[kwlug-disc] given enough eyeballs, all bugs are shallow?
tenger at iseries-guru.com
Tue Feb 16 17:01:42 EST 2010
The many-eyeballs discussions, so far as I have seen, omit a disciplnary
mechanism that seems important to me. Frederick P. Brooks, Jr., in The
Mythical Man-month was talking about hardware design and System/360 in
particular, but it is easy to translate his comment for software.
In most computer projects there comes a day when it is
discovered that the machine and the manual don't agree. When
the confrontation follows, the manual usually loses, for it can
be changed far more quickly and cheaply than the machine. Not
so, however, when there are multiple implementations. Then the
delays and costs associated with fixing the the errant machine
can be overmatched by delays and costs in revising the mackines
that followed the manual faithfully. (Addison-Wesley, 1982,
I speculate that the absence of a disciplining need to adhere to the
published API specification has absorbed quite some time from MS
engineers maintining the office suite, and I can very easily imagine
that the same kind of thing contributes to reputation that Internet
Explorer has for reliability. In promoting OOXML, which is darn nearly
proprietary to Microsoft, Microsoft is setting up a temptation to
inflict the needs of the moment upon the specification.
Mr. Brooks says clearly that that last temptation should be resisted.
The architect of a system, like the architect of a building, is
the user's agent. It is his job to bring professional and
technical knowledge to bear in the the unalloyed interest of the
user, as opposed to the interests of the salesman, the
fabricator, etc. (page 45)
On Tue, 2010-02-16 at 15:17 -0500, john at netdirect.ca wrote:
> kwlug-disc-bounces at kwlug.org wrote on 02/16/2010 02:35:15 PM:
> > Reviving this thread ...
> > Microsoft pitches in re: "given enough eyeballs, all bugs are shallow".
> > http://blogs.msdn.com/shawnhernan/archive/2010/02/13/microsoft-s-
> > many-eyeballs-and-the-security-development-lifecycle.aspx
> > Obviously, Microsoft has no love for that argument for known reasons ...
> > And the Slashdot discussion
> > http://tech.slashdot.org/story/10/02/16/0151226/Are-All-Bugs-
> > Shallow-Questioning-Linuss-Law
> I think Shawn Hernan makes some good points but his conclusions are
> faulty. This is typical of Microsoft rhetoric. Make compelling plausible
> points, aim the reader in a direction and hit them with a faulty or
> incomplete solution. In fact it's just all round good rhetoric, I bet MS
> patented it.
> How many authors have submitted patches to the Linux kernel? I bet that
> number is high. Even more may have submitted patch ideas or identified
> faulty code to kernel developers. These are all invisible "eyes" that make
> developers much more efficient than those that had to hunt for bugs
> themselves. Hunting is usually the time-consuming part.
> And open source developers aren't paid? What!! I read something recently
> that said that over 80% are paid contributors. These paid developers can
> leverage the huge advantage of extra eyeballs to make them more efficient.
> I think that we will always find software projects, maybe even whole
> categories of open source that pales in comparison to their proprietary
> competitors. Proprietary software is good when there is a niche but
> lucrative market. This is also the area where open source often takes much
> longer to root. As a result I think it is easy for Microsoft to pick a
> niche product in Open Source and compare it to a big money maker in
> proprietary source and win. I don't doubt that and I bet they use those
> examples in their research.
> I think another thing that applies is that open source has more heads and
> more heads are better than one. The deep discussion on salient points of
> reasonably popular projects has got to out-perform proprietary equivalent.
> I always like to turn to Netscape Navigator as an example. When Sun open
> sourced the code it took two years for the community (and foundation) to
> clean the code up enough to create a release. I don't think this was
> cleansing proprietary code, it was cleansing fast and loose programming
> that is natural in a lot of proprietary code.
> John Van Ostrand
> Net Direct Inc.
> CTO, co-CEO
> 564 Weber St. N. Unit 12
> Waterloo, ON N2L 5C6
> john at netdirect.ca
> Ph: 866-883-1172
> Linux Solutions / IBM Hardware
> Fx: 519-883-8533
> kwlug-disc_kwlug.org mailing list
> kwlug-disc_kwlug.org at kwlug.org
More information about the kwlug-disc