steven-2023-strangeloop.txt

## Session title:

* All the good database ideas were used up in the 1980s
* 1985: The Golden Age of Databases
* The database scene peaked in 1984

## Technical Bullets:

* unlimited disk
* vector dbs
* relational and document models

## Cultural Bullets:

* SQL
* immutable data + FP
* strong, dynamic types


## Description v1:

The goals of database research have extended far beyond technical
feasibility for decades.
In the 1970s and 1980s, disk and memory were major constraints on
database design.
The utility of the DSM (columnar data) was unclear.
Relational, document, and graph storage models hadn't proven themselves yet.

In the 2020s, we have unlimited cloud storage, productionized vector
databases, and a clear appreciation of each storage model.
In addition to loosening our technical constraints, software culture
has evolved.
A functional and dynamic query language for immutable data is possible
today, where it may not have had an audience even 10 years ago.

In this talk, we'll retrospect on the database dreams of the 70s, 80s,
and 90s.
We'll look at databases applying these decades-old ideas -- and what
problems they solve.


---

## Title Options v2:

* The 20th Century Database Prophecy
* The fun and optimism of the 80s Database Prophecy
* The Other 1984: A joyful database prophecy comes true
* 1984 to 2024: An optimistic database prophecy comes true
* The Immutable Database Prophecy of the 1980s
* The database prophets of the 1970s: their optimism and legacy
* The Other 1984: An optimistic database dream comes true

## Description v2:

Revisiting database white papers from the 1970s and 1980s, we've
found that there was a certain joy and optimism in the thinking of
that era which parallels our own childhood hacking.
Grand ambitions from that time include infinite disk, immutable
databases, columnar storage, semi-structured records, and unifying
query languages.

Forty years later, many data products have achieved these goals --
though they often do so in isolation.
If 1984 held optimism and potential, 2024 holds the power to make
storage systems as simple as researchers once dreamed.

In this talk, we'll explore the vision and ambition of a few select
papers from that time, then connect each to its modern technical
implementation.
In aggregate, these papers (and their often-unlikely endgame) tell
a fun, exploratory, and hopeful story about the immutable databases
of the 2020s.