Is ANY OS Ready for the Desktop?

That was a nice article, and I agree with the main point - that no OS will ever be perfect. But this applies to anything in the universe, and is a very weak conclusion to arrive at. In fact, it's accepted that nothing is perfect, which is why we make decisions in life.

With operating systems, there are decades of research and testing which use specific criteria to determine which operating system types are the best choice for a specific problem set. From the average user's desktop OS to mainframe OSes to clustering OSes, there are a myriad of criteria from which to pick to evaluate an OS. Selectively choosing the criteria can sway the evaluation in the favor of one OS over the other, effectively mitigating the results of real testing.

That said, you picked some great categories on which to evaluate this category of software. However, there isn't much detail. This is not a criticism, but a point of fact: it sounds as though your knowledge of operating systems is limited. At face value, the argument you've assembled and the evaluation you've made has been repeated countless times in the past, albeit in a more controlled and analytical way.

For starters, I'd suggest picking up anything by Tannenbaum (creator of the Minix "learning" OS, the inspiration for Linux). Tannenbaum His Modern Operating Systems book is a great first look at how OSes work under the hood, and includes a great discussion of the various types of kernels (although he clearly has a preference here :) ).

With that said, I'll go ahead and make my own generalizations about Linux. I first installed Linux in 1996 from a handful of floppy disks. There were only a handful of WMs at the time, and boy, did they suck. Since then, we've all watched the various free Unixes transition from server-oriented software to something more appropriate for the desktop. This transition hasn't been easy, given the nature of open-source development, and I dare say that without the always pragmatic Linus Torvalds looking out for the desktop user, things might not be as rosy as they are today. It was only recently that the Linux kernel gained user-level preemption on a wide scale, and even then only with patching. The versatility of Linux is equaled only (at times) by the frustration of the one-OS-for-all-solutions model. Decisions that are great for the desktop may not be so great for the Web server.

I'll avoid going line-by-line into the generalizations made in the article, but suffice to say, a few more facts and a little less pontificating would have served the analysis much better. As it stands, it's a provoking starting point for a colloquial discussion of pros and cons, but not a starting point for a real discussion applicability.

Reply Like

Aaronjb...you are correct. My understanding of how each OS works under the hood is very limited. I dare to say that this holds true for at least 80% of all users. The points were chosen specifically for the generalization because they apply equally to all OS's. I'll admit it is by no means an exhaustive review. I think that its the fine points that get in the way sometimes. The articles that have been posted so far have been full of fine points( install times, lack of support, lack of security)which have really clouded the issue rather than cleared it up. I really doubt it will ever be truly cleared up. Although the conclusion I arrived at may be considered weak, I think it is the most truthful one to arrive at. To often we try to apply our own viewpoint to all situations when it is so limited by our predjudices that it can't come close to applying to all situations.

To say it plainly...just because one individual had great success with a particular OS does not make it the best choice. Conversely just because one individual has a lack of success with the same OS doesn't mean that it should be ruled out as a choice. The choice must be weighed by each individual, taking into account his/her preferences and predjudices.

The point of my article wasn't to pick apart the different OS's, but rather our way of looking at them. You mention the years of study and design that have gone into all of the differing OS's and yet it still seems that there is no good choice. Maybe those studies have been to narrow in focus. I don't know. I only know that I still judge my OS by how well it fits my needs and wants. For myself Windows still is a better fit than Linux. But I'm not really concerned with security(if some hacker really wants to know that I visit this site he's welcome to that info, as well as the fact that I enjoy playing Theif 3). I am concerned with having the full functionality of my hardware and being able to play my favorite games...which happen to be designed for windows.

Eventually someone will figure out that the OS needs to be able to adjust to the user and not vice versa(even linux still requires that you learn about it and not the other way round). Maybe the real issue is that computers are straight up logic and humans are stright up messed up.

Reply Like

I am concerned with having the full functionality of my hardware and being able to play my favorite games...which happen to be designed for windows.

That's the bottom line, really. Sometimes when I think about high-level computing problems, I ask myself "Could my grandmother use and understand this?" Maybe it's a strange question, but so much of what we do is computing for computing's sake, and not computing to solve a real-world problem.

Eventually someone will figure out that the OS needs to be able to adjust to the user and not vice versa(even linux still requires that you learn about it and not the other way round). Maybe the real issue is that computers are straight up logic and humans are stright up messed up.

:) That's a pretty good summary. But in the end, who designs the comptuer systems? We do. We control the hardware and the software, and we're not doing a particularly good job at either.

Reply Like

:) That's a pretty good summary. But in the end, who designs the comptuer systems? We do. We control the hardware and the software, and we're not doing a particularly good job at either.[/QUOTE]

I know the real answer to that...it comes down to one simple thing, money.

The build an OS and design a game so they wont go over budget.So what they do is make sure it will run even with flaws which everything has, sell it off to us and then once they have gained back some of the cash they spent they then use it to "fix" it.That way more people will buy it with the "fix" added to it and so forth and so on. The "fix" doesnt always work either because get cheap with the budget for the fix and the designers say "if your not going to pay me to do it right"...were rightback where we started.

My 2cents lol

Reply Like

That is correct OSX is written on FreeBSD which is written on UNIX... Linux is also written on UNIX so there is some interoperability between OSX and Linux there.

Linux was greatly inspired by MINIX (a small version of UnIX) but has NO code from the UNIX world. The interoperability part is the POSIX standard (and a bunch of otherrs).

FreeBSD came from BSD version 4.4, BSD in turn used to be a modification of the AT&T UNIX (System V), which in turned was shipped off (without any System V code) to fill in the missing pieces because AT&T didn't allow BSD to run on top of System V.

OS X runs on Darwin kernel is a derivative of Mach kernel (which somehow comes from Unix, I am not too sure) and FreeBSD (from the 4.x branch I believe). OS X uses XFree86 4.3 (rather old compared with Xorg 7.1) for it's graphics server.

Reply Like