Xorg up and running using less than 100 MiB of RAM. Does everything a computer can do. Why should I get some bloated DE? This is not a rhetorical question. I really would like to know.
I used to run Xwindows on a 386-40 with 4MB of RAM. It ran very badly. Part of the problem was swapping; part of the problem was lack of FPU, which is needed for font rendering. I fixed that by buying a 387 floating point processor, but the machine was still barely usable due to lack of RAM. So I upgraded to a 486-25 with 16MB, and it worked acceptably. But: At that time, the only three apps that I ran on the desktop were: xclock (so I knew what time it was), xterm (for doing work, reading mail, copying data), and a huge/complex/powerful data analysis package called PAW (it later turned into ROOT, which still exists and is in use). There was no web browser at the time. Matter-of-fact, there were darn few web sites; ftp was still using 100x more traffic than http on the internet. This must have been 1994. The idea of playing live video on a computer was laughable; playing short music clips (.wav files) was at the edge of what was doable.
So: It is perfectly possible to run one or a few graphical applications that are written for a narrow use case on a small machine. The problem is that today's DEs have a much higher workload, and that is for good reason. For example, the machine I'm typing this on has 16Gig, but it also has about 50 tabs open in two web browsers (most tasks like e-mail, chat, communication are now done in web browsers), it's running a database client, a big data/analytics program, I'm monitoring a few dozen batch jobs (which are running on tens of thousands of CPUs in some data center), plus I have a youtube window in some corner for my music enjoyment.
In fact, macOS probably wouldn't even boot on that machine. and yet it is very popular as a consumer desktop operating system.
Yes, it may seem obvious as to why (Apple wants money) ...
You don't understand Apple. They don't sell you a laptop (like the MacBook Pro I'm typing this on). They also don't sell you an OS (like the MacOS I'm typing this on). Nor a telephone or tablet (which I don't use) and the associated OS, nor a cloud service for serving music or storing data. What they sell you is the complete package of many (or all) of these things. If you could buy MacOS as a separate product, and run it on arbitrary hardware, then you'd be right to complain that it wouldn't even boot. Similarly, MacBooks are not intended to be run with other OSes (even though it is technically possible). If you stay in the "walled garden" of using the Apple ecosystem, then you will have a really good computing experience (at least I think so). If that walled garden is not what you want, or you need/want/desire stuff that's outside the walled garden, then you are not the appropriate customer for Apple's mindset.
You misunderstand. There's a difference between simply asking for donations or contributed work, and doing work with the expectation that people should give back or give you money for it. The FreeBSD Foundation does none of the latter. There is a clause in the BSD license that states those who use BSD licensed software to acknowledge the author of the original work involved. That is all someone needs.
The BSD license is not what holds back the usability of FreeBSD on the desktop. I think the reason is lack of investment. Building a seamless desktop/laptop system, from hardware support through user interface standards and a large stable of useful and usable applications takes an enormous amount of manpower. Hundreds or thousands or perhaps tenthousand people in engineering and engineering management. Microsoft and Apple have those people, and use the profits from selling it to pay for the staff. Linux has that to a smaller extent, the best example being RedHat (which has thousands of engineers working to make Linux better).
One thing we must not forget: The vast majority of engineering that happens for Linux is done by paid staff. People who work fulltime and get a paycheck to make Linux better. Many of them work for RedHat, Suse and so on. Some are paid by the various foundations and non-profits (like the Linux Foundation). But the bulk of them work for companies such as Intel, IBM, Oracle, ... who use Linux, and need to see it function. As an example, look at the Wikipedia page for "Linux Technology Center": Already in 2006 (15 years ago), IBM had 300 people working full-time on Linux development, and already by 2000 (about 8 or 9 years after Linux was born), it had invested a billion (not million) $ into Linux improvements. Amateurs, hobbyists and enthusiasts are a tiny or irrelevant part of the Linux ecosystem.
Obviously, the situation for FreeBSD is different. How many hundreds or thousands of people get a fulltime paycheck for working on the FreeBSD desktop environment? Duh, zero.
In my biased opinion open source in general isn't ready for the desktop.
Simply because it requires you - as a programmer - to think outside your own bubble and try to imagine what others might want from it.
Absolutely correct. Many engineering decisions in open source (even in commercially paid open source) are driven by engineers: I want to build this, I want to work on that. This is particularly true for efforts (such as the BSDs) that rely mostly on volunteers. At companies like Apple or Microsoft (or IBM or Oracle) the decision making is completely different: What do our users = customers want? How can we make our users happy? How can we get them to love our product so much that they will buy it again and again? This means a lot of decisions will be made by user interface researchers, by marketing people who tend to understand the users (and their personal and business needs and wants), by product managers. The individual engineer implements what they are ordered to implement.
Wait a second... Revenue and open source?
There is lots of revenue and profit in open source. Exhibit 1: Look at IBM buying RedHat for many billions. But you don't get revenue nor profit by giving the source code away and selling nothing else. RedHat sold support, and made a huge amount of money. IBM sells support, hardware, the whole computing ecosystem; giving a few million lines of Linux source code away per year doesn't change that business model.
Personally I highly value 2 kinds of environments. I'll take FreeBSD over Windows as a server any day of the week, but having said that I also heavily prefer Windows 10 over FreeBSD (or any other X based desktop) as a client OS just as strongly.
Agree, except that I happen to prefer MacOS as a laptop OS. But I ran with Windows on my work laptop for about 15 years, and I wasn't terribly unhappy. And I can totally see that other people find Windows a good choice for their preferences, habits, and workflows.
adrianchadd.blogspot.com
