the death of dreams (was Not again (BSD etc. ports))

Jeffrey L Bromberger jeffrey at
Wed Mar 27 03:50:03 AEST 1991

In article <1991Mar26.121309.5351 at> mvadh at (andrew.d.hay) writes:
>this is exactly why i suggested we start with miniframe ctix 5.X.  the
>miniframe is a *very* close relative of the 3b1; binaries can be
>ported, for example (are drivers also portable? thad?).  i thought
>that by diffing the src trees for ctix 5.X and 3.51m, we could easily
>spot the hardware dependencies.  this would reduce the porting job to
>an editing job, which is something even i could do -- and i would come
>out of it knowing a lot more about the kernel and kernel programming,
>which was one of my major goals in this project/dream.

But you seem to forget one salient point.  AT&T will not release the
source code to a non-educational site.  I had a nice long (3 hour)
talk with Unix Licencing on Friday, and unless you're a
college/university, you're SOL.  And even if you *do* get source, it's
held under the one-processor thingie.  Next off,where you gonna get
CTIX source?  Scotty's gonna beam it down for you? :-)

>but response to my proposal has been sparse; i've only seen about a
>half-dozen replies from people willing to commit money toward source

I believe they quoted me (as a university person) $4000 for a source
distribution.  But first, I'd have to buy a source licence, if I
didn't have one.  Well, CUNY has one, so that's $$$,$$$ I don't have
to worry about.  But $4K is still a hell of a lot for a freaking tape.
Then, it comes 9-track format; you'd need to get it onto something a
3b1 could read.

>oh well...

[going into musical mode]
I quote Billy Joel (from an old song)

Don't you know that only fools are satisfied.
Dream on, but don't imagine they'll all come true.

It's time we realize that the problems with distribution are going to
be insurmountable for a bunch of guys (any gals out there??) doing
this as a hobby.  We're not going to invest mucho dollars in this, nor
will we invest the time.  It's a hobby, folks.  Ain't nobody gonna
make money off this.  And Ma Bell has the bottom line of $$$.

And a different question comes up.  Just *what* do you want from SVR3
that we don't have?  And I mean the kernel, not the utilities.  
Shouldn't the effort be to use what we have and build upon it?  Why do
we insist on reinventing the wheel (albeit a new and improved wheel)?

Jeffrey L. Bromberger
System Operator---City College of New York---Science Computing Facility
jeffrey at			jeffrey at ccnysci.BITNET

More information about the Comp.sys.3b1 mailing list