Re: Mildly OT: dBASE IV

David Cressey wrote:
> "Frank Hamersley" <terabitemightbe_at_bigpond.com> wrote in message
>> Kenneth Downs wrote:
>>> Marshall wrote:

[..]

>> My poison was Clipper Summer 87 and then 5.01 from Nantucket (the 2 gold
>> releases).  When Foxpro came out it was seen more as a clone of the
>> dBase user environment while Clipper was a *woo hoo* compiler (well sort
>> of)!  I ended up using Foxpro on SCO for an app that lasted for 15 years
>> before it was retired.  It was fast and quite reliable.
>>
>> When M$ bought FoxPro and Nantucket went off with VO - soon followed by
>> CA slurping Nantucket up, the future was writ large on the wall.
>>
>> About then I suspended my coding activities and became a dreaded
>> consultant and then even worse, a project manager!

> Your CV makes interesting reading.

As it happens I have rarely coded the various systems I have been associated with in any given language more than once.

Jack of all Trades - yes, master of none, not IMO. Fortunately I have always managed to out skill the other team members even if I have never "seen" it before.

> My trajectory was somewhat different.
>
> I had been programming for some 20 years, as a student, in on campus summer
> jobs, or as a professional, when I got exposed to the data centric world
> view. It changed my thinking.

My epiphany re databases was seeing dbase II on a mates Osbourne in Adelaide round about 1980 or 81 as I recall. Didn't get to use it for many years later until the Foxpro job in 1988. Thereafter I did a major Clipper and a few minor Clipper projects before moving to Access (blech!) and then consulting SQL Server and Sybase ASE. Never tangled with DB_at_ or Oracle.

> During my 20 as a programmer, I had learned along the way, such things as
> interactive debugging (using a tool called DDT on the PDP-1 in 1962-63),
> structured programming (largely self taught, although I did eventually
> read some books on the subject), and languages ranging from assembler to
> lisp to Algol and Pascal.

The glory days of hacking OS's - *nix is nothing new (under the sun).

> My switchover to data centric thinking was occasioned by a confluence of
> factors: a change of jobs, where my mission was to support a boss that was
> interested in using data for decision making, rather than in technology for
> its own sake. A change of mentors: a colleague of mine, a database
> instructor named Bob Ellis, exposed me to the relational model, albeit in
> very low level form. A change of platform, from the DECsystem-10 where I
> knew the operating system internals, to the VAX, where I didn't even know
> the command language. And a change from 3GLs to Datatrieve, a funny little
> language that allowed you to do remarkably sophisticated things with data
> without engaging in a lot of esoteric programming.

dBase (et al) had only the 10 work areas and the "SET RELATION" command with which to link the tables together. This limitation much more than Access stimulated the practice of relational purity to me even though every interaction was in essence cursor oriented.

> By the time Datatrieve's limitations got to be too much for me, DEC had
> released internal copies of DEC Rdb. DEC Rdb comprised Rdb/VMS, which got
> sold to Oracle in 1994. It also comprised Rdb/ELN, which was actually
> available internally before Rdb/VMS. Rdb/ELN eventually stimulated the
> building of Interbase... Firebird.
>
> I never got involved in Foxpro or Clipper or DBASE, except as a hobby.
> Eventually, I started using MS Access for some data management tasks that
> were NOT part of my deliverables, just because it was so easy and so
> ubiquitous.

I never found Access "easy" to produce a truly professional outcome (by my own standards).

> For me, even Oracle RDBMS was a step down from DEC Rdb, as a DBMS.
> However, as a programming environment, Oracle was a step up from Rdb.
>
> For the next 15 years or so, I spent helping people with Rdb and/or Oracle
> databases get more bang for the buck. Perhaps that's why my experience is
> so contrary to what Dawn recounts.

As is most peoples. I don't discount that Dawn hasn't built workable solutions using Pick but that doesn't say anything about the RM. I have built major gizmos using Excel but that says nothing about the RM either.

Cheers, Frank. Received on Tue May 30 2006 - 11:04:43 CEST