Hacker News new | past | comments | ask | show | jobs | submit login
R.I.P. Phil Bagwell (typesafe.com)
182 points by SanderMak on Oct 15, 2012 | hide | past | favorite | 17 comments



Some background on the kind of person Phil Bagwell was and his contributions http://blog.fogus.me/2012/10/15/phil-bagwell-rest-in-peace

Of particular note is his work creating Hash Array Mapped Tries (aka functional associative arrays) detailed in http://lampwww.epfl.ch/papers/idealhashtrees.pdf and implemented in Haskell, Clojure, Scala and Rubinius. His latest work involved lock-free versions called CTrees[1] that are likely to create more waves as others become familiar with them.

[1]: https://github.com/axel22/Ctries


is there a good spot to find all the papers he's coauthored?


Here's a start: http://www.informatik.uni-trier.de/~ley/db/indices/a-tree/b/... doesn't include technical reports though.


Phil Bagwell's VList is a really cute data structure. It's an immutable list with O(1) cons and cdr, lookups that are expected constant time and O(lg n) worst-case time, and better cache locality than linked lists. The wikipedia entry is short and understandable, which makes it easy to honor the dead in a pleasant and enlightening way:

http://en.wikipedia.org/wiki/VList

Obviously a very clever guy.


See http://infoscience.epfl.ch/record/64410/files/techlists.pdf for one of Phil's most influential papers. Surprising (and saddening) to hear of his passing. See also http://lampwww.epfl.ch/~prokopec/ctries-snapshot.pdf for a more recent contribution.


Wow, this Ctrie looks like a concurrent Swiss Army data structure. Very interesting, thanks.


Might I inquire how/of what? I know it shouldn't matter, but I'm always curious when the cause of death is not listed. I've read too many of these where smart people decide to leave us of their own accord. WikiPedia was no help.


That's just the animal inside you that's hungry for gossip talking, because it really doesn't matter. And I know you'll come up with some kind of rationalization for wanting to know, but you should respect the wishes of family and close friends, whatever the reasons may be, especially because it really does not matter, the loss being the same with or without the reason your gossip-hungry brain may expect.


Indeed, this has nothing to do with gossip, but everything with the prevalence of depression & burnout in the tech field. It was addressed by Joel from Fog creek several years ago.


Here is a collection of all his papers: http://infoscience.epfl.ch/search?ln=en&p=bagwell


Phil's ideas on data structures are a critical corner stone in making functional programming easy and practical. Though most people may not have heard his name before, his influence on some computer languages was HUGE.


Considering his contributions, can we have a black bar today?


I had the pleasure of having a very long, beer filled conversation about tech, computer languages and culture with Phil while at ClojureConj in Raleigh last year. He was a very intelligent and charismatic man. It's amazing how sad the death of someone you had only known for a short moment can be.


I never heard of this guy, but after reading this I realized I know about the data structure he invited that is used in Clojure and other functional languages. We should celebrate the important impact he had on technology. His work will live on.


I only met Phil once, but it was his suggestion that led to the formation of Underscore, where I now work. I hoped I'd get a chance to say thanks, but now that opportunity is gone.


I'm a newbie on Functional Programming.

But I'm changing that, trying hard at least. 2 things have been a great influence for me: Underscore.js and learning scala at coursera.org.

Perhaps, at some level, I owe these new skills to Phil.

RIP


I met Phil at Scaladays in 2011. He was such a nice guy. I remember having some ideas about how to improve the efficiency of hash array trie maps for certain cases and types of data and I remember how encouraging our conversation was. He'll be missed.




Consider applying for YC's Spring batch! Applications are open till Feb 11.

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: