Subject: hello from France
From: Vassily (vass@cs.washington.edu)
Date: Mon Oct 16 2000 - 02:00:44 PDT
Bonjour everyone!
Executive summary:
- spent a week with my family in Moscow;
- got settled in INRIA; see end for contact info
- giving a presentation on my work this Friday,
planning to discuss informally Craig's EML with Didier;
my project to be determined by early next week
- attended PhD defence of Jerome Vouillon and talks by
Benjamin Pierce and Martin Odersky
- talked to Didier-student (working on OCaml's type system)
and Daniel-student (working on Nice - concrete and better ML-sub)
For a longer literary exercise, read on.
Writing you from Rocquencourt, a small place near Versailles (and
that's a big place near Paris!). After having spent a week in Moscow,
visited a bunch of relatives and performed the agrarian duty, I am now
adjusting to being in this very different place and working with
people with surprisingly similar interests and views - the research
team of Didier Remy and the birthplace of OCaml - the Cristal
Project at INRIA.
In my first week (or, rather, half-week) I attended the PhD defence of
Jerome Vouillon (the implementer of the "O" in OCaml). He is
apparently very bright (and BTW the first place at this year's ICFP
programming competition, in a team with a couple of other guys at
UPenn), but his answers to the questions of his PhD committee were
very concise - of a yes/no kind. The questions themselves were much
longer and sounded like masterpieces in themselves, containing a good
deal of humor, so I was rather disappopinted to hear only short
answers by Jerome. (Well, almost everything was in French so I
couldn't follow anyway.) Anyway, that didn't affect the outcome of
the defense.
I also attended the talks of Benjamin Pierce (on his work of defining
and implementing a robust file synchronisation protocol) and Martin
Odersky (on his development of Java-compatible concurrent language).
Benjamin and Martin turned out to be on Jerome's PhD committee so they
gave talks while they were here.
Last week I discussed my work and theirs with a couple of students.
(Here they have students named Didier, Benjamin, etc., so some
confusion may arise.) Didier Le Botlan is trying to add higher-order
quantification to OCaml (i.e., to be able to parameterize over type
constructors, not just over types). Allegedly this should be
straightforward.
Daniel Bonniot is working on an out-growth of ML-sub. He was the
student of Francois Bourdoncle, but Francois quit research in favor a
start-up in the area of building - you would never guess - a web
search engine. So Daniel found a solution in becoming a student of
Didier Remy. Daniel is working on increasing the expressive power of
ML-sub, while at the same time implementing it in a language with a
Java syntax called "Nice" (former Jazz?) (one of his aspirations is to
make any legal Java program also a legal program in Nice with the same
meaning. Hmmm...). On my todo list is talking to Francois Pottier
who is working on type inference with subtyping.
Finally, installed Vortex here - everything went incredibly smoothly.
Complements go to Craig and others for spending enormous amount of
time getting the bugs out of the installation process. Now only need
to find a g++ library (to get the linker's -lg++ option satisfied).
But the pre-compiled Vortex runs as-is, which is good enough for the
typechecker.
Contact info - I will be using my usual email address. Other than that:
Vassily Litvinov
INRIA Rocquencourt
Projet Cristal
Domaine de Voluceau
B.P. 105
78153 Le Chesnay
FRANCE
Office phone: +33-1-3963-5139
Home phone - don't have
Email: vass@cs.washington.edu
Cheers,
Vass
This archive was generated by hypermail 2b25 : Mon Oct 16 2000 - 02:00:53 PDT