BSD's thoughts of my GNU/Linux Assessment

[Orang_Gila]

I am curious to know what some of the members of the FreeBSD/PC-BSD community would have to say about my personal experiences and criticisms with "GNU/Linux vs Closed/Microsoft" . When I first started my adventure into free software, I did try PC-BSD version 7(?) but it did not work well with my older hardware (i.e. liveUSB); that's one of the main reasons I moved on to GNU/Linux. Now I am reassessing that decision.

Orang_Gila

[fluca1978]

Well, a few random quotes...

who enjoy or are able to spend many, many hours putting together and maintaining systems in university or research environments

I think this is not fair: every type of system requires time to be maintained and tuned. I think you probably should refer to the fact the some systems come pre-installed so that they can provide much more basic services as out-of-the-box.

This is the nature of GNU: individuals and very small groups, with next to no acid testing (mainly user bug reports) and next to no documentation.

No, this is not the nature of GNU at all. This unluckily is what happens for a lot of projects, but is not the GNU nature.

I did try most of the recommended "Libre" distributions promoted on the Free Software Foundations' website (FSF), but most were either complete junk or too restrictive for my hardware.

I agree with you. The problem here is that FSF lists only really free (?) distributions, those that probably have also less funds to support developers, and therefore are partially mature in most cases.

When the GNOME project decided to move over to Gnome Shell, many long term users were left scratching their heads because they could not even figure out how to "power off" their computers; the only option in the drop down menu was "suspend" or "log off"... The only individual I can think of that might not have a use for a power off option is probably a programmer or network server technician. This is just one example of the kind of garbage that we (users) have had to deal with. As stated earier, I chose Debian because I "adapted" to using Gnome 2.3. All the trials of distros using KDE were not as satisfying; it was cluggy, bloated and constantly crashing. In the end, I moved over to MATE (Gnome 2.3 fork) [2], but how long will this last when only a few individuals are maintaining it?

Gnome is a good example of bad managing a project, at least in my opinion. And not for the reasons you listed, but because they are becoming every day much more Linux-oriented, cutting off the rest of us that use other systems. KDE is much more stable in this sense, even if I had to admit that switching from KDE 3 to KDE 4 was not so straightforward, and the same is true for Gnome 2 to Gnome 3.
However, haven't Microsoft changed its UI too? And did the users have the power to keep the old UI or claim about something or even maintain the old UI as for the MATE project? I don't think so.

Debian encourages users to submit using "Reportbug" but the silly part is that the user's contact information (i.e. email address) is broadcast over the internet for every spider to collect and potentially adding them to junk mail lists.

Never used the Debian bug tracking, but other bug tracking systems I used to report bugs did not work as you describe. Seems to me this is a per-project problem, not a wise one.

Rebranding the same stuff just wastes a lot of time from the users' persective and when the dust settles, it just does not feel all that satisfiying.

Again you are wrong. Or better, you are right and wrong at the same time. I agree with you that having a single (or a few) distros and a lot of developers involved in it will speed up code development. On the other hand, distros are there not only to go for a fund raising, but also because they have different aims, opinions, way of developing and testing. Developers are free to work on what they care about, that at the end is the fastest way (for them) to reach the final target. The same thing could be true for the BSD family: NetBSD, FreeBSD, OpenBSD, DragonflyBSD are the major, and then come MidnightBSD, GhostBSD, ManeshaBSD, ....

As it stands, we are all stuck in the middle of the corporate battle between Microsoft, Apple, Adobe and Google.

And when the battle is over, you will have a computer that will do exactly what the winner wants, not what you want.

Lack of respect for users by projects dictating to users how they think users should interact with the packages.

This is another reason why we have so much choices: choose the one that fits you. Do not download and burn ISO images, first read and understand the spirit of the community (if any) and of the operating system, and then try it.
There is also another reason why often developers are not taking into account users' suggestions: users' often don't know what is the best way or solution to implement and just ask for what they want to be, without having enough knowledge or understanding of implications.

I think most users can probably live with less application tool functionalities, but hardware reliability is a basic requirement of any OS. It might be a fair argument that any new user coming into this FOSS community should be expected to learn a little, but do we all really need to become programmers in order to get the hardware up to top functional performance?

You are missing an important point here: hardware vendors often have deals with proprietary software vendors, and that is why we have fancy hardware that is supported by proprietary OSs and not by open ones (at least immediately). In fact, servers that have more standard hardware are very well supported by Linux and BSD systems.

I would never use GNU/Linux for a profit making small business enterprise because of no documentation, constant package breaks and no support; but for personal use it "should" be possible - only time will tell.

I did and I will do. And besides, often open source software provide commercial support that you can get to assist you in a better way than proprietary software often does.

These are my thoughts, and all of the above is not meant to be rude.

[Cobalt]

@fluca1978: Actually, Windows 7 can be easily changed to look like Windows 2000. This is mostly to accommodate weaker computers, though.

[Orang_Gila]

@fluca1978: thank you for your critique, I did not think anything you said sounded rude. Just a few questions though:

No, this is not the nature of GNU at all. This unluckily is what happens for a lot of projects, but is not the GNU nature.

Can you elaborate a bit more about your statement?

...KDE is much more stable in this sense...

I understand your frustrations with project dropping compatibility with organizations outside of the Linux kernel. My focus was on application stability primarily, an example being my current complaint with Kmail, which has stopped displaying the bodies of IMAP emails. I did like the program up until now (once I figured out how to work with akonadi) but have to admit I'm now reaching the end of my rope with it. I chose kmail back in 2010 because I liked the interface, particularly the way I could filter incoming emails to different folders and have each folder display the html & images in different ways, a feature I could not find in any of the other packages like Claws-mail or Evolution. Kmail2 sounds even worse. I never had these problems when testing Evolution or Claws-mail.

However, haven't Microsoft changed its UI too?...

Microsoft's UI was never pretty to start with....

And did the users have the power to keep the old UI or claim about something or even maintain the old UI as for the MATE project? I don't think so."

I partly agree with you there, however I still have at least two licensed copies of WinXP running. I knew enough to stay away from Vista; Windows 7 seems to function well enough but then again I haven't spent any real time playing with it. As stated in the report, I have spent most of my time in GNU/Linux these last 18 months. WinXP had the "classic mode" (if I am not mistaken) for us old timers to revert back to. Both GNOME & Microsoft have behaved in a similar fashion... so I guess maybe in both cases - closed or open - both have restricted user movements. I guess I had high expectations from the GNU/Linux projects from everything I read at the start of this adventure.

Again you are wrong. Or better, you are right and wrong at the same time. I agree with you that having a single (or a few) distros and a lot of developers involved in it will speed up code development. On the other hand, distros are there not only to go for a fund raising, but also because they have different aims, opinions, way of developing and testing. Developers are free to work on what they care about, that at the end is the fastest way (for them) to reach the final target. The same thing could be true for the BSD family: NetBSD, FreeBSD, OpenBSD, DragonflyBSD are the major, and then come MidnightBSD, GhostBSD, ManeshaBSD, ....

Maybe I wasn't clear: I was not suggesting a "single" distro environment, I just meant fewer. You understand my point about splitting human resources (I hate this term!), but isn't that what Microsoft & Apple have been doing for decades now and what portion of the desktop market do they control...?

This is another reason why we have so much choices: choose the one that fits you.

There are many documented studies which clearly show that more choice does not equal more personal satisfaction - this is definitely my personal experience! I examined well over 26 different distros and the thing that finally sunk in was that most of these were rebranding the same packages again and again. I sincerely believe that if this "..BODY WITHOUT A HEAD" ever started to coalesce developers into fewer groups, the big two would have a real threat on their hands (maybe not if GNU/Linux became an ally instead, as I see with BSD & Apple).

You are missing an important point here: hardware vendors often have deals with proprietary software vendors, and that is why we have fancy hardware that is supported by proprietary OSs and not by open ones (at least immediately).

Trisquel forum discussion recommendations were that I should toss our current wifi card and replace it by buying the only card availiable on the market with an open source driver. My attitude was that I bought this computer used to recycle it as well as save some money and it seemed counter-productive to start throwing hardware away just to fulfill the FSF mandate.

I thought this statement made it clear that I understand the reality of proprietary hardware/software relationships.

In fact, servers that have more standard hardware are very well supported by Linux and BSD systems.

Choosing GNU/Linux as an operating system is great for programmers or others with sufficient experience/skills who enjoy or are able to spend many, many hours putting together and maintaining systems in university or research environments (where software must sometimes be written from scatch) but it does not work well for people without these resources who just want a reliable, working system to use as a tool for other endeavours.

Again, I am well aware - from a server-side perspective - that BSD, Linux & Unix servers are heavy hitters; my complaints are from the average home & small business "user" side.

I did and I will do. And besides, often open source software provide commercial support that you can get to assist you in a better way than proprietary software often does.

Can I be a bit bold and cheeky here and ask you what type of small business you run using only GNU?

When I look at most small business, I think of individuals with limited fiscal startup resources and IT costs are one of those items they would rather forgo on their budget sheets. Can these individuals afford numerous "daily" crashes from LibreOffice or Kmail? Remember that I worked as IT in two faculties at the University of British Columbia and I never had regular issues like this on either Microsoft or Apple platforms; network problems once and a while but not stuff like this. For those running a restaurant, bakery, garage, etc... spending this much time fiddling around during a busy day and at the end of it,... I think most people would just throw their computers at the wall in frustration or maybe bake them in an oven to a nice crispness :wink:

[fluca1978]

When I look at most small business, I think of individuals with limited fiscal startup resources and IT costs are one of those items they would rather forgo on their budget sheets. Can these individuals afford numerous "daily" crashes from LibreOffice or Kmail? Remember that I worked as IT in two faculties at the University of British Columbia and I never had regular issues like this on either Microsoft or Apple platforms; network problems once and a while but not stuff like this. For those running a restaurant, bakery, garage, etc... spending this much time fiddling around during a busy day and at the end of it,... I think most people would just throw their computers at the wall in frustration or maybe bake them in an oven to a nice crispness :wink:

Well, I used Linux systems (I don't agree calling them "gnu") for small shops (photographer, car repairing) and also in a quite big factory (200+ employees). But we don't have to confuse the server side with the client side: I used it on both but the desktops were mainly Windows machines (or apples) with floss software (e.g., libreoffice) and the servers were, well, full linux systems.
Your KMail example opens the discussions the need for distributions: I used Kubuntu for a while, since I'm a KDE fan, and I have to admit that it is not the best environment for a stanrdard user, since the update policy introduces a lot of bugs as the one you describe. Other distros, like Debian for example, prefer quite more old software but that has been proven to be more stable. I have a workstation now running a Debian, with KDE and KMail (not update) that is running since 2+ years and I've got no bugs at all.
I agree with you about the fact that having lot of choices is not the best environment, since it is easy to get confused and to waste efforts, but you cannot also impose a single model to all the developers around the planet. Take as an example the OpenBSD distribution. Or shells. Why do we need csh, sh, bash, tcsh, zsh....when what we only need is to input a command and get a result after enter has been pressed? That's because each shell has a philosophy and the user has it too, and when the two matches there is the best experience.
With regard to the gnu philosophy...well, I don't believe that the gnu pjilosophy is to produce some code without documentation or with the lack of stability. Take an example at KDE, that is gnu software, very well written, but with a quite poor documentation (for developers) and look at Qt, still gnu software, but with a very rich documentation. The fact is that documentation is a part of software, but often developers concentrate only on the running part of software. And this is not a fall of gnu, this is an educational fall that begins from academics.

The truth is that computers and operating systems are complex beasts, and require fine tuning to express their power. If we want something simple for home users, we have to accept the risks. Think about wireless modems and routers that come pre-configured and that are cracked every day because they are not pre-configured well. They are simple to use, but not so smart! On the other hand think about how the systems have simpllified and improved during the last uears: setting up a nas 3-4 years ago was a geek affair, now it is quite easy to get it running out of box using some distros available.