OSDev.org

Alboin wrote:?? Really?

I think Plan 9 was brilliant. They simply threw away everything from Unix. They rewrote everything. (The compilers, linkers, editors, etc.) They made a Unix for machines in the 21st century. (20th, whatever.) Meaning, they (gasp!) used the ... MOUSE! (I can hear thousands of die hard terminal fans screaming...)

I do think that a lot of what's in Unix is outdated. It just is. (This can be bluntly seen in xwindows and xlib. Although not a part of Unix per se, it shows the age of some aspects of the beast.) I just don't understand why we are still using software based of software, based of software, which was based of software from the 1970's.

This may explain why I am in osdev.

Note: I like this bold function.

All UNIX clones are "rewritten" as you stated, Not many share code from the original AT&T UNIX.. Just some of the design philosophy.

OpenBSD for example has a highly stable and secure code base.. Care to deny this?

As for X11, A great deal of work has went into the recent 7x series.. I don't see how X is outdated... X isn't exactly tied into UNIX-like systems anyway, It's just sort of the "De facto" standard for the Windowing protocol/server used..

Learn to respect UNIX & UNIX clones.. They are far superior to the crap that's passed off as an Operating System these days.

But from your limited perspective you obviously can't respect the nuances of UNIX.. Shame on you.

Brynet-Inc wrote:All UNIX clones are "rewritten" as you stated, Not many share code from the original AT&T UNIX.. Just some of the design philosophy.

OpenBSD for example has a highly stable and secure code base.. Care to deny this?

I never said Unix wasn't secure or stable, just that it needed some updating.

Brynet-Inc wrote:As for X11, A great deal of work has went into the recent 7x series.. I don't see how X is outdated... X isn't exactly tied into UNIX-like systems anyway, It's just sort of the "De facto" standard for the Windowing protocol/server used..

Are you kidding? Have you actually used xlib? It's almost as bad as the Win32 API. Really. Do we really need support for black and white monitors? There is just so much backwards compatible garbage in it that it has become bloat ware.

Brynet-Inc wrote:Learn to respect UNIX & UNIX clones.. They are far superior to the crap that's passed off as an Operating System these days.

But from your limited perspective you obviously can't respect the nuances of UNIX.. Shame on you.

I have respect for UN.... Oh... wait.. Let me just...Hmmm......Ah!! I'm sorry, I was just running something and my punch tape jammed.

It's stable/secure and most are up-to-date... A lot of the technology implemented in some UNIX-like systems is far superior then any other OS.

So X is bloat ware?.. Sorry but the GUI mess Microsoft released called "XP and Vista" fit that category..

You realize X isn't just limited to x86.. It and lots of UNIX-like systems work on a lot of different targets.. Yes.. Even ones that use black and white monitors. Making it support only the latest "PC-trend" isn't the only top priority.

I indeed have knowledge of xlib.. It's quite extensive and sophisticated.. If you can't figure it out.. You're free to use the GTK toolkit..

And your last remark was just... sad..

You're nothing but an annoying adolescent..

It's quite apparent you think that there is only 1 architecture out there..

Brynet-Inc wrote:You realize X isn't just limited to x86.. It and lots of UNIX-like systems work on a lot of different targets.. Yes.. Even ones that use black and white monitors. Making it support only the latest "PC-trend" isn't the only top priority.

I am sorry. I forgot the current "PC-trend" consisted of those new fangled colored monitors.

Anyway... Question: If a computer does not support a color monitor, than how does it run an entire unix system with xwindows? Don't those type of systems generally just run text only systems?

Brynet-Inc wrote:I indeed have knowledge of xlib.. It's quite extensive and sophisticated.. If you can't figure it out.. You're free to use the GTK toolkit..

One generally has to use GTK, as xlib doesn't offer a widget set itself.

Alboin wrote:Actually, from what I've read Plan 9 goes farther with the file design. (Therefore creating more consistency.) For example, processes are, themselves files and are located in the /proc directory in Pan 9. This allows you to do almost anything you can do with files to processes.

/proc? That has been in UNIXes for ages. Are you sure it's been an invention of Plan 9?

But don't you see? You cannot fork() a process by creating files in /proc. You cannot change a kernel config by copying the /proc/kernel/config.gz from your old system to a new one. Lots of other things.

It's nice that you can get all that information about the kernel and the processes and everything, in text form. But when they change the text format, any program evaluating the info has to be changed. To really get the information, from within a program, you either use a completely different service, or you have to do text parsing (ugh!). And it does look like just another directory, when it is not.

In my eyes, it's another point where the designers chickened out, and instead of designing a user-friendly interface (risking that they do get it wrong), they made it a developer-friendly interface (just dump it to /proc) which they could change on a whim, and made people believe they did you a favor. Same thing as the kernel-driver thing in Linux ("we won't design a proper and stable interface, and you better believe this is a feature").

As for /proc, I would much prefer a system service you can query for all that information, in both struct form (read by programs) and text form (for users). It's OK to provide easy access to this information on the "plain text" level, but making it look like just another directory is deceiving, IMHO. Same thing for network connections - sure you can represent FTP connections as just another directory, but that means every file-handling tool has to be made aware that a copy operation could time out...

Solar wrote:

Alboin wrote:Actually, from what I've read Plan 9 goes farther with the file design. (Therefore creating more consistency.) For example, processes are, themselves files and are located in the /proc directory in Pan 9. This allows you to do almost anything you can do with files to processes.

/proc? That has been in UNIXes for ages. Are you sure it's been an invention of Plan 9?

<History Lesson>
Ahem. *Clears throat*

Tom J. Killian wrote a paper about process as files in 1984 and applied his ideas to the 8th edition of UNIX. Then, in 1991, Roger Faulkner and Ron Gomes furthered KIllian's idea and added to it the creation of ps. After such, in the early 90's, Plan 9 finally finished the idea by having a single file for each process. Shortly after that, 4.4 BSD, Solaris, and Linux copied Plan 9's method of /proc.
</History Lesson>

So, Plan 9 did not technically invent the idea of /proc, but it did produce the version used today.

Ah, OK. Lesson learned. Still, I don't exactly consider it a stroke of genius. It works somehow, it's better than some of the previous methods, and there has yet to come a better-designed alternative. You could say the same for QWERTY keyboards or the Windows Registry.

Solar wrote:Ah, OK. Lesson learned. Still, I don't exactly consider it a stroke of genius. It works somehow, it's better than some of the previous methods, and there has yet to come a better-designed alternative. You could say the same for QWERTY keyboards or the Windows Registry.

There are much better keyboards than QWERTY . It's about the worst design ever (which was the point at the time).

OK, language barrier again. When I sadi "there has yet to come a better-designed alternative", I should have added "that makes it into the mainstream". I know about Dvorak keyboards (which, by the way, still have to prove their superiority), but what good does it do when I have to order them at "speciality" vendors instead of getting one together with the rest of my hardware? (Exagerating to show my point.)

Solar wrote:OK, language barrier again. When I sadi "there has yet to come a better-designed alternative", I should have added "that makes it into the mainstream". I know about Dvorak keyboards (which, by the way, still have to prove their superiority), but what good does it do when I have to order them at "speciality" vendors instead of getting one together with the rest of my hardware? (Exagerating to show my point.)

See, you didn't put the mainstream part in. It was more of a joke than anything however . How do you prove one is superior? The fastest typer in the world set the record on a Dvorak keyboard, is that to say if he used a QWERTY he wouldn't be as fast, who knows . Superior products don't always mean they will be accepted and be mainstream, especially if it means a steep learning curve, could you imagine if dell started including Dvorak keyboards with their machines instead of QWERTY, the completely computer illiterate people would probably pick it up pretty quickly, well as fast as they would a QWERTY. The people who know how to operate a computer to a small extent would be lost. Myself, I have never used a Dvorak, I type about 80wpm, I'm sure if you put a Dvorak in front of me, it'd be nowhere close to that without some serious useage. Now, take this almost impossible feat of measuring which input device is faster, and apply it to something way more complex like a kernel, and you'll see my point. Just because one method works for someone, doesn't mean they all will.

Ready4Dis wrote:could you imagine if dell started including Dvorak keyboards with their machines instead of QWERTY, the completely computer illiterate people would probably pick it up pretty quickly, well as fast as they would a QWERTY.

No, they wouldn't. I still didn't make myself clear: Just because some technology is available, in the technical sense of Joe Average being able to get his hands on it, doesn't mean it is available in a practical sense.

Let's say a guy buys a Dell computer with Dvorak keyboard. Steering a game character with QWE/ASD? Navigating vim with HJKL? Using Ctrl-C and Ctrl-V for copy/paste? Sure you can remap all those keys, at least in decent applications, but until a Dvorak keyboard can be used with the same convenience as a "normal" keyboard in all aspects, it's a novelty, with some specific uses but also several disadvantages.

But all this is rather off-topic. I meant to say that /proc is a working kludge, but a kludge nevertheless, even when the better mousetrap is not available yet to Joe Average.

Kind of OT: My brother who has no programming intrest and average computer knowledge changed to "svorak" (swedish dvorak, with åäö) , he installed a XP driver for it, and rearranged the keys on his original dell keyboard. It took him about 2 weeks, but now he writes equally fast on both types.

bubach wrote:...rearranged the keys on his original dell keyboard...

I'd be really interested in trying Dvorak one day, but right now I use three keyboards: The one of my laptop, a Microsoft Natural keyboard and a workstation at the office that, at times, is used by other people, too.

Rearranging keys? No deal.

If you need to see the characters (easier at first), you could always use stickers..

Hopefully, something like the Optimus Keyboard will finally come out (supposed to in 2007) and make this entire conversation obsolete