.NET and Mono. Call for action: ask for retraction

[Bob Drzyzgula]

I'm sorry, but I've gone through this too many times with
Microsoft. They simply are not to be trusted, period.

Windows NT 3.51, as I recall, contained many features which
gave one the impression that they were honestly trying to
build a decently interoperable system, ready to fit well
into a standards-based, heterogeneous network. Things
like the decent TCP/IP stack, Berkeley print spooling,
Posix system-call and multi-user support, as well as DNS,
DHCP, SNMP and HTTP clients and/or servers. SMS in it's
early design had a lot of promise as a general-purpose
management server; the data collection mechanisms were
based on DMI/DMTF standards and it seemed quite practical
to build the pieces one would need to manage Unix machines
with the same platform.

During the transmogrification of SMB from a
loose, dark art, truly understandable only
through the reading of incomplete, cryptic and
confidential documents as well as network traces and
if-I-let-you-see-it-I'd-have-to-kill-you-level source
code, to a more standards-based CIFS, Microsoft
actually paid to fly some of the Samba developers out
for a CIFS developer conference; the Samba guys were
some of the most vocal and serious contributers to the
CIFS standards developer's mailing list, and Microsoft,
by all appearances, was listening to them.

During the early days of Java, when Sun allowed Microsoft
to do the reference Windows port, Sun officials were
actually quoted in the trade press as saying that Microsoft
had been a "good partner" on Java.

In all of these cases, Microsoft went on to make proprietary
changes to their own implementations of these technologies
which resulted in serious damage to the interoperability
which they had only recently seemed to be working toward.
Windows 2000 broke significant amounts of Samba, for
example.  Not that the Samba guys haven't caught up with
much of it, but the curves that Microsoft threw in their
direction was a distraction that ate up valuable developer
bandwidth that could have been applied to functional
improvements. The proprietary fields in their Kerberos
interface was an egregious attack on interoperability.

This isn't to mean that you are wrong about Petreley being
wrong. This could in fact be the one time that Microsoft
has nothing up their sleeve. But, having spent the last
ten years on this roller-coaster, being alternatively
encouraged and disgusted by their behavior, I personally
refuse to trust them any further than I can throw Mount
Rushmore. All I can say is that all the shoes have yet to
drop -- sneaker mules, probably worn by Bill Gates Himself,
can only be just over the horizon.

Microsoft is like a malignant suburban whore, enticing
a john to go with her to a motel and remove his clothes,
whereupon she will whack him upside the head, steal his
wallet, dispose of his clothes, and scurry away back home
to get dinner ready for her struggling husband and adorable
kids. "Gosh, what a great wife you are, dear; I just don't
know how you manage to make dinners like this on my salary..."

--Bob

On Sun, Jul 29, 2001 at 01:10:57AM -0500, Jeff Bone wrote:
> 
> I'm as big an M$ critic as you'll find around these parts, but I
> think that Nick Petreley is both (a) way off the mark with his
> assessment / analysis of Mono, and (b) stands to infect --- a first
> for the Open Source community --- unnecessary FUD into what looks to
> be a valuable and worthwhile effort....
>