Open Government Data: Anno 1994
Just yesterday I attended a first working brunch related to the Austrian OGov Data Initiative. The workshop had the goal to detail the data we are interested in and also to define the overall goal and degree of formality ( detailed results).
The turnout was very good, with a good mix of individuals, university people, smart companies. Needless to say, I saw no observer from the public sectors.
And - being a methusalemic Internet person - I had some flashbacks, reaching back into the early 90'ies.
The BIBOS "Experience"
I was university assistent at that time. And while I was supposed to write my thesis, I rather preferred to experiment with this new thing called the Internet.
At some stage, my colleague Manfred Hauswirth and I had our eyeballs on BIBOS, the Austrian OPAC (open public access catalogue for all Austrian libraries). It was running in a mainframe environment and - think IBM haydays - could only be accessed via a 3270 terminal emulation. Adoption therefore had been poor, despite the high quality of the data.
As even we computer scientists had our problems to make it run comfortably on PCs and SunOS boxes, Manfred and I decided to create a web gateway, one which would offer HTML FORMs and HTML results on the user end.
"NO" is NOT an answer. It's a challenge.
Of course we had contacted the library people upfront to get an official API access to the database. But all we got was a "go piss off, only over our dead bodies". I may paraphrase, but only slightly.
Now this could not discourage Manfred and /me. We spent countless afterhours to translate the 3270 (differential) character stream into a VT100 80x25 virtual terminal. Some of the regexps were half a mile long. That screen we grabbed (ASCII, easy), and prettified it as HTML.
Needless to say, the application (think Perl4, think CGI) was extremely wasteful: At every request we would log into the BIBOS system, skip several intermediate control forms, and collect all results by rapidly paging through them.
Success is Your Enemy (especially in Austria)
Within weeks the popularity of the service picked up and quickly spread into the university and also student population.
I had been foolish enough to deploy this on a diskless Sun ELC, so it was some stress to see 80 concurrent Apache children in the process list and the machine swapping heavily over a 10 MBit local network.
We - now think the Frankenstein brothas - invested even more work to manage the monster we created.
Central Services Called
Not long afterwards I got a letter from the ministry of science (which was my official employer). I was ...
... required to attend a meeting regarding the "BIBOS Gateway" we created.
As Manfred cleverly had been at a business trip at that time I turned up there alone. Only an old (I probably should write "senior") Ministerialrat was there in the room, as was one OPAC library man and one manager person I did not know.
Soon the librarian accused me of abusing the existing BIBOS access. I would have had no right to open the access to the web. I - in turn - questioned whether he had any authorization to say what I can do or not. The discussion turned heated and vicious.
The Ministerialrat had remained silent, pacing around the room. But when the discussion deadlocked, he decided to back up his pawn. The ministry "would consider the future of this gateway".
While I was desparately trying to control my adrenaline rush, the unknown manager stepped in:
"The Gateway has become an invaluable service to the Internet community. There is no way the service will go away. We will provide all the necessary hardware to continue to run this service."
Within a second the meeting had come to an end.
A technical description of the gateway we published at one of the first WWW conferences.
It has happened before. And it will all happen again. But this time, I will not be Mr. Niceguy.