useless system

Status
Not open for further replies.
Unlike Windows and Linux, FreeBSD is a professional operating system for professionals and serious amateurs.

Yes, very much that way. Windows and Linux have a (more or less) clear distinction between consumer versions and business (aka "enterprise") versions - and to handle the latter, you are recommended to do courses to become some whatever-certified-whatever. In FreeBSD there is only the professional/business version, and you are as well expected to aquire some skills to handle it successfully.

And there is something else: when I finally achieve to make feature X run on system Y, then it is rather straightforward afterwards to write a tutorial on how that was achieved. But for FreeBSD this does mostly not apply, because most features were originally developed in Berkeley. There is e.g. no point in describing how to make Internet work on Berkeley-Unix, because Berkeley-Unix was what made the Internet work in the first place.

I might suppose the FreeBSD user base is not so much interested in getting many FreeBSD desktop users who wantz zu runz their desktopz on freebsdz bekause it is so kuuulz to have freebsdz as desktopz. --- Instead we would like to encourage you to learn and study and undestand a bit more about practical computing, so to get more empowered and have more of a clue about the inner workings of all that IT surrounding us everywhere today.
 
What desktop should be chosen as default? Your preferences may not coincide with mine. So maybe I don't like your suggestion of that default and I'm going to suggest a different one. You then end up with an endless debate about which desktop should be set as default.

I take your point, but isn't there already 'prior art' in the FreeBSD installer? Which filesystem do you want to use with FreeBSD? UFS or ZFS? Some want one, some want the other. Rather than insist that you have a specific one decided by the developers, there is an option within the installer. IMHO installing a GUI should be no different. Don't want one, click past it in the installation process. Do want one? Select one from a list. As part of the post-installation reboot the GUI that you specified could be pulled down and installed. That way the GUI packages need not be part of the base system (another excuse people (rightfully) give for not including a GUI in the installation).

I'm not trying to pick a fight, just play devil's advocate, and the more I use FreeBSD, the more I have come around to the current way of thinking (i.e. no GUI option in the installer). However, if we did add an option in the installer, a lot of the forum threads involving new users and GUI's would disappear. And as another user mentioned, making the OS more appealing to first time users may increase uptake and contributions.
 
Why are you so bend on having FreeBSD do it for you? It sounds like you want FreeBSD to be something that FreeBSD itself does not want to be.

Re-read Toolforgers post and you'll understand my issue. If FreeBSD isn't going to ship a desktop (which is fine), make setting one up simpler.

The whole purpose is to make the barrier to entry a lot easier - at first glance. The alternative is to use the installer.

There are a multitude of issues with the GhostBSD approach, so I don't even consider that. But that's another argument i won't get into.
 
I take your point, but isn't there already 'prior art' in the FreeBSD installer? Which filesystem do you want to use with FreeBSD? UFS or ZFS? Some want one, some want the other. Rather than insist that you have a specific one decided by the developers, there is an option within the installer. IMHO installing a GUI should be no different.

Easy answer here: You need the filesystem first in order to install the system onto it, so this has to be answered first.
But you dont need a desktop first in order to install the system onto it.
 
The data is going to be important enough to rule out Linux

Why would linux be unsafe or not trustworthy enough for your data? If you already know linux, why not spend the time to better understand it so you can make a safe setup instead of learning a new OS and then not learning it well enough so you'll end up with an unsafe setup anway?
 
Re-read Toolforgers post and you'll understand my issue.

Sorry, I don't understand the issue. I'm new to FreeBSD as well and it's quite obvious to me that FreeBSD was not designed for the complete newbies who have no knowledge whatsoever about computer systems or networks. If you are, FreeBSD is not for you. Try Mac, some Linux flavour, GhostBSD...
 
I'm new to FreeBSD as well and it's quite obvious to me that FreeBSD was not designed for the complete newbies who have no knowledge whatsoever about computer systems or networks.

Ok, let's take this argument further. Why even have an installer to begin with? The expectation is to have people understand the inner-workings of a system; why not scrap the whole concept of an installer and manually install everything the Arch way? You see where i'm going with this?

People who use this argument when it comes to FreeBSD desktop drive me up the wall. At least put a damn an WiFi SSID locator in the installer at least so i won't have to futz with config files on every FreeBSD installation.
 
You don't like FreeBSD because it has no desktop GUI. Then install GhostBSD as it is FreeBSD with a desktop.
 
You don't like FreeBSD because it has no desktop GUI. Then install GhostBSD as it is FreeBSD with a desktop.

I love it. :) I just wish certain desktop subsystems were automated at least - like WiFi and device probing for drivers. I'm fine with it not having a GUI.

GhostBSD is just more fragmentation IMO. Again, I'm not fond of what they're doing either.
 
IIRC, FreeBSD was once shipped with installer that had the ability to configure a GUI. Due to lack of maintenance by anyone that used it, it became broken by other updates, bugs didn't get fixed, etc. When deciding what to do with the installer feature it was decided that it would be removed rather than fixed because no-one came forward to fix it. (This is me remembering a segment from a past BSDNow episode, so details may not be correct)

If you want it there is precedence for it being there, but you may have to make sure it gets there yourself (be that coding it yourself, lobbing developers for it, sponsoring the FreeBSD Foundation to do it, etc)
 
Why would linux be unsafe or not trustworthy enough for your data? If you already know linux, why not spend the time to better understand it so you can make a safe setup instead of learning a new OS and then not learning it well enough so you'll end up with an unsafe setup anway?

I'll establish safety by minimizing the attack surfaces. I.e. no services that I didn't explicitly enable - with Linux, the distro may add a new shiny service (with associated security holes) with any update, and I don't expect that to happen with FreeBSD.

It's going to be a NAS, which is essentially single-user except for technical users that operate the various jails.
It is going to pull data via SSH (or rsyncd if that turns out to be both easier and safer, but I doubt it). No sshd on the NAS, no remote control of any kind - minimizing the attack surface.
The data is going to be published via a Samba stack that has been hard-configured to disallow write access. If Samba cannot do that reliably enough, I'll configure it to serve from a read-only view of the ZFS-based data filesystem.

So... I'm minimizing not just the attack surface, I'm also minimizing the impact of my relative incompetence :)

Besides, I have the idea that while I'll be hindered more by lack of experience, I'll get better results once I get up to speed. Linux limits have always been pretty soft and incomplete, and I don't really trust the somewhat incoherent mixture of ulimit, cgroups, and systemd that Linux is relying on today. (systemd is far too feature-hungry for its own good, and that's generally what I see in the Linux world: More regard than getting things IN, while FreeBSD has a focus on getting things RIGHT. Which doesn't mean that FreeBSD is free from defects, it's just that I expect less defects in the FreeBSD world, even if it's at the price of less features. I pay that price willingly to get my secure NAS; I not so sure about the effort-to-effect ratio for, say, setting up a developer workplace, but that's not my current demand so I don't worry too much about that.)

(Btw 'nuff said about installers and such. It's a no-no on this forum, and I respect that.)
(Plus forquare is right on spot that complaining isn't helpful, making things happen is.)
 
IIRC, FreeBSD was once shipped with installer that had the ability to configure a GUI. Due to lack of maintenance by anyone that used it, it became broken by other updates, bugs didn't get fixed, etc. When deciding what to do with the installer feature it was decided that it would be removed rather than fixed because no-one came forward to fix it. (This is me remembering a segment from a past BSDNow episode, so details may not be correct)

If you want it there is precedence for it being there, but you may have to make sure it gets there yourself (be that coding it yourself, lobbing developers for it, sponsoring the FreeBSD Foundation to do it, etc)

Back in the FreeBSD 4.x, days I remember this, 5.x might have still had it but 5 was kind of wonky. Then there was what I call the Apple exodus and wonky got wonky. There even used to be a port called instant-workstation which was maintained by Greg "groggy" Lehey.

I think now a quick graphical desktop could be added to the installer, something like OpenBSD does. But people would still cry about it because their sparkly new booble or win printer doesn't work with it. Or that it's some lightweight DE like fvvm instead of some heavy full fledged KDE/Gnome offering.

I think when one comes to FreeBSD you have to realize that the projects goals are not world dominance and being number one (aka expanding the user base and being user friendly).

Just my $0.02
 
Toolforger: I'm sure you will have a decent firewall protecting your NAS from the public Internet? Even if your favorite Linux distro decides to add and enable some random service in a new update (which I highly doubt), your firewall will still block traffic to it.

I'm not saying you should go back to Linux. I myself just recently made the move from Linux to FreeBSD. I just wouldn't call Linux that much more insecure than FreeBSD.
 
If this is so important to people, I would think it wouldn't be all that difficult to put together a package--mostly a script--that could piece together some kind of desktop and, perhaps, present it as a port to the committers. I don't mean it will be easy--it could take weeks or months--but a basic package might not be that difficult, just time consuming.

Be prepared for all the complaints about what you did or did not include.
 
If this is so important to people, I would think it wouldn't be all that difficult to put together a package--mostly a script--that could piece together some kind of desktop and, perhaps, present it as a port to the committers. I don't mean it will be easy--it could take weeks or months--but a basic package might not be that difficult, just time consuming.

Be prepared for all the complaints about what you did or did not include.

It's certainly worth looking into. Good enough for me to take a jab at writing myself.
 
... which is a shame because I think if more people were introduced to it it would catch up to linuxs' market share quickly.
Why would I want FreeBSD to have Linux' market share? That doesn't help me at all. On the contrary, it is likely to hurt. Today I use FreeBSD because it is a high-quality OS, put together with care, for reliability, logical operation, ease of long-term administration, and so on. These attributes are not what gets an OS really high market share, which is ease of installation and administration, and a flashy GUI. If the (small and fixed) set of developer resources get moved towards a consumer and beginner OS, then there is likely to be less effort in making it a reliable and well-built server OS.

I'd really like if the FreeBSD installer autodetected network cards and did the little zeroconf dances that all the other operating systems do for you.
And I would disable any such dance, or switch to an OS that doesn't do that. For my server, I want (and probably need) full control of the networking.

By the way, my day-to-day desktops are MacBooks (we own roughly a half dozen of them in our family). For them, I love the automatic network configuration, the flashy GUI, and the polished surface (by that I mean not just the hardware, but the whole user experience). But that's not where I deploy FreeBSD.

For my NAS, security is paramount because it's my last line of defense against ransomware and (at least some degree of) spear attacks. The data is going to be important enough to rule out Linux, ...
Linux is a perfectly fine operating system. A very large fraction of the world's data is stored on Linux, and it can be perfectly reliable, safe and secure, if administered carefully. There is no reason to hate Linux. Which doesn't mean that Linux is always the best choice; there is a reason that I choose to use FreeBSD for my home server. Mostly it is that administering FreeBSD is simpler for me, because the system is cleaner and more logical in its construction, while having features (such as ZFS) that I need. And the reason FreeBSD is cleaner is that it is simpler, because is not designed around GUI. All I need to say is "systemd", which came fundamentally out of the desire to better integrate the needs of a desktop with the init process, and then went sideways because of the personalities involved (Lennart) and the Linux development process and quality mindset.

Honestly: If FreeBSD dedicates a lot of effort to being a better desktop OS, I'll switch to Linux (which I use heavily, both at home and at work) or OpenBSD (which I used for my home servers before FreeBSD).
 
Today I use FreeBSD because it is a high-quality OS, put together with care, for reliability, logical operation, ease of long-term administration, and so on.

Are those attributes somehow anti-desktop?

If the (small and fixed) set of developer resources get moved towards a consumer and beginner OS, then there is likely to be less effort in making it a reliable and well-built server OS.

Am I supposed to feel guilty for stealing precious developer time?
 
Are those attributes somehow anti-desktop?
Yes. Desktop use adds a lot of complexity. Not just a lot more software that needs to be maintained, but also fundamental changes. Just one example: A single-user desktop system needs the ability to change permission and ownership of devices to match the person who is using the desktop (for example so a non-root user using the GUI can just insert a USB stick and it is mounted). This kind of complexity is not needed on a server, and actually harmful: The identity of the logged-in user on the console in most cases should not influence what happens when a USB device is inserted.

Am I supposed to feel guilty for stealing precious developer time?
Guilty? No. But we have to recognize that ultimately, there is a zero-sum game here: there are only so many developers and most of them are unpaid volunteers (unlike Linux, where a significant fraction of the developers are paid professionals). They can work on one thing, or they can work on another thing, but they don't have enough time to work on everything.
 
There is no reason to hate Linux. Which doesn't mean that Linux is always the best choice;
I agree. There are lots of threads and posts in this forum about FreeBSD vs Linux which could be exact copies of the Linux vs Windows flamewar discussions of years ago. This has been at the time when Linux has become more and more polular, about 10 to 15 years ago. There is no need to hate any OS.

The only thing one can really dislike are persons who push an OS as a religion and consider all others as evil. This includes persons who dislike non-mainstream stuff just for the reason they can not blame others in case of problems. Those guys who only care about having a clean slate, whatever it costs ( of course for the organisations they work for, not for themselves ...).

Let us enjoy all the options we have!
 
Guilty? No. But we have to recognize that ultimately, there is a zero-sum game here: there are only so many developers and most of them are unpaid volunteers (unlike Linux, where a significant fraction of the developers are paid professionals). They can work on one thing, or they can work on another thing, but they don't have enough time to work on everything.

We should not forget that desktop brings awareness. I'd say, a general purpose OS should be able to serve as a reasonable developer workstation to newbie/hobby developers, otherwise there eventually would be no one to advocate it. Anything above that is probably not something that FreeBSD as a project should be concerned with, although if somebody wants to work on flashy GUIs — let them.
 
a general purpose OS should be able to serve as a reasonable developer workstation to newbie/hobby developers, otherwise there eventually would be no one to advocate it.
First there was FreeBSD with plenty of advocates. Then, years later, came the newbie/hobby "developers". There have been plenty of advocates though we could always use more.
 
Right everyone, I feel like I've created Frankensteins monster with this thread here. I came here to ask for help and some here have been in my opinion great ambasadors for BSD and got me up and running with advice and encouragement. My comments and suggestions came purely as a new user of BSD and an uninformed one at that. My intention was not to start a flamewar albeit it mild mannered one. I know I don't have the authority to close this thread but I'm asking the admins to do so. I've used the advice given here to get a second laptop up and running with BSD (ironically had the exact same issue setting up my Dell as I did with the fujitsu-siemens) and think the advice given from most was pretty solid and it certainly helped me, but this thread had definitely gone on a wild tangent and is now in Narnia somewhere.

Is there any way of closing it and renaming it to point to the advice given so that it can be useful to anyone else who might have came across the same problem when setting up a new laptop?

Regards,
Nick
 
Status
Not open for further replies.
Back
Top