[Date Prev][Date Next] [Thread Prev][Thread Next]
[Thread Index]
[Date Index]
[Author Index]
RE: Pick?
- From: KThorpe <kthorpe pricetrak com>
- To: michael twopoint com, redhat-list redhat com
- Subject: RE: Pick?
- Date: Mon, 7 Sep 1998 13:55:00 +0100
>Date: Wed, 02 Sep 1998 20:19:02 +0000
>From: Michael Jinks <michael twopoint com>
>To: "redhat-list redhat com" <redhat-list redhat com>
>Subject: Pick?
>
>Just saw the press release on redhat's site, and I've been looking
>around Pick's web sites, and I've gotta say it all sounds a little too
>good to be true. All I know about Pick is that its earlier incarnations
>predate UNIX.
Ah. You've just stumbled upon the best kept secret in the database world.
Pick began in 1967, written by a chap called Dick Pick (I'm not kidding) to
support the parts inventory for the Cherokee helicopter design project at
McDonnel Douglas. It was written on paper and the hardware then built round
the database. As such it is a 'real world' database, not some fancy
post-grad research project like everything else I've seen.
We used to have a GBP50,000 minicomputer, we now have a GBP2000 server
running Linux and jBase (a Pick cloneish.) This is running a database of
over 1Gb. Some files have over 5 million records in them. It widdles all
over Squirrel Server and Orrible.
>Could anyone with some Pick experience comment on the general layout,
>ease of use, that sort of stuff? Much of our business involves chunking
>through large databases in a hurry, and there are some Pick tools that
>look like they might be pretty serious.
I've been using and developing for Pick for over 10 years (oooohhhhh that
makes me feel old and I'm only 32). It is significantly different to
anything else, but then you're using Linux so you're not afraid of steep
learning curves.
The database itself is post-relational (even though it existed before
relational databases). All records are multi-dimensional. Each
field(attribute) in a record(item) can hold multiple values, each of which
can hold multiple sub-values. Fields are variable length so don't waste
space.
The programming language is a version of Basic called DataBasic. NO Don't
switch off just yet. This version of Basic is on steriods. All variables are
dynamic, three dimensional arrays. Writing database apps it a breeze. There
is also a query language, Access, which makes SQL look like nursery rhymes.
>What's the OpenSource/Pick situation like? I get the impression that
>freely available tools do exist. Otherwise, what's the state of the
>Pick art like? Worth the $300 to try it out?
State of the art Pick isn't Pick anymore, but jBase......
jBase is written by a bunch of defectors from McDonnel Douglas Information
Systems - the design team for McD's own Pick.
There is a version for most IXes, especially Linux (RH5.0) and for NT/95 if
you're really sad. This is a shell environment and front end for gcc or MS
Visual C++ and produces real object code (.so or .dll as well).
The NT version also has 'OBjEX' which is a DCOM interface to the system. We
have used this to integrate our database into MS Excel among other things.
Go to www.jbase.com and blag the FREE 3 user developer's trial copy for
Linux and start asking me questions.
>-- Michael Jinks
>mailto:michael twopoint com http://www.twopoint.com
>Systems Administrator, Two Point Conversions, Inc.
=========================================
Kevin Thorpe
Purchasing Index (UK) Limited
LONDON, England
Tel: (44) 171 350 4200
[Date Prev][Date Next] [Thread Prev][Thread Next]
[Thread Index]
[Date Index]
[Author Index]