBSD uses ports through a GUI frontend and has an easy installer as well.
http://www.desktopbsd.net/
http://desktopbsd.net/forums/index.php
Why is the PBI system not an option Kernelpanic? Have you even tested is to see what it does/can do?
Just because you dislike it, doesn’t mean it can’t be a great idea (although, i’d agree, it has a lot to rogress).
I agree with him on the .PBI problem. Their solution to “library hell” is to bring all libraries with you in the .PBI and put them in local-per-program subdirectory for each program. This has 2 consequences:
1. Programs which PRODUCE other programs (compilers/IDE combos etc.) will work but their “product” won’t because it won’t find the required libraries.
2. Memory usage will grow as there will be many instances of the same (or slightly version different) library per each program. This is actualy braking the usefulness of .so altogether.
This is not just some random rant, I am writing this from PCBSD and I also contributed one .PBI myself (Lazarus)
I think however that the .PBI concept is nice for “end-user” apps which don’t “provide” (typical is openoffice or firefox) but not for programs which “produce” something which needs a framework. I think that putting all possible(or most atleast) frameworks through ports and then using end apps via .PBI might be the best solution with possibility of combined checking or atleast one way (so a PBI can see if it has the libs and if not, will install a pkg which it has with itself)
Actually, I think we can have best of both worlds here. What needs to be done is to clarify what belongs to the “base system” (in good FreeBSD tradition) – and make PBI creator skip those libraries that we can expect to be present on all PC-BSD installs.
Ring 0 would be FreeBSD base system + xorg and all their dependencies. Kdelibs, qt, and kdebase might go there as well, with all its dependencies. If you think in an integrated solution (KDE part as the OS for instance) – users will have at least that on each PC-BSD install. And if you come to think of it: FreeBSD + xorg (server, libs, fonts, etc.) + kdebase + qt + kdelibs = that pretty much covers the majority (wild speculation: 70-90%) of all shared libraries you’ll ever need for your PBIs. Which means the “overhead” (in memory for instance) is either minimal or non existent (pretty much all “independent” kde apps – scribus, kdissert, tellico, k3b – these are just the ones I use – will be satisfied without any additional libs in the PBI
Their solution to “library hell” is to bring all libraries with you in the .PBI and put them in local-per-program subdirectory for each program. This has 2 consequences:
Doesn’t Mac OS X work that way?
You can have the same desktop with FreeBSD+KDE
Not exactly. PC-BSD is full of enhancements, utilities, custom configurations, etc… On FreeBSD you may need a few days to customize it to your liking, and still it won’t have all the PC-BSD ease-of-use. FreeBSD is very rough around the hedges. PC-BSD is still, but a lot less
I don’t understand why it takes so many months to finish 1.0, considering it’s based on a current release of FreeBSD.
It’s taking time because we’re improving many features, and we’re adding a lot of additional configurations, for instance, now fonts look just like on Windows with custom rules in fontconfig. Also we should have more than 50 different languages available for 1.0 release, we are fixing many minor annoyances, bugs, we’re testing, etc… It takes a lot of time…
PC-BSD is based on FreeBSD but there is still a big difference between both of them. Making a server OS an easy-to-use desktop OS takes hell of a time.
Also, we’re a very small team of devs ATM.
If you want to join us and help out, you’re mostly welcome
Charles, Please try to find a way to turn of the anti-aliasing in FF and TB so when we make the fonts look like Windows they also work on FF. I don’t know why it doesn’t work anymore but I have done it before on a Linux ditsro. The only differnce was, I had to turn the byte-interpreter off manually and re-compile FreeType. In BSD it’s already off as you mentioned once. I had fonts in FF look same as in Windows along with KDE and GNOME. I am not sure why it didn’t work in BSD.
Edited 2006-02-15 12:12
Yes, we’re working on it: http://www.pcbsd.org/forums/viewtopic.php?t=2737
If you ever find a solution, please tell us… This has been a real challenge lately… But early or late we’ll find a solution!
I edited my comment above, please re-read it
I think I found out exactly what I did in Linux. I will post another post soon. Please stay tuned 
I sent you a private message on the PC-BSD forum. Please check it out.
Actually, the Bytecode Interpreter is off in the former linux_base-7 that we used. Now we use linux_base-rh9 and it is enabled, but the problem is that all fonts are antialiased now. The bytecode interpreter needs to be turned on so that the antialiasing rules of fontconfig work properly though. This is the way it works on FreeBSD at least.
I just wanted to add: take you time, there is really no point in rushing this. PC-BSD 1.0-Final, when released, should hit with a BANG, so in my opinion, it is better to try to work out even minor issues than releasing something that is rough around the edges. We have plenty of that kind of releases in linux-land, especially among desktop oriented distributions (Mandrake/driva, Kubuntu Breezy – the last one I tried – PC-BSD is already as good as these, but that is not good enough imho… we should do better
)
The rock solid stability of FreeBSD with the pretty gui of KDE. I tried out PCBSD and even though KDE isn’t my favorite DE, I liked it and will keep it around on a spare partition to play with when I am bored.
It was great from a Slackware/Gentoo perspective, but ports were weak for upgrading the system. Too much recompiling (even more than Gentoo).
IMO it’s such a non-argument. Why the f*ck do you wait for a compile. Just type “portupgrade -a” and go to the pub. I can’t understand why people are always in a hurry..
first of all – congrats for trying to establish freebsd on the desktop! i hope this will raise the visibility for freebsd in general – it’s about time that it will get more commercial support…
however, what’s still missing to establish really free unix or linux desktop distris (pc-bsd as well as ubuntu are currently totally lacking in this respect!) in the corporate market is network-wide desktop management software like xandros’s xmds or redhat’s satellite server!
of course, you don’t have any problems in this respect with your single desktop at home or 10-20 to manage, but if you have to manage 100 or even 1000 of desktops corporate-wide, you need to have some sort of gui-based centralized management software to deploy, maintain and customize your desktops.
if you could provide something like this based on .pbi, i expect large-scale and commerical interest and support for pc-bsd to really take off which in turn would be beneficial for the whole project!
you wouldn’t even have to give away this software for free, for example you could spin-off a commercial entity which offers add-ons to pc-bsd, but for a moderate fee.
current solutions are roughly as expensive as win xp!
—-
To the other reader that said put KDE on FreeBSD and you get the same thing. Yea but the thing with PC-BSD is KDE is already put up for you and the installation of new programs is piece of cake so I find your argument invalid. A normal user won’t be able to put KDE on a FreeBSD. PC-BSD is a “desktop” OS so I believe you are missing the point with that comment. That’s why it’s called a desktop OS so that you don’t have to put things up yourself. Another thing, even if you put up KDE yourself, it won’t be the same. In speed? Yes, it will be, but you won’t have PBI (PBI is a must for a “desktop”), the menu structure won’t be the same etc.
—-
Well if you are incapable of installing software on an OS, that really leaves you two options.
1. Read the handbook and learn how to use it properly.
2. Stick with Linux
Edited 2006-02-15 19:03
Well if you are incapable of installing software on an OS, that really leaves you two options
Incapable? I don’t think so. Who is talking about me? The point is “end users”. You are missing the point again. PC-BSD is a “desktop” OS, FreeBSD isn’t.
Edited 2006-02-15 20:05
there a common misconception that pkg_add -r can’t get “newer packages” but the fact of the matter is that you can upgrade rather easily with binary packages. You need only set the environment var PKG_SITE to a place with updated packages, something along the lines of
ftp://ftp.freebsd.org/pub/FreeBSD/ports/i386/packages-6-stable/Lat…
works perfectly well.
Combine that with a nice dependency updating/checking script like portupgrade and you’ve got yourself a binary upgrade path. I have yet to compile a new copy of big suites of software like kde for years now.
That next year will be, “The year of the BSD desktop?”
Learn to use portsnap instead of cvsup to update the ports tree, and you will be a much happier (PC|Free)BSD user. It fetches a compressed snapshot of the ports tree, which is *much* faster than cvsup. (and the commands are easier to remember)
portsnap fetch
portsnap update
Combine that with portupgrade -P or -PP, and you will rarely have to compile an update.