un-triviality is saved - Mailing list pgsql-sql

From Dmitry Turin
Subject un-triviality is saved
Date
Msg-id 17323140312.20070412142320@narod.ru
Whole thread Raw
List pgsql-sql
>What does this do that inheritance doesn't already do? I don't think I see anything.

- So, we have proved, that all, what you offer to make on your a-a-a... Fortran, is possible to make on assembler.
- First, I know this without you, second, speech was not about that.

Newness is not "what", newness is "how" !

As to processing data,
new way allow to think in natural terms of tree, instead in artificial
terms of Decart's production and PL-extension for SQL.

As to communication,
new way allow to extract XML from database directly, without any
gasket (php, etc) - that is VERY important to untrained, inexperienced user.
Direct way open gates (doors) for applied specialists !

>> We speak about database, thus what user wants to do is
>> predetermined partly - there are at least two things: database and
>> visualizer (browser is only one case).

>This sounds to me entirely like a user-interface problem.

Situation is system.
And as any item, it has system's characteristics.
What's about difficulty:
difficulty-of-request + difficulty-of-transportation = invariant.
If you make request without transport's possibility,then you must to enter one more language!It's very, very bad for
un-experienceduser.
 

And the main: un-triviality (un-banality) is saved !!!
you are forced to get this un-triviality in other language.
But in addition to un-triviality, you have place of dockbetween SQL and other language.

Representation of data as tree (instead of as matrix) is not aim itself,
it's sideline-result.

---

>We don't need to make SQL more complicated (and less relational)

There are two question: "more complicated" and "less relational".

First:
SQL doesn't loss and doesn't gain any functional characteristics.
SQL does not become more complicated.

Second:
SQL doesn't loss and doesn't gain any relational characteristics, they are saved.
SQL does not become less relational.

>point of SQL was to move away from the hierarchical model

"Was", pay attention.
It doesn't mean, that it should be always.
Let's distinguis between two characteristics: harmness and orthodoxality.
Orthodoxality is not argument, i think.

AS> inexperienced user are willing to learn

What about user want to learn, user think:
what is the least, so is the best.
Minimum quantity of tools is the best.
This is fact instead of opinion.




pgsql-sql by date:

Previous
From: "Rodrigo De León"
Date:
Subject: Re: Replace string
Next
From:
Date:
Subject: ERROR: column "crc" does not exist