7 ms·
Interview with Don Chamberlin, designer of SQL
- petarb 9y agoTook an undergrad comp sci class taught by Don Chamberlin at UCSC in 2009. He is a fascinating professor!
- dunkelheit 9y agoDisappointingly low on technical detail... For me the most interesting part of an interview with the original author of some well-established technology is when he or she is discussing something that you take for granted and says something along the lines of "oh, that thing is more of a historical accident, I would design it a bit differently today". These nuggets can really broaden your perspective, but I couldn't find any of them in this interview.
- taf2 9y agothat's because SQL is perfect obviously /kids
- maxxxxx 9y agoThat's a problem with a lot of interviews. Journalists almost always go for the "human" side but I enjoy reading about the technical evolution and how things developed. I just read a book about the development of windows NT. Same thing, a lot about people but not much about technical stuff.
- zengid 9y agoI think the main nugget is that they took Codds powerful mathematical system for defining and manipulating relations and created a simpler way to interface with it through the English keywords SELECT FROM WHERE. I see a similar thing happening in Elm, as it tries to translate Haskell/ML ideas into simpler concepts without all of the category theory.
- rhombocombus 9y agoAs someone who uses SQL extensively every day at my job, I really appreciate the thought that went into it's design. It is by far the most sane part of what I do every day.
- cosmie 9y agoCompletely agree! Although most of the developers I interface with look at me like I'm an alien. So I don't know if SQL being the sanest part of my day is a point for SQL or a point against the general sanity of my day.
- oneweekwonder 9y agowhat dialect of sql you find yourself in? I try to keep my env to pl/pgsql and like some pl/python on the side.
- Haul4ss 9y agoI've been subscribed to this podcast since his 2nd or 3rd episode. He's had some interesting guests. I think the podcast is a one-man show. He could probably use some help with mixing/production (and a new intro theme), but for a one-man show it's pretty darn good. Edit: I'm an idiot and just realized OP is the podcaster.
- pramodhs 9y agoThank You. Yes, I'm a one-man army and would use some help. Finding interesting guests have not been hard but doing this week after week has been hectic, but I love it. Maybe someday I can run the podcast as a team.
- Haul4ss 9y agoWell considering you're doing it all by yourself, you are doing an awesome job! I look forward to each new interview. I like that you give the guests time to provide an in-depth answer without peppering them with too many questions. When you're a famous podcaster you can do a Mapping the Journey episode about Mapping the Journey! :)
- berbec 9y agoMr. Chamberlin was my dad's roommate at Harvey Mudd and worked with him at IBM. It amazes him how much Sequel runs the world now.
- jason_slack 9y agoOne thing I thought about after reading this interview was just how much these grassroots projects (at the time) were done just as a labor of love. Also, just how free companies were with information. Publish a paper and share it with the world.
- gumby 9y agoI was glad for a transcript! I find it interesting that the technology of the time meant it was hard to write ∀ and ∃ (though APL users seemed to go OK) and so they chose a complex syntax that reminds me of COBOL. The early database applications were business applications (SABRE being #1 I believe) so this wouldn't have been at all an unreasonable basis for SQL's syntax.