[mdlug] Plan 9 from Bell Labs

[Adam Tauno Williams]

> > I ran Plan 9 once, went "Huh, Ok?", and that was pretty much it.  If
> > someone really wants a blast-from-the-past that is interesting find a
> > copy of NextSTEP/OpenSTEP [and hardware that can run it].  WOW!  Was
> > that ahead of its time (and *glacially* slow).
> NextSTEP/OpenSTEP ran on several hardware platforms, including x86.

OpenSTEP did, not NextSTEP.  I don't recall if I ever used an actual
OpenSTEP system, just an early NeXT system.

> Perhaps the early 68000 systems were slow -- 

Yep, they were.  Of course so was WfWg and Motif [Argh!] on the early
generations.  I remember watching grey block appear,  border paint, then
the title part fill in, then... oh hey, that's a new window!

> If you're interested in experimenting, GNUstep provides the UI (and
> more) for Linux.

I'm very familiar with GNUstep.  I'm a developer of OpenGroupware which
is a very large Objective-C / SOPE / GNUstep code base.  There is a lot
to love about Open/GNU-step.  Things like EOFilter ages before .NET gave
us LINQ [although LINQ is now light-years beyond EOFilter,  I love LINQ]
And the whole Objective-C paradigm is cool.  But on the flip side my
largest project now is reimplementing OpenGroupware's ZideStore in
Python [OpenGroupware Coils: <http://sourceforge.net/projects/coils/>]
because the honest truth is that GNUstep and Objective-C are dead-ends.
No one is investing any significant resources into carrying them forward
and GNUstep in particular has repeatedly broken backwards compatibility.
Obj-C development is forked into the Apple stack and the GNU stack.  The
GNUstep tool-chain compared to that available for .NET / Python / C++ is
really lame and the documentation [for example, but pick anything:
<http://www.gnustep.org/resources/documentation/Developer/EOControl/EODebug.html>] is LOL hilarious.