Nickname: baobab

Review: how much did they give you ? What you say is really stupid open code is much much better quality code than proprietary

Date reviewed: May 29, 2008 11:26 AM

Nickname: Yayati

Review: Mission Critical!! wWat is this guy trying to say: "Proprietary software applications scatter across the quality continuum, but the best ones tend to be considerably better than open source, and customers don't choose software based on industry averages." Where he has compared hobby, volunteer based desktop open source applications as a yard stick for comparison, that's absurd and contradicting his own research. Another commercial gimmick to lure customers.

Date reviewed: Sep 25, 2007 5:03 AM

Nickname: Jeff

Review: Most of the feedback I've seen from open source developers about Coverity's scans has been quite positive. Sure, it's not a perfect instrument, but the results are a big help for under-resourced projects. Mozilla's bug DB has had a bunch of bugs that were caught by Coverity scans and subsequently fixed.

Date reviewed: Nov 3, 2006 10:32 PM

Nickname: Sally

Review: It's worth noting that the vast majority of "bugs" that Coverity has found on the scan site have not been addressed. And most open source tools have not tried to fix all the bugs. We used Coverity and most of the problems found were bogus. In between Coverity's own stats and my experience, its clear this opinion piece is worthless. The data simply isn't reliable.

Date reviewed: Oct 11, 2006 5:07 PM

Nickname: Constantine Tsardounis - http://www.costis.name

Review: What about giving to open-source developers the same amount of money that is given to closed-source companies? I think that the results will be of higher quality than the current state of closed-source programs.

Date reviewed: Oct 10, 2006 5:41 PM

Nickname: Blackwind

Review: Who is making "critical-mission" software using open-source technology? This isn't hobby-programming. One bug can kill many people. In this article was compared dogs and elephants. I like such stories - it makes me smile.

Date reviewed: Oct 10, 2006 11:12 AM

Nickname: John

Review: What were those properietary software analyzed? Is it fair to compare Overdose (a Yahoo client) with a soft made by aerospace industry? And, BTW, those BSOD, what kind of software produced them?

Date reviewed: Oct 10, 2006 7:53 AM

Nickname: shish-ka-bob

Review: Finding bugs is only one part of the process of getting more reliable software. You should also look at the rate at which bugs are being fixed. There are several open source projects that have removed all of the Coverity discovered bugs. To see for yourself, look at scans.coverity.com and see the raw data. I would have appreciated Mr. Chelf dsiscussing the rate at which bugs are fixed in closed and open source applications. I am astounded at the rate at which the KDE team is cleaning up defects. Every week, dozens of the Coverity discovered defects are removed. (The KDE project has over 4.5 million lines of code, so even a thousand defects is a defect rate that is much above average. I suspect that the work of Coverity & the open source developers is going to result in significant improvements in the stability and security of major open source projects.

Date reviewed: Oct 10, 2006 4:14 AM

Nickname: kjn9

Review: The selection of projects determines the conclusions. OpenBSD is an open-source project that makes serious efforts to achieve code quality, including a rigorous code auditing process. I expect this project would be near the top of the quality scale, although not as high as life-or-death applications that use formal methods to verify their code. Five months ago I suggested to Coverity that OpenBSD should be on their list, in order to evaluate whether its development methodology gives fewer bugs than are found in other projects. OpenBSD is still not on the list reported at http://scan.coverity.com/ - yet FreeBSD and NetBSD (respected projects, but with a conventional development methodology) are on the list. I would actually like to know whether the OpenBSD methodology "works", according to Coverity's metrics. If it does, then this shows one way that any software project, open-source or proprietary, can improve its quality.

Date reviewed: Oct 10, 2006 12:29 AM

Nickname: Unbiased

Review: Shame on BusinessWeek editors for not vetting this article.

Date reviewed: Oct 9, 2006 7:04 PM