[Search for users] [Overall Top Noters] [List of all Conferences] [Download this site]

Conference ulysse::rdb_vms_competition

Title:DEC Rdb against the World
Moderator:HERON::GODFRIND
Created:Fri Jun 12 1987
Last Modified:Thu Feb 23 1995
Last Successful Update:Fri Jun 06 1997
Number of topics:1348
Total number of notes:5438

1165.0. "" by SCAACT::ADISESHAN () Wed Jun 17 1992 22:46

We're working on a State Govt. RFI.
They are getting to some pretty detailed database specifications.

From the following description, can you get a feel whether they are
more inclined to Rdb or AS/400 or Oracle ?

Does Rdb comply to FIPS 127-1 ?

Is the following description part of FIPS 127-1 ?

From the Rdb SPD, certain parts of the table have been filled.
Help needed on the rest of the table, to come up with a suitable response.
We have very little time to respond to the RFI,
need all the help one can get.



							Req'd
							for
				Rdb	Oracle	AS/400	FIPS 127-1?
				---	------	------	-----------
Database-limits:

255 columns/table		yes	___	___    	___
1,000,000 rows/table	        _?_     ___     ___     ___
255 characters/columns          yes     ___     ___     ___
2,000 tables/database           yes     ___     ___     ___
32K characters/row              yes     ___     ___     ___
120 characters for "key"        yes     ___     ___     ___
32K tables/view                 _?_     ___     ___     ___
32K tables joined @ once        _?_     ___     ___     ___

Data-types:

Character			yes	___	___     ___
Decimal				no_  	___	___     ___
Packed decimal			no_	___	___ 	___ 
Floating point			yes	___	___	___
Null Value			yes	___	___     ___
Logical Data			no_	___	___     ___
T.RTitleUserPersonal
Name		DateLines
1165.1Join of how many tables?TRCOA::MCMULLENKen McMullenThu Jun 18 1992 15:168
    1,000,000 million rows per table - NO PROBLEM the real limitation for
    table cardinality is disk space.
    
    32 K tables in a join - who could construct/envision such a thing! The
    government must be smoking and not destroying the drugs they
    confiscate.
1165.2RE: FIPSNOVA::FEENANJay Feenan Rdb/xxx EngineeringThu Jun 18 1992 22:185
please read note 933 and its replies.

cheers,

jay
1165.3Clarification?BROKE::HIGGSSQL is a camel in disguiseFri Jun 19 1992 16:1841
                    <<< Note 1165.0 by SCAACT::ADISESHAN >>>

	I think you need to go back and clarify what is meant by some of
	these terms:


							Req'd
							for
				Rdb	Oracle	AS/400	FIPS 127-1?
				---	------	------	-----------

Data-types:

Character			yes	___	___     ___
Decimal				no_  	___	___     ___
Packed decimal			no_	___	___ 	___ 
Floating point			yes	___	___	___
Null Value			yes	___	___     ___
Logical Data			no_	___	___     ___

 
	1) Please have them define what is meant by 'data-types'.
	   For example, Rdb can define columns within a table as
	   being of type DECIMAL or NUMERIC (along with other types).
	   And the host variable types that correspond to these types
	   are also supported (in those languages that support them)
	   But the physical representation on disk is not in DECIMAL
	   or NUMERIC form.   So we could say that Rdb supports 
	   Decimal/Packed decimal, or not, depending on the customer's
	   definition.

	2) What is meant by Decimal, as opposed to Packed decimal?
	   It could be that they mean NUMERIC when they use the term
	   Decimal, and DECIMAL when they use the term Packed decimal,
	   but I don't know.

	3) What do they mean by logical data?   Boolean?   Enumerations?
	   Something else?


	Bryan
1165.4Comments...no direct answersCOOKIE::MELTONThe zen of character setsSat Jun 20 1992 00:3815
I'd like to echo Bryan's comments in .3; the table that you supplied in .0
is a bit ambiguous, so it's hard to know what they really want.

Also, the last 2 lines of the "Database-limits" section are absurd.
Nobody (and I mean "nobody"!) would really want 32,768 tables in a view or
in a join.  Perhaps this was a simple typographical error and "32" was
intended instead of "32K"?  The other values are reasonable.

As far as FIPS 127-1 goes, I no longer know where my copy is, but I'm
quite sure that Steve Horn must have a copy (or know where to access one).
After all, Rdb/VMS must have had one to ensure that their limits covered
those required by the FIPS.

Good luck,
   Jim