Binary Relational Modeling call

[Ivan]

about brm: a simple, yet powerful rules and cool graphs for database definition. Join as at:
https://groups.google.com/forum/#!forum/binary-relational-model

[Bob Badour]

Ivan wrote:

> about brm: a simple, yet powerful rules and cool graphs for database definition. Join as at:

How does this supposedly new model deal with data that are naturally
n-ary relations for n > 2 ?

[Eric]

On 2011-09-07, Ivan <ivanv...@gmail.com> wrote:
> about brm: a simple, yet powerful rules and
> cool graphs for database definition. Join as at:

> <link removed>

Are you actually claiming to have invented something? I can't see it.
And in general we are not impressed by "cool graphs". If we were
actually getting something useful, but I can't see that either as yet.

Eric

--
ms fnd in a lbry

[Ivan]

i hate crossing. you can discuss the materia by visiting the link.

[Eric]

On 2011-09-09, Ivan <ivanv...@gmail.com> wrote:
> i hate crossing. you can discuss the materia by visiting the link.

Probably not, after that remark. If you don't get talked about outside
a little world of your own creation, you don't get talked about. Your
world has obviously not yet provided a sufficient level of interest to
entice anyone else in.

In any case, what are you aiming at here? An abstract theory, a design
methodology, a diagramming convention, a piece of software? At the moment
I can't tell.

[Ivan]

all right, i suppose moderator of the group won't be too mad at us, that's all i was having in mind.

i'm honored to have something to share with anyone interested.
i sincerely fill that BRM could aim at SQL's place in programming.

yes, it can manage n-ary relationships.

is it new technology? i found some resources on internet claiming that BRM exist since the year 1984. what is new is its graphing capabilities. to give you a sneak peek, very BRM can be also implemented in itself.

i hope i at least interested you. if i'm not, i'd like to thank you for your time. that's the way to make a progress.

[Bob Badour]

Ivan wrote:

> i hate crossing. you can discuss the materia by visiting the link.

In other words, you are just spamming us to get people to come to a
pointless group you created.

[Ivan]

@Bob

do you really think no one could be interested?

[Bob Badour]

Ivan wrote:

> @Bob
>
> do you really think no one could be interested?

I don't really think any educated person could be interested in
something that cannot easily, simply, and directly answer the question:

[Eric]

On 2011-09-09, Ivan <ivanv...@gmail.com> wrote:

> all right, i suppose moderator of the group won't be too mad at us,
> that's all i was having in mind.

**digression**
This is a Usenet group.
There is no moderator.
There is no owner.
It does not reside on any single host.
There is no reliable way to delete your own or anyone else's messages.
Messages may or may not be retained somewhere public essentially for
ever.
**end of digression**

> i'm honored to have something to share with anyone interested.
> i sincerely fill that BRM could aim at SQL's place in programming.

I would have thought that that was an apples/oranges comparison. Unless
you are thinking about diagram-driven software, but that is an
interesting concept not noted for its success except in very specific
areas.

> yes, it can manage n-ary relationships.
>
> is it new technology? i found some resources on internet claiming that BRM
> exist since the year 1984. what is new is its graphing capabilities. to
> give you a sneak peek, very BRM can be also implemented in itself.

I found a 1982 paper which referred to a 1980 paper (it didn't take me
long, or I wouldn't have bothered), as well as at least one quite recent
paper (2009). I haven't read any more than the synopses, but they all
seem to come from the large collection of people who do not properly
understand the relational model.

I'm afraid that, on the evidence so far, I have to put you in that
category too.

http://www.dbdebunk.com/page/page/1147347.htm

> i hope i at least interested you. if i'm not, i'd like to thank you for
> your time. that's the way to make a progress.

Just for the interest of others, I will point out that you have had
a very similar conversation with someone on comp.databases , and that
you have this stuff on more than one website. So if you have a purpose
other than getting yourself noticed and generating website traffic,
the sooner it becomes more apparent the better.

[Ivan]

https://docs.google.com/drawings/d/1xYyqPxmlzpG8_HHU-qfiWbDjuBkA0zbJ8btkwPn6EiE/edit?hl=hr

[Ivan]

Dana subota, 10. rujna 2011. 13:55:25 UTC+2, korisnik Eric napisao je:
> On 2011-09-09, Ivan <ivanv...@gmail.com> wrote:
> > all right, i suppose moderator of the group won't be too mad at us,
> > that's all i was having in mind.
>
> **digression**
> This is a Usenet group.
> There is no moderator.
> There is no owner.
> It does not reside on any single host.
> There is no reliable way to delete your own or anyone else's messages.
> Messages may or may not be retained somewhere public essentially for
> ever.
> **end of digression**

we're supposed to be moderators of ourselves. don't you think our posts should be given to anyone interested? or you have to hide something? well, i have solution for these two problems: NO CROSSING, if you mind!

> > i'm honored to have something to share with anyone interested.
> > i sincerely fill that BRM could aim at SQL's place in programming.
>
> I would have thought that that was an apples/oranges comparison. Unless
> you are thinking about diagram-driven software, but that is an
> interesting concept not noted for its success except in very specific
> areas.

for now, i didn't see visual representation of BRM. as i thought, BRM existed before, but graphs are new. and if i'm allowed to have my opinion that's a very cool news. just to invite you to our little group, let me tease you a little bit:
https://docs.google.com/drawings/d/1sa2AyprqIFvy2de3BoKse-4vfh2_sJq6DiRe4ZiXgu8/edit?hl=hr

what happens when things get ruff? let's see definition of the BRM in SQL visualization compared to the BRM in very itself:
https://docs.google.com/drawings/d/1_jY_wLwqzuGMuff1-nOK8fgqcTY4GUQ2iENKfYEVC4k/edit?hl=hr

> > yes, it can manage n-ary relationships.
> >
> > is it new technology? i found some resources on internet claiming that BRM
> > exist since the year 1984. what is new is its graphing capabilities. to
> > give you a sneak peek, very BRM can be also implemented in itself.
>
> I found a 1982 paper which referred to a 1980 paper (it didn't take me
> long, or I wouldn't have bothered), as well as at least one quite recent
> paper (2009). I haven't read any more than the synopses, but they all
> seem to come from the large collection of people who do not properly
> understand the relational model.
>
> I'm afraid that, on the evidence so far, I have to put you in that
> category too.
>
> http://www.dbdebunk.com/page/page/1147347.htm
>
> > i hope i at least interested you. if i'm not, i'd like to thank you for
> > your time. that's the way to make a progress.
>
> Just for the interest of others, I will point out that you have had
> a very similar conversation with someone on comp.databases , and that
> you have this stuff on more than one website. So if you have a purpose
> other than getting yourself noticed and generating website traffic,
> the sooner it becomes more apparent the better.
>
> Eric
>
> --
> ms fnd in a lbry

as for those conversations, why didn't they take a place, that's up to you, folks. if anyone interested, "the code project site" (8 000 000 programmer members) actually this morning published an article i wrote on this subject:
http://www.codeproject.com/KB/database/brm001.aspx

if you can promise me that you really took this concept under consideration, and if you find it offensive for programmers community, i'll be glad to wipe up the group i'm trying to give a breath in. for xrist's sake, this is not about money and this is not a visit count pursuit. i actually think that we have something new we could use in database programming. and a friend doesn't deserve such humiliation i'm going through.

but again, maybe i'm living in imaginary clouds. so explain me it's so.

[Ivan]

guys, i'm sincerely sorry i lose temper back there. You certainly didn't deserve that kind of treatment from me. i think You were just trying to protect each other from my nonsense.

just, it took me so long to reach this stage of knowledge (in whatever infancy state it is). particularly, 15 years of thinking and thinking and thinking. you can imagine under what pressure i am for reaching so little in so much time.

Since I've got some experience about databases and SQL, one assumption constantly twinkled in my head: there has to be a better way for defining databases. I was imagining some cool and neat model for representing tables, records and relations that will simply "be the one". Since then I have experimented with different exotic visual representations of data that was crossing my mind. And then I saw it! It was literally love at first sight. The same power of SQL's database defining, but with much simpler rules. So simple that I decided to give it a name. The name was Binary Relational Modeling.

After setting up a dedicated site, I entered the same name in Google and realized that similar (if not the same) approaches already exist, even under the same name. After initial shock, when I gathered myself in, I found myself surprised why a wider use of the technology isn't already in a full swing by now. So I decided to share my experiences with programmers interested in alternative database technologies.

i hope You will forgive me for loosing temper, i was under great pressure. i'll try to be a good one from now on, i promise :x

[Bob Badour]

Ivan wrote:

Sorry, I don't have access to that. Actually, when I clicked on the link
from my newsreader it was just a bunch of garbled characters in pretty
pastel colors.

[Bob Badour]

Ivan wrote:

For a conversation to happen, you would have to say something. Anything
more substantive than: "Hey! Come on over to our new group! There'll be
ponies and ice cream and ..."

[Ivan]

this is another link to n-ary. this time is bitmap:
https://sites.google.com/site/binaryrelationalmodeling/home/n-ary%281%29.png

yes, my fault, i'm aware now i didn't give you a reason to distinct me among spammers (if there exist any spammer in the world, as they are people also).

i'm really not interested in that it has to be "my" group. my interests lay in proposal of actual development of application that implements the concept. as i don't have a time for development, all i thought is that someone else might be interested in programming. and i don't have any interest either in earning from, or in being referenced in end user app. totally royalty free using. after all, the concept isn't all my idea.

all of you, and i mean it, all of you are invited to ask any question about any of your interest, as i feel capable of answering matter related to programming, either in private or in public (remember, i'd like to see actual app running).

once again, my apologies for promising much, as now i see that i don't really have so much. anyway, if you have an interest, you may count on me, totally royalty free.

also, please note that i am not a vampire.
thank you in advance
bip bip :}

[Bob Badour]

Teacher looks an awful lot like a tenary relation not a binary relation.

The arrows seem to indicate one can easily access the Text for teacher,
for example, from Teacher, but one cannot so easily access the Teacher
for being taught, for example, from Text. Is that correct?

[Ivan]

teacher is a set (rounded rect). it's related to three other sets by named binary relations (arrows). one can track binary relation from some set in direction of arrow and backwards.

[Ivan]

if someone is subjective enough to give it a shot, i recommend reading of:
https://sites.google.com/site/binaryrelationalmodeling/

[Bob Badour]

Ivan wrote:

> teacher is a set (rounded rect). it's related to three other sets by named binary relations (arrows). one can track binary relation from some set in direction of arrow and backwards.

If Teacher is a set, what's it a set of?

So, your data model has at least two structural elements: sets and
binary relations. Does it have any other structural elements?

Note that at 2 structural elements, BRM is already structurally more
complicated than the relational model, which has only relations. What
compensating benefit does one get for the added complexity?

[Ivan]

> If Teacher is a set, what's it a set of?

elements are contents of sets, but they are not shown by BRM graph. they should be available for browsing only in user application.

binary relations have their name, source set and destination set. they state possible bindings between elements of source and destination sets. Basically, a set can be seen as a table with its fields represented by relations that source from that set. elements are records of that table. here is example:
https://sites.google.com/site/binaryrelationalmodeling/compare-invoice%286%29.png

> So, your data model has at least two structural elements: sets and
> binary relations. Does it have any other structural elements?

those 2 are crutial structural components of BRM. however, notation of uniqueness of elements is welcome to take place in graphs (one-to-one relationship is done in the same way many-to-may relationship is, with intermediate set, but with uniqueness constraint). although i'm aware of importance of uniqueness, uniqueness notation is not specified and programmers have to implement one in the way they find appropriate.

> Note that at 2 structural elements, BRM is already structurally more
> complicated than the relational model, which has only relations. What
> compensating benefit does one get for the added complexity?

the real value of BRM graphs is in its simpleness. its simpleness takes place in complicated definitions. compare BRM definition in SQL to BRM definition in itself:
https://sites.google.com/site/binaryrelationalmodeling/compare-brm.png

[Cimode]

<< elements are contents of sets, but they are not shown by BRM graph.
they should be available for browsing only in user application.>>

Are you defining elements of a set through a GUI feature ? That must
mean that sets can not be defined if browsing does not exist.

<<binary relations have their name, source set and destination set.
they state possible bindings between elements of source and
destination sets. Basically, a set can be seen as a table with its
fields represented by relations that source from that set. elements
are records of that table. here is example:https://sites.google.com/

site/binaryrelationalmodeling/compare-invoic...>>
What are you talking about ?

<<those 2 are crutial structural components of BRM. however, notation
of uniqueness of elements is welcome to take place in graphs (one-to-
one relationship is done in the same way many-to-may relationship is,
with intermediate set, but with uniqueness constraint). although i'm
aware of importance of uniqueness, uniqueness notation is not
specified and programmers have to implement one in the way they find
appropriate.>>

Subjective need for uniqueness. I wonder what would happen in a
family, if one of the two parents decides to name all or some of the
children with the same name.

<<
> the real value of BRM graphs is in its simpleness. its simpleness takes place in complicated definitions. compare BRM definition in SQL to BRM definition in itself:https://sites.google.com/site/binaryrelationalmodeling/compare-brm.png
>>

BRM or whatever you call it is neither simple nor clear. It is an
obscure and fuzzy creation of your imagination. Give it up: you are
wasting your time. Educate yourself then create a model.

[Bob Badour]

Ivan wrote:

>>If Teacher is a set, what's it a set of?
>
> elements are contents of sets, but they are not shown by BRM graph. they should be available for browsing only in user application.
>
> binary relations have their name, source set and destination set. they state possible bindings between elements of source and destination sets. Basically, a set can be seen as a table with its fields represented by relations that source from that set. elements are records of that table. here is example:
> https://sites.google.com/site/binaryrelationalmodeling/compare-invoice%286%29.png
>
>
>>So, your data model has at least two structural elements: sets and
>>binary relations. Does it have any other structural elements?
>
> those 2 are crutial structural components of BRM. however, notation of uniqueness of elements is welcome to take place in graphs (one-to-one relationship is done in the same way many-to-may relationship is, with intermediate set, but with uniqueness constraint). although i'm aware of importance of uniqueness, uniqueness notation is not specified and programmers have to implement one in the way they find appropriate.

So, you are saying BRM doesn't have logical identity?

>>Note that at 2 structural elements, BRM is already structurally more
>>complicated than the relational model, which has only relations. What
>>compensating benefit does one get for the added complexity?
>
> the real value of BRM graphs is in its simpleness.

Whoa! Hold on there! We have already established BRM is structurally
more complex than the relational model not simpler. It's real value
cannot come from something it doesn't even have.

[Bob Badour]

Cimode wrote:

> <<those 2 are crutial structural components of BRM. however, notation
> of uniqueness of elements is welcome to take place in graphs (one-to-
> one relationship is done in the same way many-to-may relationship is,
> with intermediate set, but with uniqueness constraint). although i'm
> aware of importance of uniqueness, uniqueness notation is not
> specified and programmers have to implement one in the way they find
> appropriate.>>
> Subjective need for uniqueness. I wonder what would happen in a
> family, if one of the two parents decides to name all or some of the
> children with the same name.

Bad choice. See George Foreman.

[Ivan]

> Ivan wrote:
>
> >>If Teacher is a set, what's it a set of?
> >
> > elements are contents of sets, but they are not shown by BRM graph. they should be available for browsing only in user application.
> >
> > binary relations have their name, source set and destination set. they state possible bindings between elements of source and destination sets. Basically, a set can be seen as a table with its fields represented by relations that source from that set. elements are records of that table. here is example:
> > https://sites.google.com/site/binaryrelationalmodeling/compare-invoice%286%29.png
> >
> >
> >>So, your data model has at least two structural elements: sets and
> >>binary relations. Does it have any other structural elements?
> >
> > those 2 are crutial structural components of BRM. however, notation of uniqueness of elements is welcome to take place in graphs (one-to-one relationship is done in the same way many-to-may relationship is, with intermediate set, but with uniqueness constraint). although i'm aware of importance of uniqueness, uniqueness notation is not specified and programmers have to implement one in the way they find appropriate.
>
> So, you are saying BRM doesn't have logical identity?

logical identity? please explain.

> >>Note that at 2 structural elements, BRM is already structurally more
> >>complicated than the relational model, which has only relations. What
> >>compensating benefit does one get for the added complexity?
> >
> > the real value of BRM graphs is in its simpleness.
>
> Whoa! Hold on there! We have already established BRM is structurally
> more complex than the relational model not simpler. It's real value
> cannot come from something it doesn't even have.

well, if you really want my opinion, sql has:
1. tables,
2. fields,
3. types of fields.

relations are just query language utilities that allow combining tables.

brm has these equivalents in the same order:
1. sets,
2. relations,
3. destination sets.

what i find interesting in brm is that (types of fields)/(destination sets) and tables/sets are noted in the same place by the same rectangle. meaning, when you specify (type of field)/(destination set) you can immediately from it draw relations to other sets. also, note the absence of ID codes (major thing, if you ask me)

[Ivan]

here is another teaser:
https://sites.google.com/site/binaryrelationalmodeling/brm.swf

[Eric]

On 2011-09-14, Ivan <ivanv...@gmail.com> wrote:
> here is another teaser:
> https://sites.google.com/site/binaryrelationalmodeling/brm.swf

Now you seem to be re-inventing Mumps with a graphic interface. Not
relational at all then, which fits in because you don't seem to know
what a relation is.

[Ivan]

"...Relations used in BRM are actually many-to-one relationships. Besides their regular use, they can be used to express definitions of tables and their columns. It also turns out that many-to-many and one-to-one relationships can be represented using BRM relations. Many-to-many relationship can be expressed using a complementary pair of BRM relations (experienced database users will be familiar with this technique). One-to-one relationships can also be expressed in similar way like many-to-many relationships, but with some unique constraints..."

[Eric]

If that's a quote from somewhere you should say where.

That quote, and many of your previous posts here, assume that relation
and relationship are the same thing, and in the Relational Model they
are not. Many people assume that the "Relational" in the Relational
Model refers to relationships between tables, but it refers to a
Relation, defined as a set of tuples[1], which in fact corresponds to a
table. Until you sort that out for yourself you will continue to fail to
make sense.

Eric

[1] with various restrictions. Look it up!

[Bob Badour]

Eric wrote:

> On 2011-09-15, Ivan <ivanv...@gmail.com> wrote:
>
>>"...Relations used in BRM are actually many-to-one relationships. Besides
>>their regular use, they can be used to express definitions of tables
>>and their columns. It also turns out that many-to-many and one-to-one
>>relationships can be represented using BRM relations. Many-to-many
>>relationship can be expressed using a complementary pair of BRM
>>relations (experienced database users will be familiar with this
>>technique). One-to-one relationships can also be expressed in similar
>>way like many-to-many relationships, but with some unique constraints..."
>
> If that's a quote from somewhere you should say where.
>
> That quote, and many of your previous posts here, assume that relation
> and relationship are the same thing, and in the Relational Model they
> are not. Many people assume that the "Relational" in the Relational
> Model refers to relationships between tables, but it refers to a

It's probably best to avoid even the word "relationships". Use
"referential integrity" instead.

e.g. It's a common misconception--heavily promoted by some vendors who
wish to keep their customers ignorant and "locked in"--that "relation"
in the name "relational model" somehow has something solely to do with
referential integrity.

> Relation, defined as a set of tuples[1], which in fact corresponds to a
> table. Until you sort that out for yourself you will continue to fail to
> make sense.
>
> Eric
>
> [1] with various restrictions. Look it up!

Actually, relation is a mathematical concept with algebraic meaning.
Codd's 1972 proved equivalent expressiveness between the algebra and
predicate calculus.

When one really thinks about it, the use of tuples is almost as
notational as the use of tables.

But certainly, while one can easily describe referential integrity
constraints using relations and either the algebra or the calculus,
relations themselves have little to do with referential integrity among
themselves.

[Ivan]

wiki:
"A relation is a data structure which consists of a heading and an unordered set of tuples which share the same type."

wiki:
"In mathematics, a binary relation on a set A is a collection of ordered pairs of elements of A. In other words, it is a subset of the Cartesian product A2 = A × A. More generally, a binary relation between two sets A and B is a subset of A × B."

says nothing about difference between "relation" and "relationships".
binary relations? exactly what i have in mind: every binary relation can be implemented as table of pairs (2 field tuples) of references to sets' elements.

[Bob Badour]

What does the ever-so-omniscient wiki way about n-ary relations?

[Erwin]

And why "references to sets' elements" instead of just "sets'
elements" ?????????

[Tegiri Nenashi]

Binary relations have rich history, arguably a theory of Binary
Relations is more elaborate than that of n-ary Relations. The fact
that there is no killer application for Binary relations is
unfortunate, but it is not for the lack of trying. If you are
genuinely interested learning the subject something like RelView
(http://www.informatik.uni-kiel.de/~progsys/relview/) would be a good
start.