Advanced

One would think, to be very passive aggressive...

Story: Report: Open source software quality is better than proprietary softwareTotal Replies: 7
Author Content
Bob_Robertson

Feb 28, 2012
9:51 AM EST 		One would think that running this test against, say, the Linux kernel source, would provide a list of questionable elements.

One might then provide a flurry of little patches to correct the questionable elements, by initializing variables, and such other.

Maybe there is some reason why this has not been done?
Khamul

Feb 28, 2012
12:26 PM EST 		I thought I had read (elsewhere) that they've been using Coverity on the kernel for quite some time.
Bob_Robertson

Feb 28, 2012
1:10 PM EST 		It might be considered logical, were such tools being used, that the perceived error rate would be substantially lower by now, no?
gus3

Feb 28, 2012
2:08 PM EST 		Well, that's assuming the results from using the tool were actually, you know, analyzed and applied.
Khamul

Feb 28, 2012
4:48 PM EST 		What perceived error rate? Can you point to any real errors or bugs in the kernel that have affected you personally? The kernel is probably the most bug-free large FOSS project out there, and the performance keeps getting better and better.

Unfortunately, a lot of the software that rides on top of the kernel (namely various desktop environments) can't make this claim at all, and is full of bugs and problems.
gus3

Feb 28, 2012
5:36 PM EST 		@Khamul, the perception doesn't have to be on the part of the end-user. It also be on the part of someone doing a code review, someone doing hardware testing, or someone reading a Coverity analysis.
Khamul

Feb 28, 2012
5:41 PM EST 		Ok, but that still doesn't answer the question. I haven't read anything lately that alleges any significant error rate in the kernel.
Bob_Robertson

Mar 01, 2012
1:16 PM EST 		"I haven't read anything lately that alleges any significant error rate in the kernel."

Other than the article this thread is attached to?

As for "performance", the Phoronix benchmarks show regressions all the time. That's not to say that on balance the things that take longer because they are abstracted to work across multiple hardware platforms are worth the price for Linux's legendary variety of supported hardware.

Post a reply to this thread:

A free account is needed to post. Write your comment here first: your text is kept in this browser while you sign up or log in, and is waiting for you when you come back.

Your message:
|||||
10
Markdown: **bold**, *italic*, [text](https://link), - lists, > quotes, $x^2$ for math. Formatting help

Sign up and post (it is free)  or  Log in and post