Why Linux sucks

[Whaleboy]

"Actually... no. Since the point of auditing code is that audits should be done primarily before releases, less bugfixes means either better auditing or more laziness/secretiveness. See OpenBSD - a prime example of good code auditing."

Look at the linux development and release "traditions". Release early and release often gets code and concepts out in the open fast. They have generally been tested to an extend. What I mean by this is that Linux does pre-Alpha, Alpha and Beta testing usually in public. As such, the majority of bugreports are for beta/alpha software that is in common use. Fixes for production stuff are much rarer, like the BSD's app-level software. Try finding bug-fixes for Bash!!! Its a piece of code used on my OSS *nix systems, is fully GPL'd, yet 2.05 went smoothly from alpha to beta (current) IIRC. Someone correct me if I'm wrong, I dont keep up to date with Debian as much as I should lol.

[Asher]

Bash is older than WinNT, why does it prove anything if a rather simplistic shell program that's old as hell doesn't have many bugs these days?

[Whaleboy]

I dont know exactly how old it is, but I am pretty sure NT is older, but they are certainly of comparable age. Not that its relevant mind.

Bash evolved from BSH, the former is a critical part of non-X programs and server use, as it is the main CLI (there are others of course).

Bash I assure you is not the only production piece of linux software with few bugs. My point was that alpha and beta software, which is tested in the open in the Linux community (what better way to test it? seems to have worked nicely for the last 12 years), has bugs, and its understandable. They are found on the spot and fixed, the fixes made available to the community immediately. Having said that, unlike some windows cases, you are not forced into upgrading, and if you do, you know and can trust what you are installing.

You have thousands of independent, skilled programmers, with different motivations, not working under one corporate roof with one corporate agenda, fixing code, sharing it with each other and everyone else. Its like a big novel being written in the open. Not only is it more poetic, but it is more effective as a means of developing, testing and fixing software than proprietery methods, as the tests on the kernels show, where Linux 2.4.19 came out with 0.25 the coding errors of anonymous proprietery systems (they offered source on the grounds that their names wouldnt be released, but Im guessing MS, Mac, Novel and possibly IBM were among them).

BTW, I'd hardly call bash simplistic!

[Whaleboy]

... sorry 0.25 times the coding errors...

[Q Classic]

hey, hey, guess what?

if you like the os, why does it matter?

it's almost as pointless as arguing which browser is better: mozilla, netscape, opera, ie, konqueror, safari, camino, or lynx.

[Angelo Scotto]

Originally posted by elijah
I dont know exactly how old it is, but I am pretty sure NT is older, but they are certainly of comparable age. Not that its relevant mind.

Bash evolved from BSH, the former is a critical part of non-X programs and server use, as it is the main CLI (there are others of course).

Bash was the GNU replacement of UNIX sh, the first version of Bash came out in 1989 so i think it's a bit older than NT...

Originally posted by elijah
Not only is it more poetic, but it is more effective as a means of developing, testing and fixing software than proprietery methods

I'm happy you're so sure but i've to disagree, open source methodology is NOT inherently more effective than closed source.
It depends from a lot of other factors, not simply the possibility to browse the code; tell me, when an OSS crashes you start yelling "Where is my debugger!" and keep examining the code for weeks just to patch it?
The majority of users will say "Oh well, i'll download the next version when someone will fix it..." and that's all.
I agree about the importance of having the source code available but it isn't a miraculous cure for buggy software, and it's dangerous to think it is.

Originally posted by elijah
As the tests on the kernels show, where Linux 2.4.19 came out with 0.25 the coding errors of anonymous proprietery systems (they offered source on the grounds that their names wouldnt be released, but Im guessing MS, Mac, Novel and possibly IBM were among them).

I think that a similar analysis is a bit simplistic:
how this rate is calculated?
they compared just the kernels? or they compared Linux Kernel with other OSes? and what about severity of bugs? and what tells you that all bugs were caught in both systems?
Finding a bug in a third party software is a hard task, find ALL of them in 2 or more third party software and EVEN compare them...
I don't think this kind of results are reliable

Originally posted by elijah
BTW, I'd hardly call bash simplistic!

Well, for sure it's not simplistic but it seems to me exaggerated to compare it with an OS like you did few lines above...

[Ted Striker]

Guys I'm sure you're all familiar with the Netcraft longest uptime survey.

Of the Top 50 servers with the longest uptime, ALL 50 OF THEM RUN BSD. You might see one Linux server once in a while but it usually disappears after a day:

http://uptime.netcraft.com/up/today/top.avg.html

[Ted Striker]

Some of these servers haven't been rebooted in YEARS, by the way...

[Q Classic]

it's a shame that the longest running windows server died because of that slammer worm.

[Whaleboy]

WOOOOOOOO BSD!!!!

I still maintain that while I prefer Linux to windows, I prefer BSD to linux. The true master.

Each system has its charms though (Im mellow today). People can choose what they want, each has its advantages. Like I say, ease of use out of the box, no privacy concerns, want to run games, run windows. Linux for security, stability and privacy (most important imo in this day and age, with USA PATRIOT act). BSD takes those aspects of Linux into the extreme, and does it damn well!

I think threads like this suck, because they are inherently subjective and pitch systems against each other, that ideally play on different fields. With regards to Linux vs Windows, few would argue that the former has much great potential, while the latter has much inertia (maybe little else, who knows). It is too early in the game to tell. No-one really spoke about Linux being a new desktop champ 2 years ago to my knowledge (although I've only been in the field for 4 months), so give it a couple more years, then we'll see how it goes on the desktop.

Today, if people really dont want to run windows, then at a lil push, they can run linux in most cases (providing they dont have winhardware). For most casual desktop users who wanna play games, run latest software, not concerned about privacy and spyware, then run windows by all means. If they take the effort, theyll reap the rewards. Same thing, to a greater extent, with the *BSD's.

Wait for the tech to develop with Linux, its a young system on the desktop, maybe a couple years old at most. I wouldnt even have put it as a serious competitor until KDE 3.0 was released! . Then we'll talk in 2005.

[Whaleboy]

"If they take the effort [with linux], theyll reap the rewards."

[Ted Striker]

In 1998 they said we'd talk in 2002.

[Whaleboy]

"In 1998 they said we'd talk in 2002."

[Fve Crathva]

Originally posted by Asher
... you need to jump through 100 hoops, sideways while on fire, to do simple things like ... change the desktop resolution.

You have a hard time with CTRL+ALT++ and CTRL+ALT+-?

That is all I have to say, except that Joytech's WinXP video drivers suck more than anything mentioned in this thread.

Edit: And I forgot to mention, this is the best thread ever.

SP