[ih] History of Online Databases

[John Day]

Dave,

I was just answering another email off list with Rick and remembered 
what might be the first occurrence of the kind of thing he is 
thinking about.  Do you remember SAIL?

The Stanford AI Lab had the AP wire online 24 - 36 hours worth).  You 
could query it for stories having a boolean expression of keywords. 
We often telneted in and would use it.  If you were a registered user 
of SAIL, you could have it send you mail when articles you were 
interested in came by.  And believe me, it got a lot of use this was 
a very exciting time:  Agnew's resignation, Watergate, Allende's 
overthrow, etc. etc.

But I concur with your list below.  Although these were not 
"generally available."  SAIL was to us!  ;-)

Take care,
John


>Hey guys, depends on what you mean by "online" database and 
>"network".   Online databases predated the ARPANET by years. 
>Project MAC/CTSS had online databases accessible remotely over the 
>telephone network, used by ordinary people for research in the early 
>1960's. In the late 1960's and early 1970's there were lots of 
>databases that were replicated across networks and accessed across 
>networks - Lockheed, DRI, NEXIS, are just some examples.
>
>IBM had distributed databases that were accessible over SNA in the 
>early 1970's, using CICS.
>
>SABRE was distributed quite early, and again was a networked 
>high-volume transaction system.
>
>SAGE was a distributed database system itself, giving real-time 
>access to a wide variety of dynamic data, in the 1950's (I think).
>
>If you are looking for WWW browser accessible databases, the first 
>PERL-script systems that built dynamic web pages from databases that 
>I know of were created at U of Illinois, by folks like my friend 
>Brygg Ullmer, who was 14 or 15 years old at the time (1992 or so). 
>Probably at other early locations like CERN.
>
>But years prior to that, in 1983 or so Ray Ozzie built the core of 
>Lotus Notes, which is a fully distributed, replicated database 
>system.
>
>And the idea of decentralized replicated databases on networks 
>precedes my 1978 Ph.D. thesis about coordinating atomic actions on 
>them.   Examples include work by Stonebraker at Berkeley and System 
>R* at IBM in 1975 or 1976.
>
>At 10:21 PM 1/17/2003 -0500, John Day wrote:
>>At 21:17 +0000 1/17/03, David L. Mills wrote:
>>>Rick,
>>>
>>>I submit the first online ubiquitous database was the ARPANET host name
>>>file. This may not be what you have in mind.
>>
>>;-)  Well, strictly speaking yes, but . . .
>>
>>Perhaps more closely,  actually I would say the NIC and NLS was 
>>really the first database system on the net.  Then there were the 
>>CCA guys who were really trying to do distributed databases.  But I 
>>don't remember when they were up and running.  And also the 
>>National Software Works!
>>
>>Take care,
>>John