Assume someone from a purely RDBMS background wants to learn Pick.
Let's say they've downloaded and installed UniVerse. What advice would
you give them to help them get started so that they can quickly gain
an appreciation as to why we like the MV/Pick model so much? Assume
they are highly intelligent and very knowledgeable about relational
theory.

Mike.

In article <1b0b566c.0407141656.2568b3c8 (AT) posting (DOT) google.com>,
michael (AT) preece (DOT) net says...

The "problem" I run into with these people is that it is hard to get
them to stop thinking in terms of how they would accomplish the same
task in a RDBMS.

An analogy would be to learning a foreign language (spoken). Many
people make the mistake of trying to translate (in their head)
everything they're hearing, back into their native language. Not only
is this too slow for them to keep up, but it hinders their progress.

I'm not saying the RDBM person should forget everything they know, just
try not to keep thinking how they would do it in RDBMS.

I remember that I was not overly impressed with Pick when I first
started using it (mid 80's). It takes a little time working on real-
world projects before one starts to realize the niceties of the data
model.

The one thing that most RDBMS people will probably cringe about is
having to maintain data integrity at the application level. Lack of
"modern" admin and dev tools can also put them off.


--
Kevin Powick

Kevin,

It might be considered a bit rude to respond to a legitimate question by
refering to "these people". This isn't comp db theory after all.

Check out www.jes.com for documentation and Pick Basic, the basic language
of Pick and Pick-like databases.

Do a search on "Malcolm Bull" and find his on-line encyclopedia.

Mark Brown


"Kevin Powick" <nospam (AT) nomail (DOT) com> wrote

Kevin isn't rude - he is just refreshingly direct. A woman would probably
say that he lacks "sensitivity". So be more sensitive, Kevin, and refer to
them as "these idiots".
or better yet, "these idiots who drive blue cars".
BobJ
PS - Dawn, please note that the feminism was intended as irony.
BJ
"Mark Brown" <mbrown (AT) drexelmgt (DOT) com> wrote

He isn't being rude. Having worked with a group of people that moved
from a Honeywell/Bull mini to Pick and watching them create an
application system of fixed length file names(8.3), fixed length ids
and fixed length fields. Oh yeah, they also created records with
fixed number of values and sub values, but unlimited number of
attributes. No, it wasn't an index file. But it did ED better.

rg

If you are using an engine such GOOGLE to search for Malcolm Bull's Pages, it
might be more productive to use the search strings

"malcolm bull" multivalue mb-cyclopaedia

as this sifts out loads of non-relevant stuff - such as 1200 Pages for a local
history website which I run for my local newspaper.

Note my spelling of MB-CYCLOPAEDIA

Regards

Malcolm Bull

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

If you have difficulty reaching me at AOL, please try
*
****MalcolmBullTraining (AT) hotmail (DOT) com

Remember our websites:

+ Books & & MB-Master Self-Tuition courses for users of Pick & MultiValue
systems:

**** http://members.aol.com/_ht_a/mbtraining/index.html

+ This is ... series of books for users of Pick & MultiValue systems:

**** http://members.aol.com/_ht_a/mbtraining/thisis.html

+ PDF versions of our Books & MB-Master Self-Tuition courses for users of Pick
& MultiValue systems:

**** http://members.aol.com/_ht_a/mbtraining/pdf.html

+ MB-Cyclopædia for users of Pick & MultiValue systems:

**** http://members.aol.com/_ht_a/mbtpublish/index.html

+ MB-OnLine Publications for users of Pick & MultiValue systems:

**** http://members.aol.com/_ht_a/mbtexts/index.html

+ Our SB+ Reference Pages:

**** http://members.aol.com/_ht_a/mbtpublish/sbplus.html

+ Our This is ... series of Books for users of Pick & MultiValue systems:

**** http://members.aol.com/_ht_a/mbtraining/thisis.html

+ The general Directory to our Training & Consultancy services:

**** http://members.aol.com/_ht_a/mbtraining/directry.html

+ Our series of MB-Master Self-Tuition Courses for users of Pick & MultiValue
systems:

**** http://members.aol.com/_ht_a/mbtraining/mbslist.html

+ Our series of MB-Reference Quick Reference Cards for users of Pick &
MultiValue systems:

**** http://members.aol.com/_ht_a/mbtraining/mbqlist.html

+ Request a No-Obligations Quotation for our Training Products & Publications:

**** http://members.aol.com/_ht_a/mbtraining/proform.html

In article <bqoJc.43741$Fy.31590 (AT) twister (DOT) socal.rr.com>,
mbrown (AT) drexelmgt (DOT) com says...

Excuse me?

What then? RDMBS people? These guys, These girls, These wonderful
human beings of no specifically mentioned gender with a background in
RDBMS?

Set follow-ups to: alt.society.politicalcorrectness.

--
Kevin Powick

Mike Preece wrote:
I think stressing some of the key benefits so he might like some of the
flexibility he gets with the pick platform.

show them how much data you actually have in pick and how much disk
space it uses. show him how you can easily share databases by using q
pointers. show him how easy basic is and how you can overcome pretty
much any design dilema with a tiny bit of code. show him how you don't
necessarily need to create large amounts of "bridging" tables just to
facilitate the design.

I'd say the best way to learn it would be to take a n sized project
from relational world and then create the database in pick. first
start in pick creating the relational database (same tables and such)
and then _fix_ it to use the MV features.

You mentioned that he is technical. Show him raw dumps of frames so he
understands how pick keeps its speed up and its filesizes down.

-Adam

Mark:

The exact site at jes.com is:

http://www.jes.com/olbooks.html

He has one of the original PICK books in pdf format, plus links to mvDbms
documentation.

Pretty cool.

Bill

P.S. Its a darned shame that all those PICK books aren't in .pdf format and
available for download for $4.95 each.

"Mark Brown" <mbrown (AT) drexelmgt (DOT) com> wrote

michael (AT) preece (DOT) net (Mike Preece) wrote in message news:<1b0b566c.0407141656.2568b3c8 (AT) posting (DOT) google.com>...
Hi,

I would not even bother attempting trying to convert them to a pure
pick environment, and with UniVerse there is no need. UniVerse comes
with the best SQL implementation on a MV database that I have used,
but the learning curve can be a bit steep, depending on your background.

First, review the 'UniVerse System Description' manual to get a overall
view of the database structure. The idea here is that they will come to
accept that UniVerse (or any pick system) is not just a database per se',
but is a complete database and development system.

Next, study the 'UniVerse SQL Administration for DBA's','UniVerse SQL
Reference',' then 'UniVerse SQL Users Guide'. Once you absorb all this
information, you will be up and running with a MV based SQL system and
all the relevent advantages.

Layering this beast would then present all the usuall problems, refer to
the UCI interface manual, or Java and / or UniObjects from Windows, maybe
ODBC could be used as well.

Now, how you would approach this problem with D3 I will leave for those
people who are still using it.

HTH

Barry