Re: Abbreviation List Tables Design, aka OTLT

Hi
The bit about dictionaries is interesting Simon. The problem I find with
Pick dictionaries is that they are file oriented instead of system oriented.
The first thing that we did was to design a System dictionary. That is one
related to the particular package that you are working on say Oil
Distribution. Every field that appears on a screen must be defiined in this
system dictionary. Thus when programming one picks up the same definition
for a field, say Company Code, across the whole system regardless of how
many files it is used in.
We also defined a file layout that is picked up by all programs referring to
the file and this of course can link back to the system dictionary.
Regards
Peter McMurray
"Simon Verona" <nomail@xxxxxxxxxx> wrote in message
news:43dbd052$0$82641$ed2619ec@xxxxxxxxxxxxxxxxxxxxxxxxxxxxx
> I find Dawns threads interesting...
>
> As a "computer professional" who learned my trade initially with MV, but
> have used relational databases as well, I find it interesting and thought
> provoking to accompany Dawn on her journey to "revalidate" (if thats the
> correct word!) the MV approach over and above the RD approach.
>
> Whilst I generally agree that MV data-theory (if you can call it that!) is
> generally very flexible over RD theory, I'm not so stupid to believe that
MV
> is perfect or that RD is completely invalid.
>
> MV's problems is that sometimes it's too flexible (it gives you the rope,
> you choose whether to hang yourself or not), perhaps a little
> too-unstructured (there is no requirement to provide any sort of data
> dictionary) and lacks in development interfaces....
>
> On the other hand, RD structure is generally too clinical, too inflexible,
> but well defined in structure (has data-typing, strict referential
integrity
> and a well defined data dictionary). It is also well supported by
> development tools.
>
> Personally, what we need is a MV structure, with better tools for
> data-definition and more integrated development tools.. Whether we
will
> ever get this is interesting....
>
> This is why I participate in these kind of threads.. It's not that I have
> lots of time on my hands (I don't) but that I believe I can learn from
> participating from these discussions and build better applications as a
> result...
>
> Regards
> Simon
>
> "Dale Benedict" <dale_knowspam_benedict@xxxxxxxxxxxxxx> wrote in message
> news:BZyCf.114011$m05.68865@xxxxxxxxxxx
> >
> > "(latimerp)" <"(latimerp)"@comcast.net> wrote in message
> > news:FrGdnVNDD_JILEfeRVn-oA@xxxxxxxxxxxxxx
> >> I agree Simon. Of course you could write a program to create the
> >> dictionaries if this is your *preferred* data structure. Lord
> >> knows you could use some *exotic* tool if this is the way you
> >> want to play the game. What did Henry E, say about *rope*. This
> >> is more like a noose approach. ;) That said the darn DB can
> >> handle it, abet with great effort. Amazing !!!!
> >>
> >> Patrick, <;=)
> >>
> >> P.S. Have you checked the length of this thread(s)?
> >>
> >>
> >
> > Like I said in another thread, any time Dawn starts a thread my reading
> > requirement goes way up.
> >
> > Nothing against Dawn, she just seems to stir up great conversations.
> >
> > regards,
> >
> > Dale
> >
> >
>
>


.

Relevant Pages

  • Re: disk based dictionaries
    ... Best Regards, ... >> I want to store dictionaries on disk. ... >> you are not the intended recipient, please notify the sender at Wipro or Mailadmin@wipro.com immediately ... >> and destroy all copies of this message and any attachments. ...
    (comp.lang.python)
  • Re: Abbreviation List Tables Design, aka OTLT
    ... data-definition and more integrated development tools.. ... Regards ... >> I agree Simon. ... > Like I said in another thread, any time Dawn starts a thread my reading ...
    (comp.databases.pick)
  • Re: Object (string) serialization
    ... serializing/deserializing any ... Best regards, ... Paul Gielens ... > I have a nested object (containing dictionaries) that i want to be able to ...
    (microsoft.public.dotnet.languages.csharp)
  • Re: "literal" objects
    ... >> A thought that occured to me is that classes are implemented as dictionaries ... >very compelling reason to do this, in which case it would be fine. ... Regards, ... Bengt Richter ...
    (comp.lang.python)
  • Re: Dictionary problem
    ... > thanx for any hint how to achieve that ... dictionaries, it won't be necessary to bind a dictionary to a variablename ... and the bind a copy of it to an new element of a list. ... Regards ...
    (comp.lang.python)