Maximum PC

How does that one grab you?!

From a recent Linux Magazine article:
http://www.coverity.com/files/linux_article.pdf

This is the website of Dawson Engler, the Stanford professor who was responsible for this research: http://www.stanford.edu/~engler/


This might prove to be an interesting research area for one or more of us programmers around here.

And before I forget, let's have a round of applause for Stanford. Their grad students and faculty are doing some incredible work and have become the 1930's Yankees of the compsci world. Seriously, what happened to Berkeley, CIT, MIT and CMU?! What the hell does IIT stan-ford again?

sounds like something that microsoft would like to take off with and over publize till we all got sick of hearing about it.

Very sharp double-edged sword there. MS knows that they'll get their turn soon enough. Software is very complex - the fact that there are over 1000 bugs in the kernel probably doesn't surprise anyone who has worked on a large project like Linux or OpenBSD. 99% of all software has bugs - the differences come down to how gracefully you fall.

OK - this makes the fourth thread that I haven't recieved an email notice from.... wtf is going on. /me goes to check his account settings.

I'd be very surprised if MS actually did that.

Wouldn't surprise me if some MS advocate out there started advertising in some marketing scheme they have though and it ends up being more public then we're used to.

Don't underestimate pMS! What I meant was that MS possibly has or is likely to use a similiar tool on their own code base. If they go and make a big deal about the bugs in Linux, not only are they going to catch a bunch of negative feedback from the OSS guys, but they also run the risk of someone that conducted the code analysis saying "wtf, there were over 3000 bugs in win2k".

Anyways, this has nothing to do with MS at this point - hopefully this will help make linux better. That is what is important.

They'll just use it as a selling point: Windows, now with 2000 more bugs than That Other OS.

One would hope. OTOH, it might slow things down too..after all, I suspect you'll see a bunch of programmers pointing fingers at each other.

Kinda like it is now, except with all the hype being about MS?

Yeah I am getting tried of the news talking about the many bugs in windows. Yes we know it has not changed in over 8 years dont need to talk about it any more.

Speaking of Windows..... I did an install for a friend a couple of weeks back. I really wish they would make a version that you didn't need to reboot 15 times to finish a damn install!!!

Heh.

I consider rebooting for anything other than a new kernel to be heretical.

I guess the folks in WA don't share your ideals...

It doesn't. The open source community has far more human resources to fix these bugs much faster than the relatively handful of people Microshaft has working on their kernel.

Another thing, at least they FOUND these bugs. Much of the problem is finding these run-time bugs in the first place.

You need to read the article. The OSS people did not find these bugs. A professor at Stanford and his grad students found them. Sure, the only reason they were able to analyze the code was because the software is open source, however, that doesn't mean that a software company hasn't already paid for their services - and I doubt they'll be doing many more OSS code analysis for free now that they have generated some buzz. So what do we know? Well, we can't assume just because OSS has more eyes that numerous bugs don't exist. We also can't assume that a software company hasn't used a service like this to fix their software. We can't assume that they have either. We just don't know.

I used to buy into the OSS mantra hook, line and sinker, but after partaking in some software projects and taking a couple of software engineering classes, I realize that having MORE people doesn't necessarilly equate to better software. I'm not saying OSS is inferior either, but it shouldn't surprise people that there are still bugs. Everyone wants to quickly code some cool appy, but no one wants to do the dirty work for free, right? Some OSS projects reach a critical size that brings in a large number of programmers, but most OSS projects have 10 or fewer devs, and even in the case of the kernel, one of the larger oss projects, you still see a high number of bugs.

Anyways, I didn't intend for this to become an OSS vs commerical software or Linux vs MS thing. I just wanted to point out a cool research project that emphasised how difficult it really is to write bug free software.