The cadence of Software Engineering.

Bill Cattey wdc at MIT.EDU
Tue Dec 22 00:15:40 UTC 1998


Dan's note appeared a couple days ago in the seueak list:

Excerpts from pm.squeak: 14-Dec-98 What's happening with Versi.. Dan
Ingalls at wdi.disney.c (1102*)

> Folks -

> I thought some of you would want to know what's happening with Version
> 2.3, since we usually put out a release within about a week after beta.

> Well, exciting things have been happening and, since there wasn't a huge
> amount of pressure on this release, I thought I would drag things a bit
> in order to make the final release more interesting and, hopefully, more
> enduring.

> Several things have happened since the beta release...

> 	We have coaxed another 30 percent out of the interpreter (some figures
> are even higher).
> 	We have the beginnings of Squeak Pages actually working.
> 	Major improvements have been made to the Balloon graphics engine
> 	Ian has a preliminary Jitter II working.
> 	A number of you have found (and fixed!) various little bugs in the beta
> release.

> The new Jitter work is integrated through the Pluggable Primitive
> interface.  As soon as that interface is stable, we will be able to
> issue new Jitter plug-ins independently of the Squeak release schedule. 
> It's my hope to reach that point and issue 2.3 by the end of the week.

> Please stand by

> 	- Dan

The parts that struck me hardest were:  Paragraph 1:  one week after a
beta, the real version comes.  The mention of re-admission of the
Jitter, and a 30% speedup in the couple weeks since beta, and the
expectation that it was all gonna be ready by the end of the week.

In answer to your query about what I meant by conventional software
engineering:  Well, these days I manage C and perl programmers.  In the
past I have coded C and the object-oriented layering of C that the CMU
Andrew Toolkit folks did.  The closest brush with C++ I got was when
they migrated the Andrew Toolkit to C++.  Seeing what had to be done to
the simple CMY OO semantics to shoe-horn them into C++ left me with a
permanent bias against C++.

I have also, in the dim ages programmed in Lisp.  I'm aware of what
kinds of productivity improvements programmers can enjoy in such
environments.  But the Lisp folk always seemed more intent on a
pedagogically correct Lisp world than a useful application delivered.  I
think Lisp fell into the chasm between conventional and highly
productive software engineering, but that Squeak may be bridging that
chasm.

-wdc





More information about the Squeak-dev mailing list