On Tue, Feb 27, 2007 at 12:07:24PM +0000, Tuomo Valkonen wrote: > On 2007-02-27, Tuomo Valkonen <[EMAIL PROTECTED]> wrote: > > I don't know about that.. The thing is, I like writing algorithms and > > data structures more than learning others' APIs, and the help provided > > by a DB isn't _that_ big... > > There's one considerable benefit from using a ready DB, though: one > could write the thing in Haskell, as no low-level bit manipulation > is needed (which is practically impossible in Haskell and other > high-level languages).
There are some very nice high-level libraries available[1] for Haskell that abstract away even writing the SQL. Of these I've only played with HaskellDB and was reasonably impressed, but I think a lot has changed since then and it may be even better. > But then again, if one decides to write a > more efficient implementation at a later date, a complete rewrite > is needed. (It's quite stupid to wrap from low-level C code to > Haskell and then back to C for FUSE.) I don't think it's implemented anywhere at the moment, but have you seen the ICFP2005 paper by Iavor Diatchi et. al.[2]? The reason I heard about it was through a safe low-level langauge I'm interested in called BitC[3] that's thinking of using a similar technique for representing "(low level hardware) data-structures which involve nested union discriminators"[4]. BitC has a bootstrapping compiler written in C++ at the moment, but hasn't reached the self-hosting stage yet. Sam [1] http://www.haskell.org/haskellwiki/Libraries_and_tools/Database_interfaces [2] http://www.cse.ogi.edu/~diatchki/papers/bitdata.pdf [3] http://www.coyotos.org/docs/bitc/spec.html [4] http://www.coyotos.org/pipermail/bitc-dev/2007-February/000881.html
