Home > Archive > Cobol > October 2006 > RDBMS as a trademark
You are viewing an archived Text-only version of the thread.
To view this thread in it's original format and/or if you want to reply to
this thread please [click here]
| Author |
RDBMS as a trademark
|
|
| Paul Robinson 2006-10-08, 3:55 am |
| According to the records of the U.S. Patent and Trademark Office, the
only mark using "rdbms" is one consisting of the words "MXInsight RDBMS
Data Loader" in connection with software used for weather information.
There is no mark using RDBMS for any other usage on file with the U.S.
Patent and Trademark Office.
| |
| Alistair 2006-10-09, 7:56 am |
|
Paul Robinson wrote:
> According to the records of the U.S. Patent and Trademark Office, the
> only mark using "rdbms" is one consisting of the words "MXInsight RDBMS
> Data Loader" in connection with software used for weather information.
>
> There is no mark using RDBMS for any other usage on file with the U.S.
> Patent and Trademark Office.
Sergei mentioned, in the original thread, that Codasyl used RDBMS in
the name. I know lots of other rdbms products use the acronym too, and
incorrectly. Is there an rdbms which meets all 12 rules for an rdbms?
| |
| Richard 2006-10-12, 6:55 pm |
|
Alistair wrote:
>
> Sergei mentioned, in the original thread, that Codasyl used RDBMS in
> the name.
No it never did, the CODASYL DBTG Data Model was a network database.
Oracle (not CADASYL) did use the name 'CODASYL DBMS' for one of its
products for VMS.
Note: CODASYL and RDBMS are acronyms and should be expressed in upper
case.
> I know lots of other rdbms products use the acronym too, and
> incorrectly.
In what way do they use this that you assert is 'incorrect' ? By
writing them in lower case parhaps ?
> Is there an rdbms which meets all 12 rules for an rdbms?
Probably not, but that is irrelevant because systems are designed to
achieve the user's needs not some theoretical but unobtainable
perfection.
| |
| Alistair 2006-10-12, 6:55 pm |
|
Richard wrote:
> Alistair wrote:
>
>
> In what way do they use this that you assert is 'incorrect' ? By
> writing them in lower case parhaps ?
No, by not satisfying the rules for relational (I note that in other
places some people refer to truly-relational). Either a product is
relational or it is not. If you can satisy 10 out of 12 requirements
then you can not claim to be relational. Quasi-relational or
pseudo-relational perhaps.
I know of one product which claimed to be pseudo-relational but was in
fact inverted-list.
>
>
> Probably not, but that is irrelevant because systems are designed to
> achieve the user's needs not some theoretical but unobtainable
> perfection.
Why add the rider? Either there is a product which is relational or
there is not.
| |
| Richard 2006-10-12, 6:55 pm |
|
Alistair wrote:
>
> No, by not satisfying the rules for relational (I note that in other
> places some people refer to truly-relational). Either a product is
> relational or it is not. If you can satisy 10 out of 12 requirements
> then you can not claim to be relational. Quasi-relational or
> pseudo-relational perhaps.
"Codd's 12 rules" are 13 rules that defined an _ideal_ relational
database system. Most users want a _practical_ database and use the
'12 rules' as a guideline only.
The actual difference between types of database systems are the way
that they represent links. A hierarchical database may simply have the
data physically ordered in a particular way. A network database (eg
IDMS) uses hard links between related records, often these are physical
addresses. So for example, an order header has a pointer to the first
order line record which points to the next. The order lines need not
have the order number in the record, it may merely be a pointer to the
address of the header.
A relational database is one which each record has the keys of the data
item which it wants to relate to.
With a hierarchical database it is necessary to start reading at the
top of the hierarchy and navigate down through it to reach the data you
require.
With a network database there are certain points at which the data can
start to be read (eg an order header) and then it is necessary to
navigate from there to the data required.
A relational database caters for access to any individual piece of data
directly (rule 2).
One that does follow Codd's rules may well be referred to as 'Codd
compliant' or somesuch but it can still be a Relational database
without being Codd-Relational.
> I know of one product which claimed to be pseudo-relational but was in
> fact inverted-list.
Which product is that and where can I find such a claim ?
| |
| Alistair 2006-10-12, 6:55 pm |
|
Richard wrote:
> Alistair wrote:
>
>
> "Codd's 12 rules" are 13 rules that defined an _ideal_ relational
> database system. Most users want a _practical_ database and use the
> '12 rules' as a guideline only.
>
> The actual difference between types of database systems are the way
> that they represent links. A hierarchical database may simply have the
> data physically ordered in a particular way. A network database (eg
> IDMS) uses hard links between related records, often these are physical
> addresses. So for example, an order header has a pointer to the first
> order line record which points to the next. The order lines need not
> have the order number in the record, it may merely be a pointer to the
> address of the header.
>
> A relational database is one which each record has the keys of the data
> item which it wants to relate to.
>
> With a hierarchical database it is necessary to start reading at the
> top of the hierarchy and navigate down through it to reach the data you
> require.
>
> With a network database there are certain points at which the data can
> start to be read (eg an order header) and then it is necessary to
> navigate from there to the data required.
>
> A relational database caters for access to any individual piece of data
> directly (rule 2).
>
> One that does follow Codd's rules may well be referred to as 'Codd
> compliant' or somesuch but it can still be a Relational database
> without being Codd-Relational.
>
>
> Which product is that and where can I find such a claim ?
Adabas. I saw the reference once, many years ago. I have not seen it
repeated. I think the claim had been made in training material (and
probably not Software AG's training material).
| |
| Richard 2006-10-12, 6:55 pm |
|
Alistair wrote:
>
> Adabas. I saw the reference once, many years ago. I have not seen it
> repeated. I think the claim had been made in training material (and
> probably not Software AG's training material).
You are such a mine of misinformation.
There was an original Adabas which was an inverted list system. In the
80s Software AG purchased from Nixdorf a completely different
relational database system which it named Adabas D. This is now
available as SAPDB, MaxDB and from Sun as Adabas D.
You are  . In the 70s and early 80s Adabas was an inverted file
systems. Adabas D was not an inverted file system but was a RDBMS.
| |
| William M. Klein 2006-10-12, 6:55 pm |
| Richard,
I checked the SoftwareAG site, and you are certainly correct that "ADABAS D"
is their current product, but I also seem to remember that when they FIRST came
out with their RDBMS system that they used the acronym in their name. Google
finds quite a few uses of that, but none from 'SoftwareAG" itself. So I, too,
may just have a faulty memory - or MAYBE they used that name briefly but found
they could NOT trademark it - so they changed to the new name.
--
Bill Klein
wmklein <at> ix.netcom.com
"Richard" <riplin@Azonic.co.nz> wrote in message
news:1160589745.185598.304320@b28g2000cwb.googlegroups.com...
>
> Alistair wrote:
>
>
> You are such a mine of misinformation.
>
> There was an original Adabas which was an inverted list system. In the
> 80s Software AG purchased from Nixdorf a completely different
> relational database system which it named Adabas D. This is now
> available as SAPDB, MaxDB and from Sun as Adabas D.
>
> You are . In the 70s and early 80s Adabas was an inverted file
> systems. Adabas D was not an inverted file system but was a RDBMS.
>
| |
| Alistair 2006-10-12, 6:55 pm |
|
Richard wrote:
> Alistair wrote:
>
>
> You are such a mine of misinformation.
>
> There was an original Adabas which was an inverted list system. In the
> 80s Software AG purchased from Nixdorf a completely different
> relational database system which it named Adabas D. This is now
> available as SAPDB, MaxDB and from Sun as Adabas D.
>
> You are . In the 70s and early 80s Adabas was an inverted file
> systems. Adabas D was not an inverted file system but was a RDBMS.
Interesting. Perhaps the inverted list database product on my pc
masquerading as Adabas 3.3 with Natural 6 as the programming product is
not the inverted list product that I think? The documentation certainly
refers to the ASSOCIATOR which is used in the inverted list product and
which I have no reason to believe would appear in a relational database.
| |
| Richard 2006-10-12, 6:55 pm |
|
William M. Klein wrote:
> Richard,
> I checked the SoftwareAG site, and you are certainly correct that "ADABAS D"
> is their current product,
Adabas D was sold, or licenced, to SAP and is now available as SAPDB
and as MaxDB from MySQL. Their current product 'Adabas 2006' is the
legacy' inverted file system that they _do_not_ claim to be relational.
Adabas D, which they also sell, they do claim to "fully implements the
relational model".
> but I also seem to remember that when they FIRST came
> out with their RDBMS system that they used the acronym in their name.
It is quite common to use a product type as part of the name, for
Example "Empress RDBMS" to distinguish it from "Empress BASIC" (if
there was one) but only the "Empress" is a registered trademark.
> Google
> finds quite a few uses of that, but none from 'SoftwareAG" itself. So I, too,
> may just have a faulty memory - or MAYBE they used that name briefly but found
> they could NOT trademark it - so they changed to the new name.
| |
| Richard 2006-10-12, 6:55 pm |
|
Alistair wrote:
So it was not, as you claimed, the _product_ which was said to be
sortofrelational, but someone else entirely which said that.
[color=darkred]
[color=darkred]
> Interesting. Perhaps the inverted list database product on my pc
> masquerading as Adabas 3.3 with Natural 6 as the programming product is
> not the inverted list product that I think?
Do try and follow this. The original 'Adabas' was/is an inverted file
system and your 3.3 may well be that, as is 'Adabas 2006'. But is is
not the _only_ 'Adabas' because Adabas D was a completely different
system that was/is an RDBMS and is now also available as SAPDB and
MaxDB.
> The documentation certainly
> refers to the ASSOCIATOR which is used in the inverted list product and
> which I have no reason to believe would appear in a relational database.
So a claim that 'Adabas' is an RDBMS may well be true if it is Adabas D
they are referring to. As you don't seem to know that there are two
completely different products, even though I had told you that was the
case, you were probably about what they referred to.
But in any case saying that a database can act like a relational one,
or is pseudo-relational, is not unreasonable.
| |
| Alistair 2006-10-12, 6:55 pm |
|
Richard wrote:
> Alistair wrote:
>
>
> So it was not, as you claimed, the _product_ which was said to be
> sortofrelational, but someone else entirely which said that.
Seeing as you are getting pedantic perhaps you should re-read what I
wrote:
"I think the claim had been made in training material (and probably
not Software AG's training material)." This does not mean that the
claim had not been made in SAG training material but it does allow for
the possibility that the claim had been made on behalf of some other
organisation (probably Grafton) in referring to SAG's product. I am
aware of how porous my memory is and have been generous in allowing you
to poo-poo it so far.
>
>
>
> Do try and follow this. The original 'Adabas' was/is an inverted file
> system and your 3.3 may well be that, as is 'Adabas 2006'. But is is
> not the _only_ 'Adabas' because Adabas D was a completely different
> system that was/is an RDBMS and is now also available as SAPDB and
> MaxDB.
So I am wrong in claiming that Adabas 3.3 is inverted-list because you
have made the claim that there are Adabas products (C AND D) which may
not be inverted-list? I am aware of different versions of Adabas dating
back to the mid-80s. I am also aware that Adabas D (of which I have a
copy) refers to tables and using SQL but that doesn't make it a
relational database.
>
>
> So a claim that 'Adabas' is an RDBMS may well be true if it is Adabas D
> they are referring to. As you don't seem to know that there are two
> completely different products, even though I had told you that was the
> case, you were probably about what they referred to.
No confusion. I do know about different Adabas databases as the
reference to C proves.
>
> But in any case saying that a database can act like a relational one,
> or is pseudo-relational, is not unreasonable.
If it looks like a fish, smells like a fish and sounds like a fish
(some fish, such as herrings, do make sounds) then is it a fish? No.
| |
| Richard 2006-10-12, 6:55 pm |
|
Alistair wrote:
>
> So I am wrong in claiming that Adabas 3.3 is inverted-list because you
> have made the claim that there are Adabas products (C AND D) which may
> not be inverted-list?
How could you possibly read "The original 'Adabas' was/is an inverted
file system and your 3.3 may well be that," as meaning that it was not
that ?
The point about Adabas D is that there is no connection at all to
Adabas the inverted file system except the name and they were both
owned by SAG. It is not a version of Adabas, it is completely
different code written by different people (U of Berlin and Nixdorf)
originally with different requirements.
'Adabas 3.3' _is_ Adabas C, the legacy Adabas which is now Adabas 2006.
Your grouping of "(C AND D)" seems to indicate that you don't know that
they are completely different products, one of which _is_ inverted list
(C) and the other is not (D).
But it is not me that makes the claim that Adabas D is not an inverted
file system, it is MySQL, SAG and others:
"""MaxDB is an ANSI SQL-92 (entry level) compliant relational database
management system (RDBMS) from SAP AG, """ (MaxDB is another name for
Adabas D, see
http://dev.mysql.com/doc/refman/5.0...b-overview.html).
"""Adabas D provides you with a database system that fully implements
the relational model - including support for domains, primary keys,
updatable join views, referential integrity, triggers and database
procedures."""
> I am aware of different versions of Adabas dating
> back to the mid-80s. I am also aware that Adabas D (of which I have a
> copy) refers to tables and using SQL but that doesn't make it a
> relational database.
I suggest that you take up that issue with SAG, MySQL and others who
may be able to tell you what does make it a relational database.
| |
| Richard 2006-10-12, 6:55 pm |
|
Alistair wrote:
> Seeing as you are getting pedantic perhaps you should re-read what I
> wrote:
> "I think the claim had been made in training material (and probably
> not Software AG's training material)." This does not mean that the
> claim had not been made in SAG training material but it does allow for
> the possibility that the claim had been made on behalf of some other
> organisation (probably Grafton) in referring to SAG's product. I am
> aware of how porous my memory is and have been generous in allowing you
> to poo-poo it so far.
I know what you wrote. What I question is you interpretation of what
was said.
"in referring to SAG's product."
SAG has several products one of which is an inverted file system,
another completely different product is a relational database
management system (RDBMS).
Both have Adabas in the name.
It may well be that the training material was for Adabas D, the RDBMS,
in which case it is completely reasonable to make claims about it being
relational in some way.
As you do not appear to understand that they are different products I
can see where your confusion comes from.
| |
| Kelly Bert Manning 2006-10-30, 6:55 pm |
| "Alistair" (alistair@ld50macca.demon.co.uk) writes:
>
> I know of one product which claimed to be pseudo-relational but was in
> fact inverted-list.
I remember a late 1970s Honeywell IDS intructor trying to convince us that
IDS could support Relations. Someone in the course mentioned that C. J.
Date cited IDS as a Network/DBTG/CODASYL DB in "An Introduction to Database
Systems", at least in the 2nd edition.
|
|
|
|
|