info
Google Groups no longer supports new Usenet posts or subscriptions. Historical content remains viewable.
Dismiss
An anecdote about Codd on the "back side" of Wikipedia
91 views
Skip to first unread message
Nicola
Mar 18, 2015, 4:41:42 PM3/18/15
to
I have stumbled upon the Talk page about Codd in Wikipedia:
https://en.wikipedia.org/wiki/Talk:Edgar_F._Codd
where a user has posted the following story:
«Starting in 1967, Luther J. Woodrum implemented a searchable patent
database at IBM using the APL language. The database was arranged
similar to a large No-SQL table with classes and sub-classes associated
with the patent number. The classes and sub-classes represented the
different searchable fields (columns), eg Inventor, Assignee, Filing
Date, Claims, etc for any patent. The database held several million
world-wide patents that were relevant to IBM's business interests. The
API also had a Boolean based query capability which Luther described as
n-tuple relational operators. Luther presented this as an example of a
"database" system in 1970 at his Poughkeepsie IBM Education Center
class. Ted Codd was a student in this class.»
This anecdote seems to imply that Codd may have been influenced by the
work of L.J. Woodrum (of whom I've found only a couple of references in
DBLP and a LinkedIn profile). Of course, in 1970 Codd had already
written about the relational model (maybe the poster misremembering?),
and applications of relational formalisms to data were not new (see the
work by Levein and Maron, cited by Codd itself) but if the above were
true, I think it would be interesting from a historical point of view,
and it would be interesting to know more about that system at IBM.
Nicola
--- news://freenews.netfront.net/ - complaints: ne...@netfront.net ---
https://en.wikipedia.org/wiki/Talk:Edgar_F._Codd
where a user has posted the following story:
«Starting in 1967, Luther J. Woodrum implemented a searchable patent
database at IBM using the APL language. The database was arranged
similar to a large No-SQL table with classes and sub-classes associated
with the patent number. The classes and sub-classes represented the
different searchable fields (columns), eg Inventor, Assignee, Filing
Date, Claims, etc for any patent. The database held several million
world-wide patents that were relevant to IBM's business interests. The
API also had a Boolean based query capability which Luther described as
n-tuple relational operators. Luther presented this as an example of a
"database" system in 1970 at his Poughkeepsie IBM Education Center
class. Ted Codd was a student in this class.»
This anecdote seems to imply that Codd may have been influenced by the
work of L.J. Woodrum (of whom I've found only a couple of references in
DBLP and a LinkedIn profile). Of course, in 1970 Codd had already
written about the relational model (maybe the poster misremembering?),
and applications of relational formalisms to data were not new (see the
work by Levein and Maron, cited by Codd itself) but if the above were
true, I think it would be interesting from a historical point of view,
and it would be interesting to know more about that system at IBM.
Nicola
--- news://freenews.netfront.net/ - complaints: ne...@netfront.net ---
Ed Prochak
Jun 23, 2015, 2:16:16 PM6/23/15
to
Nicola
Jun 24, 2015, 4:00:16 PM6/24/15
to
I don't know if it is possible to contact him through LinkedIn to ask
for a comment.
In fact, I was hoping that someone would reply saying: "Ah, I wrote
that comment!" But now I am not positive about finding more details.
0 new messages