[NTLUG:Discuss] Microsoft stock slips after report, Dell comments

[Christopher Browne]

On Sat, 12 Feb 2000 14:14:35 CST, the world broke into rejoicing as
Steve Baker <sjbaker1 at airmail.net>  said:
> sunflower1043 at juno.com wrote:
> 
> > which said in part, quoting ZDNet:
> > 
> > " 'Our customers do not want us to sell them products with over 63,000
> > potential known defects'....stated one of Microsoft's Windows development
> > leaders, Marc Lucovsky, in the memo. 'How many of you would spend $500 on
> > a piece of software with over 63,000 potential known defects?'
> > .....
> > "According to the Microsoft memo, the Windows 2000 source-code base
> > contains:
> > .....
> > "Overall, there are more than 65,000 'potential issues' that could emerge
> > as problems, as discovered by Microsoft's Prefix tool. Microsoft is
> > estimating that 28,000 of these are likely to be 'real' problems..."
> 
> Much though I hate M$ products for their lack of reliability, I think
> you have to put that figure of 65,000 (or 63,000 or 21,000) known bugs
> in context.
> 
> I was reading the comments about this revelation on Slashdot - and there
> are some important things to realise.
> 
> 1) The Debian bug tracking database for Linux has over 10,000 bugs listed.

Add an extra bit of context; the Debian database is listing 10,000
problems associated with a diverse set of applications from a wide
variety of sources.

To make the W2K database comparable, it would have to include defects
not only for W2K, but also for things like:
  - MS Office
  - Lotus Millennium Suite
  - Lotus Notes
  - Netscape Communicator

> 2) Of the 65,000 or 63,000 or whatever in the W2K bug tracking
> database, only 21,000 are considered "bugs".  The rest are things
> like ergonomic issues that are labelled "defects" - not "bugs".
>
> 3) Even then, some of these so-called bugs are things like spelling
>    mistakes in dialog boxes, variables declared global when they could
>    be local.  We are not talking 21,000 BSOD bugs here.

Defects can indeed extend from a box being drawn a little too wide to
BSODs.

> 4) Lots of those errors will be in the gazillions of device drivers
>    out there.  Since there are dozens and dozens of (for example)
>    video adaptor drivers - and you'll only be using one of them,
>    the number of bugs you'll actually experience will be a lot
>    smaller than the full 21,000.  Most people (aparrently) actually
>    install less than 10% of the code that comes with W2K, so perhaps
>    2,100 bugs is a more realistic figure for the actual number you'll
>    experience in practice.

I'd think it necessary to count these up together at least for the
more common sorts of devices.
--
The question  of whether a computer  can think is  no more interesting
than the question of whether a submarine can swim.
-- Edsger W. Dijkstra
cbbrowne at hex.net - <http://www.ntlug.org/~cbbrowne/lsf.html>