[Date Prev][Date Next][Thread Prev][Thread Next]
- Subject: Mining Code Repositories for Language Design
- From: steve donovan <steve.j.donovan@...>
- Date: Tue, 26 Oct 2010 08:44:54 +0200
Sometimes LambdaTheUltimate comes up with something entertaining:
A paper by J. Garrett Morris discussing using a Haskell code repo to
discover common language idiom patterns. The paper says that
"serve as a valuable resource for language designers: by providing a
large collection of code written by different programmers and in
different styles, it allows language designers to see not just how
features could be used theoretically, but how they are (and are not)
used in practice."
Naturally, somebody found this notion suspect, and there was this
"Suspect compared to what? The alternative method that I know is to
drive language design decisions by divine inspiration."
PS. Sounds like an interesting project for someone looking for a
thesis; the problem with Lua would be finding a sufficiently large
corpus of code in one place. David M's LuaInspect would be a good
starting point as the analysis engine.