PC-BSD 0.7.8 has been released and I also recently conducted an interview with PC-BSD Project leader Kris Moore.
1) What made you decide to start PC-BSD?
I’ve been an off / on user of various linux distros over the past few years. I would always end up getting frusterated at things such as package installation, and bloat. Eventually something clicked, and I decided to try my hand at it, and PC-BSD was born. FreeBSD was always a “safe” place for me, things just seemed to make sense, and I spent more time actually using the OS, then actually trying to make things work.
2) Why use FreeBSD as its base opposed to NetBSD, OpenBSD, etc?
Stick with what you know. To be honest, I’ve only tried the other BSD’s a few times, and FreeBSD has always been my bread and butter. Their documentation and availability of information on the web has always been a huge help when troubleshooting problems. I’ve just always felt more comfortable with it.
3) What benefits over other package systems does PBI offer?
Probably the most important one of all… Usability! Don’t get me wrong, from a technical standpoint the ports system, or RPM may be very advanced, but from an “ease of use” standpoint, it really is a step back for your average user. People migrating over from Windows or Mac just want to download a file, and double click it to install. No and’s, if’s, or’s, or but’s. This is something that I think has greatly hindered other open source products on the market. Being forced to A) download programs in a project’s app list, or B) being forced to compile from source, are really not options for 98% of normal computer users out there. I couldn’t imagine having to download apps for my OS from microsoft.com, but this is normal for many Open Source OS’s out there. My goal is to instead keep the OS and the packages as seperate and distinct as possible. Which is in my opinion, the way it should be.
4) Other projects like Desktop BSD and NetBSD-Office have appeared since PC-BSD’s arrival on the scene, what do you think of their offerings and ideaology?
To tell you the truth, I haven’t had a chance to take a look at any of the other projects yet. Just trying to keep up with PC-BSD has been crazy enough.
5) What do you think it is about PC-BSD that makes it unique when compared to these other “Desktop” BSD releases?
Well, as I said before, I’m hoping that my take on software installation will be the key.
6) As we know PC-BSD is a KDE orientated BSD Distro, are there any plans for a GNOME based version? Such as in Ubuntu and Kubuntu?
Not at the moment, but it could happen down the road a bit. For now I’m just trying to focus on getting to 1.0, and have the OS be as full-featured as possible. Trying to support multiple WM’s would slow me down a bit in this reguard. However, I will be open to the idea once things are more finalized here.
7) What are your views on the newly started Kamel project to provide PC-BSD with a GUI based on KDE?
I’m excited to see people taking ideas like this and trying to put them into action. Hopefully it’ll come to fruition and I’ll probably be among the first to try it out.
8) What features are going to be present in the next release of PC-BSD?
Well, 0.7.8 was just released, but I’m already thinking ahead. A couple of the biggies I want to get implemented is the Online Update Manager, so that people won’t need to keep reloading the OS to get the latest features. Along with that, I want to get the installer setup so that you can “upgrade” your existing OS, I.E. install new version of PC-BSD, and not format the HD, keeping your programs / settings on the box.
9) How will these features benefit the end user?
It will help people commit to the OS, by knowing that they don’t need to reload in a month when I get something else, something newer finished.
10) When FreeBSD 6.0 is released will PC-BSD’s base system be updated accordingly?
Depends. I haven’t had a chance to play with 6.0 yet, and I’ll probably wait a bit just to ensure that its stable and solid before looking at updating. Once 6.0 gets there, I will probably consider migrating.
11) Can you see PC-BSD breaking away from its FreeBSD roots to form a new BSD kind of like what Apple did with Darwin, or are you dedicated to maintaining compatibility with FreeBSD?
I want to try and stay as 100% compatible with FreeBSD as possible. I don’t want PC-BSD to become a fork of FreeBSD at all. My goal is to provide the tools / utilities for the graphical desktop, so that your casual computer user can sit down and be productive on the machine, with no command line interaction at all. But for the power-users out there, I want to keep FreeBSD as pure as possible, so that they can modify / tweak things to their heart’s content. It’s an ambitious goal, but one worth having.
I have tried PC-BSD, I like it a lot. I stuck with Linux, just becuz it is what I know. But when I have some free time I plan to give PC-BSD a real work out. I did try the package installer that he is talking about, WOW is all I can say. It is as easy as a Mac to install software, good job!
I also liked that I could still use the FreeBSD ports system, the real neat feature of the FreeBSD platform. If CrossOver Office, and Cedgea ran on FreeBSD, I could really see my self making a total switch to PC-BSD.
<cite>If CrossOver Office, and Cedgea ran on FreeBSD</cite>
You can use the Serenity Virtual Station:
http://www.serenityvirtual.com/
It is a competitor to VMWare, but it supports Windows, Linux, FreeBSD, and IBM OS/2 as hosting platforms.
Also, the TextMaker Office suit got rave reviews a while ago, they said it was the most compatible to MSOffice product. It runs on FreeBSD also:
http://www.softmaker.de/tm_en.htm
As I write this, I come to realize there’s no excuse really not to use FreeBSD. BSDs rocks. Linux sux. (Let me get my asbestos suit on.)
When I googled for it, I saw mention by one of the developers from 2003 that they hoped to have a BSD-version of Crossover Office later that year. Anyway..it seems you can try the Linux compatibility layer in the BSDs if you’re slightly masochistically inclined.
“I want to try and stay as 100% compatible with FreeBSD as possible. I don’t want PC-BSD to become a fork of FreeBSD at all. My goal is to provide the tools / utilities for the graphical desktop, so that your casual computer user can sit down and be productive on the machine, with no command line interaction at all. But for the power-users out there, I want to keep FreeBSD as pure as possible, so that they can modify / tweak things to their hearts content. Its an ambitious goal, but one worth having.”
I love this guy! Keep up the great work!
This guy is a genius! I tried PC-BSD and it’s excellent! I’m a long-term FreeBSD user and with PC-BSD (still beta version) I can use all FreeBSD features such as the ports, etc…
It took 15 minutes to install PC-BSD, very easy and all my hardware was detected and properly configured (incredible!). It was very easy. Installing software is also very easy. Keep up the good work!
I guess it’s going to get much momentum pretty quickly.
Good interview! This guy really knows what he wants with his project. I’m pretty curious about his package system too. I really need to download PC-BSD because I got this old compaq p400 thats just waiting to get a new OS installed on it.
I just wish Linux distros could take a hint from this guy. LOSE the FLUFF.
” An operating system should be just that, an operating system, upon which users can load and unload the software of their choice.”
-AMEN
Linux distros seem to be scared to say this is ‘OUR’ default app. Anything different must be installed by the user from the repository.
I also love the way this guy has solved the dependency hell problem. They set it up so that all applications and required dependencies are stored in its own folder. I can see where this might add some bloat to the over all system, but you know what? I’m sitting on 250 GB of storage of which I have used 35.6gb, and if a little bloat can save me from linux dependency hell, I’d sacrifice the extra space.
I’ll keep my eye on this one.
> I also love the way this guy has solved the dependency
> hell problem. They set it up so that all applications
> and required dependencies are stored in its own folder.
Yes, that’s very clever. Not only clever, but awesome and gorgeous: If there’s a bug in a library, not only one package needs to be fixed, but dozens.
It’s the best feature that one could copy from Microsoft Windows[TM]. Apart from that, there are enough UNIX operating systems for x86 computers.
>Yes, that’s very clever. Not only clever, but awesome >and gorgeous: If there’s a bug in a library, not only >one package needs to be fixed, but dozens.
>It’s the best feature that one could copy from Microsoft >Windows[TM]. Apart from that, there are enough UNIX >operating systems for x86 computers.
ahhh no. You’re right about each package but as far as there being enough UNIX OS’s for x86…. hmmm.. who makes that determination… you?
Sounds like sour grapes to me. At least this guy is trying to do something original and attempt to solve problems.
This really is a double standard. Suppose Windows used a system similar to ports. Could you imagine the outcry if microsoft was able to upgrade the libs of all your various software on your box? Talk about crazy, this would drive me nuts! Everybody would hate microsoft’s guts for it
But with *nix its ok to do it?
They (Microsoft) does do just that.
They invented the dependency hell.
Again, it sounds like there’s more of an axe to grind here than you’re letting on perhaps?
If zlib has a security flaw, upgrade zlib and then have the package application recompile the apps that rely on zlib. It’s not hard.
Well, Microsoft does the very same, with their security updates.
debian:
aptitude checks available packages in a central repository, checks which need to be installed on the system, downloads and installs them. you can see the list of installed packages, when you run aptitude.
ms:
windowsupdate checks available patches in a central repository, checks which need to be installed on the system, download and installs them. you can see the list of installed patches in control panel / something.
if you are running a stable version of debian, you only receive security updates.
the only difference is that debian is your vendor for your os, office applications, games, browsers, sound editors, movie players, etc – almost all of the software installed. ms is your vendor for only a part of the software installed.
with a bit of abstraction, you can see, that ms is already using a package-manager + central repository to distribute a part of their software.
Seems like a dead point, since:
1. This is a beta product
2. The guy is working on an software update program.
So… If it is a HUGE issue for you, don’t use the product yet, and wait… Or, if you can work with the current limits, enjoy your self.
it is not a HUGE issue for me, as i do not use the product and probably never will. i was just presenting my point of view – that i consider his solution worse than what we have now.
If there’s a bug in a library, not only one package needs to be fixed, but dozens.
Dozens is probably an overstatement, but it is a valid observation. However, look at it this way – I’d rather replace ten copies of a library (if you need to, most of the programs will work fine with the original lib) than to replace one copy of a lib and have it break ten programs.
Sometimes, updating a lib means all programs using the lib must be updated as well. By having a seperate copy for each program, you don’t need to update programs which only changed because a lib changed.
Dozens is probably an overstatement, but it is a valid observation. However, look at it this way – I’d rather replace ten copies of a library (if you need to, most of the programs will work fine with the original lib) than to replace one copy of a lib and have it break ten programs.
Dozens is actually an understatement. And this isn’t about bug fixes. It’s about security.
“apt-cache rdepends zlib1g” returns 1,848 packages that use zlib. Imagine if a user just had 10% of these packages installed. That’s 184 pieces of software or libraries that would have to be ugpraded. And 184 of those must be upgraded to plug the security hole.
Shared libraries solve this problem by centrally updating a single library. The linked programs never even have to know about it.
Dozens is actually an understatement. And this isn’t about bug fixes. It’s about security.
“apt-cache rdepends zlib1g” returns 1,848 packages that use zlib. Imagine if a user just had 10% of these packages installed. That’s 184 pieces of software or libraries that would have to be ugpraded. And 184 of those must be upgraded to plug the security hole.
First, security doesn’t affect most programs. Who cares if zlib has a hole if you’re only using it with a SNES emulator? That’s one less copy of zlib.
Second, you’re still over-estimating it. I doubt the average person would have even 1% of those programs installed, so that’s less than 18 packages, of which, most won’t be a security issue as they don’t connect to the net.
Third, if I DID have that many copies to replace, I’d write a bash script which found and replaced all the copies. It would take about three lines and two minutes. I would be surprised if you checked the PC-BSD forum and there wasn’t already such a script available.
You’re making a mountain out of a molehill to justify your position. It just doesn’t hold any water.
Surely you are joking. Using “I’d write a bash script” to defend a feature, which tries to be a usability feature.
On one hand, you claim it is a good feature for the regular user, for whom it is too hard to do (click synaptic, type string, select, click install), on the other you expect him to go to forums, copy/download a bash script and run it?
Get your arguments and expectations of a “regular user” straight.
<cite>Third, if I DID have that many copies to replace, I’d write a bash script which found and replaced all the copies.</cite>
Right. As I said, this is an easily solvable problem, but I propose metadata to keep track. This can installed atop of what they have.
Sure, there’s bloat. But that’s the only work around to avoid cross-dependencies. We know, from experience, that the traditional approaches get you:
1) Either source compile.
2) Or, use the smart BSD approach: ports.
3) Package-management hell. Worst-case scenario: Debian. Package explosion (1000->8000->14000) breaks down release management, because they don’t have /any/ automation.
With 2 and 3, either way, you’re looking at forcing a 6 months overall release-cycle: Ubuntu, OpenBSD, just about everybody who’s serious (except Debian 🙂 Sorry).
Can’t escape from library dependencies in 1, either.
It’s an undeniable fact that software installation under WIndows is easier. In fact, that concept started with NeXT, not Windows, so be cool everybody!
Besides, people are overlooking the fact that a package is installed under a user directory with permissions. This isn’t Windows, so it’s safer still. The argument holds no water, really.
It’s an undeniable fact that software installation under WIndows is easier. In fact, that concept started with NeXT, not Windows, so be cool everybody!
any other arguments than emotional? i already mentioned the problem with upgrading packages. it has an O(n) time cost, compared to O(1) for a system with package management.
it is similar with installing the first time. compare:
* download firefoxe bundle
* drag&drop it somewhere
* download thunderbird bundle
* drag&drop it somewhere
* download gaim bundle
* drag&drop it somewhere
* download gvim bundle
* drag&drop it somewhere
* download gimp bundle
* drag&drop it somewhere
* download openoffice.org
…
to:
* start package manager
* select gimp, gvim, gaim, thunderbird, firefox, OO.o
* click install
yes. the second option requires a central repository and the user learning a different way to do things.
however, i would not equate “don’t know something” to “something is more difficult”.
I also love the way this guy has solved the dependency hell problem. They set it up so that all applications and required dependencies are stored in its own folder. I can see where this might add some bloat to the over all system, but you know what? I’m sitting on 250 GB of storage of which I have used 35.6gb, and if a little bloat can save me from linux dependency hell, I’d sacrifice the extra space
it would be ok, if it was only your disk space. it is, however, also the space on the server, which provides you all these packages and all the mirrors.
bigger packages also require more bandwidth to download. and bandwidth is NOT cheap everywhere. and why the heck should i download the same library ten or twenty times. just imagine all gnome apps. i want the gnome environment, but i don’t want the totem player, or the gedit editor. either, everything is in a big fat “gnome” package (so much for “little” bloat), or it is nicely divided as in e.g. debian, but i have to download libgtk+2.0 for every gnome application. makes zero sense.
also, there is the developer time – instead of building libfreetype2 once, they have to make sure it builds correctly for each appplication that uses it.
all this for a NON-problem. sorry, but no matter how you try to twist it, i have not experienced dependency-hell since i moved to debian. apt-get (aptitude) solves the dependencies for me. you know, the computer is the tool and it should work for me (figure out the dependencies and download only the minimum, use only the minimum space), not the opposite way (me buying more disk, paying more bandwidth) to have the same library installed 20 times.
someone has already brought up the security problem – e.g. zlib. widely popular. many apps linked statically to it, which is almost the same (from the distribution point of view) as having it bundled with each app and linking dynamically. once a bug was found, you had to re-install all the packages. i’d rather have one non-functional program on my system (due to a new version of a library) than one copy of a library with a security problem.
having the same library installed more times could also break the benefit of dynamic linking – only one copy of a library in memory. either, all the library directories from all installed self-contained applications are in LD_LIBRARY_PATH. in that case, the first copy of a library will get loaded for all applications that use it, hence no point having a separate copy for each applications. or, when a self-contained application is started, only its own library directory is in the LD_LIBRARY path. in that case (and i’m not 100% sure on this), the linker would consider /app/foo/lib/libxml2.so to be a different library than /app/bar/lib/libxml2.so and load both of them – instant waste of RAM. i’d rather use my RAM on file cache than 20 copies of the same library.
talking about upgrades/reinstallation. try to think about, how fast you can update a system, where a package managers keeps track of the installed software and how fast a system, where you have to download and double click for each application installed.
i use both debian (@home) and w2k (@work). there is a set of applications i use on both systems – firefox, thunderbird, gimp, gaim, gvim, openoffice. to update them on debian, i do “aptitude update ; aptitude upgrade”. 2 commands for 6 applications. actually, it is 2 commands for all the applications installed. on w2k, i need to download 6 files, unzip one of them (gvim), run 6 executables. with 10 applications, that would be downloading 10 files (from different places !) , running 10 executables. with 100… i hope you get the idea. how can someone consider the latter an easier way escapes me… (instead of typing aptitude blah blah, one could click-start synaptic a click a few buttons to do the same). yes, people are used to it. people were used to driving without security belts, licking their pencils. not everything that people are used to do is the correct way to do things.
if you really need to have installation from within a browser (which i also don’t like – you know, the right tool for each job…), it could be solved by mime types and respective handlers.
<cite>all this for a NON-problem. sorry, but no matter how you try to twist it, i have not experienced dependency-hell since i moved to debian.</cite>
What a load of bull! Anyone who’s used Debian for a couple of years know how problematic it has been for Debian stable to get stuff rolling.
Debian represents the utmost failure of a free software project. It’s become unmanageable for all those people to create packages in a timely manner.
Look how long sid took! YEARS! Ubuntu has taken its place.
C’mon! You think we’re all newbies here?!
You think we’re all newbies here?!.
No. I think you are someone without a sound argument hiding behind exclamation marks, rude words and not signing your words.
if you read my entry, instead of just finding a bit to flame, you would have noticed i was discussing the technicalities of installing packages, not their quality or up-to-date-ness.
call me, when you get ubuntu running on something else than an intel machine. still, i consider ubuntu a good thing.
It would be ok, if it was only your disk space. it is, however, also the space on the server, which provides you all these packages and all the mirrors.
Where have you been? PC-BSD is for desktop, but for the server!
jziegler, I think you miss the point completely.
If you’re comfortable with an advanced package manager, it means the PC-BSD simplified package management is not for you: so, even on PC-BSD, you would use the excellent FreeBSD ports, since they’re totally available there too (it has a complete FreeBSD operating system under the hood).
Your mistake is thinking that everybody has the same priorities as you.
If somebody has a big HD (and today, most HDs are) and they happen to want to get the work done *the fast way* – without even knowing or caring about what a dependency *is* – then the PC-BSD package manager is the best solution for them. It’s as simple as that.
Many users of a desktop OS are exactly like that. And it doesn’t mean that they’re stupid or something – they simply *don’t care*. Maybe they’re not interested in a PC’s inner workings, or maybe they’ve got more urgent things to do.
ulib, have you even read what i’ve written? the HD usage is only one part of the problem. downloading the same things 20 times also cannot be faster than downloading it only once.
i was not discussing user preferences or ease of use. i was discussing the technical implications of such a solution. it’s a bit sad that no-one seriously answered me on that points.
but yeah, i’m strange. i try to understand things around me.
One thing, I read, which by my understanding means, it only maintains a copy of a library if the package calls for a version that isn’t already on your system.
Kind of shoots down the whole 200 versions of the same library arguement you got there. It only grabs another library version and places it in that app folder so you don’t have to recompiling all your programs for an updated or outdated lib.
OK. That sounds reasonable.
First, i think his new Package manager is excellent — it’s exactly what a windows user would like: simple and straightforward, all files needed in a single download.
As for the Kamel side-project, this is the first i’ve heard of it, and looks very interesting. I’m glad he’s focused on the end-user experience.
the desktop choice for pc-bsd is kde :>
Yes, I’m glad KDE is the default environment. I don’t like other desktop environments that much…
One thing that would also be helpful would be a GUI tool to help the newbie use the ports system without sending them to read the “handbook.”
I’ve tried out the last 2 releases of PC-BSD, and quite honestly I didn’t really like the package system because you couldn’t keep the software up to date with it…which is why a ports gui would be better, IMHO. But it sounds like he’s working on it, so I may just wait till that functionality is release before I install another incremental release.
Other than that, I’m really excited about this project!
I tried PC-bSD 0.75 and it rocks. But KDE 3.4 feels a bit sluggish. But it looks nice.
I really like PCBSD . It’s very very good . Detects all the hardware I have . Package system is very easy to use , even easier than Windows . My only complaint is about the speed of Internet – it’s too slow
.
As someone said above “I love this guy! Keep up the great work!”
I really like PCBSD . It’s very very good . Detects all the hardware I have . Package system is very easy to use , even easier than Windows . My only complaint is about the speed of Internet – it’s too slow
.
As someone said above “I love this guy! Keep up the great work!”
If I understand with PC-BSD if you have X apps that use zlib you have to update each app to get a security fix for zlib?
Any Apps you have installed with ports, can be updated using the ports system. Appliactions you download from PC-BSD’s website and install, will have to be re-downloaded and installed, or you can just install the next release… Since he seems to be releasing about once a month, thats about as fequent of an update I can take!
It can be very boring to update every application and I guess that libraries are statically linked just like I guess it happens on MacOS X.
IMHO system libraries are part of the operating system, application specific libraries *may* be included in a self-contained environment for the app itself.
With dpkg is possible to create a kind of package – a task, so you can create a task named xyz, the user install xyz and apt downloads and installs the whole dependencies chain, so the user experience is just the same as download a .pbi that installs a self-contained app.
Can any one tell me about an install of PC-BSD they have done on a laptop? Does it support suspend and standby? What about 802.11b/g?
I would like to know if KOffice is included. I would think that it is, but if this is true then it seems to violate his mantra of having “just” an operating system.
No Koffice on the system, or Kdevelop either. Just the essentials
You can check for wireless support in FreeBSD’s project page or PC-BSD’s forum. People answer fast there.
Security udates aren’t in place yet but as you can imagine, it is a question which will logically be adressed. Right now, people are focused on making more mundane things work.
I am enjoying this OS very much. Kris actually hears people and helps them. No elitism on that forum whatsoever!!!
You are welcome with your sugestions!
no KOffice is not included only text utils that are included are KATE and Kedit
Thanks!
I also love the way this guy has solved the dependency hell problem. They set it up so that all applications and required dependencies are stored in its own folder.
It will be insteresting to see how they can deal with the GNOME dependencies. I doubt it will be an easy task. Check here url:
http://people.freebsd.org/~adamw/gnome_kde_deps/
Keep in mind, it’s only gnome2-lite. I think, if they keep storage the dependencies in its own folder will have more problems in future when the things are growing bigger and bigger.
KDE is dying with its bad GPL license and too many germans controlling it. Nobody can trust KDE.
This can be easily solved with metadata, or a sort of “registry” that keeps track of where the libraries are. Any security problem and a program would just follow a tree, updating everything.
Sure, this creats the problem of multiple copies and a larger update. But that’s the price you pay to keep installations as free as possible from dependency-hell. Dependency-hell kills. Look what it did to Debian.
UNIX-to-the-core fans will scream. But then again, UNIX-hardcore people: 1) are mostyl network guys; 2) only code in C.
Some people have other type of work to do besides setting up a firewall, and distros, particularly Linux, SUCK in the software installation thing.
Kudos for PC-BSD! UNIX FOR THE PEOPLE!
I agree with your solution however I do take exception to the generalization that all hard core Unix people code in C.
Ada2005 is my favourite language. I’ve been using it since it was Ada83. 🙂
Cheers,
Nick
I agree with your solution however I do take exception to the generalization that all hard core Unix people code in C. Ada2005 is my favourite language. I’ve been using it since it was Ada83.
Well, Ada is nice! in fact, I like the fact that you can find Forth code in FreeBSD’s bootloader and Modula-3 in CVSup. These were chosen for a very good reason.
Maybe in FreeBSD it’s different. I’ve used more OpenBSD, where’s there’s a definite trend to keep it simple and small (C). For better or for worse, and I can’t really see much non-C get in the core system. Maybe FreeBSD people are more open, it sure looks that way. In fact, this C-fixation is an aspect I don’t like in OpenBSD, but I understand the rationale.
I tried PC-BSD before OSNews broke the news and guess what it works on R3000Z. I have been modded a couple of times and screamed at and also shown links to FreeBSD books to go fix it myself. Finally I can use FreeBSD on my Lappy.
Comments above prove yet another time – security and usabilty are mutually exclusive.
Well, before you start flaming, take a break. Maybe these things are not mutually exclusive, after all.
Assume that usability of encapsulated apps is near perfect (simple install/uninstall, no dependency problems). What about security?
OS base system security doesn’t depend on apps/packages at all.
Having encapsulated apps on system, application security model is simpler than in usual case – you can focus on securing single applications (by upgrading these or their components/libraries). You don’t need to worry about breaking system or dependencies or other apps – they are isolated anyway. IMHO this is positive both for security and productivity. Of course app maintainer needs to watch for libraries problems – but it’s needed anyway.
Only big problem seems appear with widely used libs like zlib – if someting breaks in such libs, almost all apps need to be updated. Fortunately such libs tend to have very stable interface and mature code (otherwise they wouldn’t be used so extensively), thereby it isn’t so hard create simple upgrades for all encapsulated apps (or entire package system).
Of course such “mass upgradeabilty” has to be designed properly and its possibility needs to be included into base package system. As far as I can understand, PC-BSD makers will include something similar into their package system anyway; they just have not finished it yet:)
Comparing to ports system, such mass upgrade of packaged apps is not any worse – all apps are upgraded, no apps will break, time to upgrade is about same.
BSD is not copylefted, therefore freedom zero is not guaranteed. Don’t use it!
http://www.gnu.org/copyleft/copyleft.html
I haven’t used PCBSD yet, so I’m just wondering how it deals with printing? Trying to set up the default lpr system FreeBSD, OpenBSD and NetBSD is pretty traumatic. The excellent book “Absolute OpenBSD” pretty much says that setting up printing in BSD is so difficult that you might as well forget it. On the other hand, FreeBSD does have a CUPS port, though it’s not real elegant. If PCBSD had CUPS already set up, that would be grand.
Power Management & Security updates
You mean you like the GUI stuff. For the OS thank FreeBSD team. 😉
all this for a NON-problem. sorry, but no matter how you try to twist it, i have not experienced dependency-hell since i moved to debian. apt-get (aptitude)…
but Debians Pacakge system is dpkg not apt-get, apt-get is just a repository tool, so for Debian you even need an additional tool to make sure you dont have dependancy problems.
i know. so you use two tools in tandem. big deal. one is “frontend” – apt-get, one is “backend” – dpkg. the point is you tell the computer which package you want and the computer makes sure it gets installed and is functional (has al the dependencies present).
Firs off all, for average users (not por advanced or power users) the pc-bsd aproach is the best, without any doubts.
From this point, pc-bsd or better said: people who package their software to be used with pc-bsd (as pc-bsd should be only the base system) have to care about security, the same as in windows world.
A possible way could be that applications in pbi has is our dynamic libs, and yes we will have a lot of zlib dynamic libs in our HD. If is there a security problem with this lib you can download ONLY the security patch for this app that will replace her lib, not the whole app.
Only a suggestion.
If is there a security problem with this lib you can download ONLY the security patch for this app that will replace her lib, not the whole app.
Fair suggestion. Still, you have to download it for each app you have installed. And you would have two types of packages – complete app, security update. If you can explain this to the users, why not “click these two buttons to receive all security updates”. And if you can explain that to the users — the same can be done today in e.g. synaptic.
the pc-bsd aproach is the best, without any doubts.
Care to explain? I don’t want to flame on this? I’m honestly curious, why people are claiming this. I have not heard any arguments except of “people are used to this from windows”. Which is an argument, but I’d be more interested in more technical ones.
Well, I’ll try to explain my view.
Using pbi system is not really a better way for finnal users only, it’s for users and above all for distributions developers.
As you said the finnal user don’t care about click on synaptic or click on other package manager tool. The problem today is, as we can see on debian (my work system) and other linux main distributions, the big ammount of work to be done to maintain the whole system with thousands of packages.
If you provide a system, as I think is trying pcbsd, where third parts can provide their software without care about dependencies, libraries, etc… developers can have all their efforts in main system.
MS don’t care about provide a flash package for windows is Macromedia who care about provide a flash package…
I am a linux person and i liek linux it works for me just great. At the same time i have read a few articles and spoke to people who recommend one form of BSD or another. This is an excellent stepping stone for me personally as i can find my way around most distro with no head ache at all so i think its time i give BSD a go.
I like the install alot !!!!
ps will there be a gnome version.
“will there a gnome version.”
have you read the interview?