SCOUG-General Mailing List Archives

>> Next Message >>

Date:	Fri, 28 Feb 2003 07:24:59 PST8
From:	"Lynn H. Maxson" <lmaxson@pacbell.net >
Reply-To:	scoug-general@scoug.com
To:	< "scoug-general@scoug.com" > scoug-general@scoug.com >
Subject:	SCOUG-General: Connectix announcement

Content Type: text/plain

Steven,

I added VPC to the ODIN-associated list as an afterthought. In
so doing I failed to qualify the Office-associated statement
with the possible exception of VPC. I also thank you for
mentioning "Codeweavers" as it was the other fee-based
effort whose name I forgot. That said I otherwise stick with
my statements.

The difficulty I have coming from a mainframe background
into a UNIX(and C)-diluted and -dominated universe lies in the
distinction that exists at times between being clever and being
smart. It's the one when being clever fails the test of
"working smarter not harder". When you look at open source
and specifically open source tools it really comes home.

"Hardly. It's the lack of folks sitting at a desk writing code.
OS/2 got to where it is because IBM paid lots of folks to write
code in assembler and C using Microsoft development tools.
The design was the easy part. Multithreading OS's were
hardly an innovation when OS/2 was implemented."

I never thank you enough for making my case for me.
Open source needs more people writing source code. We
agree. Open source doesn't need more people writing more
source code than necessary, in more languages than
necessary, for more tools than necessary. That's a fine
difference between clever and smart as well as "working
smarter instead of harder".

So what have I suggested? One, eliminate two
compiler-based restrictions: (1) use multi-pass instead of
single pass and (2) allow an unlimited number of unordered
"external" procedures to a compile. Two, retain source in a
database access through a data repository/directory instead
of files in directories. Three, seamlessly merge the functional
capabilities of editors, compilers, and CASE tools into a single
tool with a single interface in a single language. Four,
upgrade to a fourth generation language with assertions, a
two-stage proof engine, and granularity of reuse down to the
statement level.

Of course my single language is based on PL/I and not C,
another example of working smarter and not harder. PL/I
has a simpler syntax: every program element is a statement
and every statement ends in a semi-colon. PL/I implements a
more intelligent, i.e. smarter, form of strong typing, supports
aggregate operands on an element by element basis, supports
the broadest range of data types, allows multiple left-hand
targets in an assignment statement, and the list simply goes
on.

The point of all this lies in using existing (and thus known)
software technology in smarter ways to make code writing
easier, require less of it, reduce the number of languages, and
allow the software to do more of the writing and specifically
the rewriting.

So we begin with HPCalc and the emx set of tools. We will
rely on our experience in writing source, in porting
applications, e.g. from gcc to Watcom, to understand the limits
the tools place on our productivity and the changes to
eliminate them. In that manner we will begin to make the
contributions to open source to encourage more users to act
as contributors.

Somewhere along the way we will come to understand why
ODIN, WINE, Codeweavers, Lindows, and VPC may offer clever
but not smart solutions and how in the aggregate they force
us to work "harder and not smarter". When we do we will
see a common thread of the UNIX and C mentality and the
principle of least effort that leads to one-pass compilers and
the use of "filters" instead of integrated (and seamless)
design.

Meanwhile we have the imminent but oft-delayed arrival of
eCS 1.1. My point exactly.

=====================================================

To unsubscribe to this list, send an email message
to "steward@scoug.com". In the body of the message,
put the command "unsubscribe scoug-general".

For problems, contact the list owner at
"rollin@scoug.com".

=====================================================

Return to [ 28 | February | 2003 ]

SCOUG, Warp Expo West, and Warpfest are trademarks of the Southern California OS/2 User Group. OS/2, Workplace Shell, and IBM are registered trademarks of International Business Machines Corporation. All other trademarks remain the property of their respective owners.