anyone using Uniquery? 
Author Message
 anyone using Uniquery?



Quote:
>Does anyone have any opinions on Unidata products Uniquery and Unidesktop?

Came from the PICK world to PRIME then on to UNIDATA, and I have found
Uniquery to be almost the same as PRIME INFOS and have found them to be very
easy to use and very inituitive and use friendly. My experiences have been as
a tool developer in several software houses. Our user base has
developed Uniquery statements that are very powerful compared to the user base
that had PICK poducts.

jkoca



Wed, 26 Aug 1998 03:00:00 GMT
 anyone using Uniquery?



Quote:
>Does anyone have any opinions on Unidata products Uniquery and Unidesktop?

A few more experiences using UniServer/UniDesktop:

1. It creates separate "views" for each association of multivalues.
   However, if you have sub-values, these are returned as a space
   delimited list. No separate "views" for sub-values.

2. The Pick translate function, in an I or V type attribute, does not
   work.  However, if rewritten using the Information syntax as a TRANS
   function, it works.

3. Using "views" of associated multivalues seems to be inefficient.
   In many instances, where the max number of values in a multi-value is
   known, I have written I-types to extract each value into a separate
   attribute.  These I-type attributes are then added to the main view
   in the schema for the file.  Big performance improvement.

4. It is important to get the schemas as complete and accurate as possible
   before you invest much time building queries and reports.
   If you have to modify the schema in Unidata, then the schema has to be
   reimported to the PC query tool, which usually forces you to redo the
   queries and reports.

I will shortly experiment to determine the affect, if any, that indexes
have on the performance of UniDesktop queries.

Benefits of UniDesktop seem to be:

1. Users can define selection criteria to extract data, using familiar
   (to them) query tools

2. Only the extracted data is transferred to the PC

3. The data can be easily "joined" to tables held on the PC
   eg: to convert codes to English descriptions

Anyone used UniDesktop/ObjectCall ??



Tauranga District Council
New Zealand



Sat, 29 Aug 1998 03:00:00 GMT
 anyone using Uniquery?
Associated multi-values, and especially SUB values, were often
inaccessible in a way that made any sense.  And having to
constantly maintain the views / schemas was too much work in an
environment where nothing was static.  In other words, it's
something you plug in at the END of your design process, instead
of something that's easily incorporated into the design itself.
  The Objects piece is a neat idea, but not all there, and
something I accomplish much more easily with dynamic normalization
under Hyperstar ODBC, which automatically makes virtual tables of
my multi-values and associations.  Sometimes Hyperstar breaks off
things I'd rather keep grouped in the original table, but there's
an easy workaround for that.  We gave Unidesktop, and moreso the
Objects, an honest workout, but said no thanks.  
Sorry, should have been more detailed in my original post.
jeff george

P.S. SOrry about the format of this post, the emulation on this
lousy laptop has gone to hell.




: >although problems persist with dynamic files.  Unidesktop we tried out
: >and said forget it.

: Could you expand upon this a little?

: Thanks,
: Ron

: Ron Walenciak
: The Software Design Group, Inc.



Sat, 29 Aug 1998 03:00:00 GMT
 anyone using Uniquery?
Jim,
Having worked in all 3 arenas (as well as most others), I have found that
Pick can produce as powerful an Access report as the other 2.  And what
people can do, they usually do.  It just requires a little different
approach in each case.
Regards,



Sat, 29 Aug 1998 03:00:00 GMT
 anyone using Uniquery?


Quote:
>Associated multi-values, and especially SUB values, were often
>inaccessible in a way that made any sense.

I agree; it hardly handles subvalues at all

Quote:
>  And having to
>constantly maintain the views / schemas was too much work in an
>environment where nothing was static.

We fortunately, have relatively old apps with more or less static
data structures.  The down side is that the structures used are
not suitable for direct conversion to views.

IMHO most sites will define a separate set of data structures for
SQL reporting, and write utility programs to regularly extract from
the transaction databases to the query database - ie Data Warehousing

Quote:
>  In other words, it's
>something you plug in at the END of your design process, instead
>of something that's easily incorporated into the design itself.

That said - you can make queries easier if you apply design conventions
which keep the data structures uncomplicated - eg: don't use a field for
multiple purposes depending on context

Quote:
>  The Objects piece is a neat idea, but not all there, and
>something I accomplish much more easily with dynamic normalization
>under Hyperstar ODBC, which automatically makes virtual tables of
>my multi-values and associations.  Sometimes Hyperstar breaks off
>things I'd rather keep grouped in the original table, but there's
>an easy workaround for that.  

Could you clarify what you mean about the Objects piece being
not all there?  I have no experience of UniDesktop/ObjectCall

Could you point me to a reference to Hyperstar?

Quote:
>We gave Unidesktop, and moreso the
>Objects, an honest workout, but said no thanks.  

Thanks for your posting - most interesting



Tauranga District Council
New Zealand



Mon, 31 Aug 1998 03:00:00 GMT
 anyone using Uniquery?
Are there any ODBC driver for Pick systems (specifically Mentor Pro from
GA) out there besides Liberty and Winlink?

- Michael



Mon, 31 Aug 1998 03:00:00 GMT
 anyone using Uniquery?


: >Associated multi-values, and especially SUB values, were often
: >inaccessible in a way that made any sense.
: I agree; it hardly handles subvalues at all
Ok, I'll bite. Can you be more specific here? I rather
thought in my admittedly not objective way that we
handled subvalues rather well.

: >  And having to
: >constantly maintain the views / schemas was too much work in an
: >environment where nothing was static.
Hadn't thought about it much but can see your point of view. I
think new tools to come may make this easier.

Jackie B
Unidata, Inc.



Tue, 01 Sep 1998 03:00:00 GMT
 
 [ 7 post ] 

 Relevant Pages 

1. Problems with UniQuery reports chnaged from System Builder 2 to V5

2. Problems with UniQuery reports chnaged from System Builder 2 to V5

3. UniQuery Question

4. Uniquery to Lotus format

5. Uniquery Problem

6. Run UniQuery commands in a batch?

7. Uniquery Problem

8. UniQuery LIKE question

9. Placeholder char avail in UniQuery?

10. Anyone using 9i Dataguard? Question using CLI

11. Anyone used/using Coromandel Integra VDB/Integra BAB?

12. Anyone ever used MS SQL Server with large databases


 
Powered by phpBB® Forum Software