Re: theoretical question on the RDBMS

On Tue, 13 Aug 2002 16:31:38 GMT, paul_at_not.a.chance.ie (Paulie) wrote:

>"Martin Hungerford" <Martin.Hungerford_at_dsto.defence.gov.au> wrote:
>
>> > Access isn't an RDBMS.
>
>> Without being too rude, would you care to describe what MS-Access is, if it
>> is not an RDBMS (or at least a version of a SQL-DBMS)
>
><polite mode>
> Access is a flat file database.
? </polite mode> Do you mean when you open it in notepad? <polite mode> Please explain since I really don't get this.

>You could call it an SQL DBMS but it
> does not support transactions
? </polite mode> Just out of curiosity, what version are you talking about? <polite mode>

>, and is flakey under multi-user conditions
> (that's true for every DB, but with Access it starts with n very low)
Very true. One must be very careful with Access even at 20-ish concurrent users and there were serious bugs in DB engine 3.x and 4.x for almost whole year after releasing.
Also Access procedural language (VBA) is built as a modification of VB so more bugs were introduced. To top that object model is not very neet. Also it is failing SQL syntax standard compliance at basic level (different wildcard chars and a like).

Still, I would call it RDBMS (but certainly not a good, reliable and scalable RDBMS).

>, it won't run on Unix (true for MS SQL also, which is an RDBMS).
>
> Overall, it lacks features that a true RDBMS needs - however, you
> do get lots of other stuff thrown in (GUI designer, ease of use)
></polite mode>

What's your checklist for R in RDBMS compliance? Not that I am a fan of Access nor MySQL, but I don't like when people claim these don't support transactions, ref. integrity or similar just like that...

( GoranG79 AT hotmail.com ) Received on Wed Aug 14 2002 - 10:20:18 CEST