Subtle bug in Data.ByteString and/or GHC (perhaps)

Donald Bruce Stewart dons at cse.unsw.edu.au
Mon Jun 18 21:44:08 EDT 2007


nad:
> On Mon, 18 Jun 2007, Nils Anders Danielsson <nad at cs.chalmers.se> wrote:
> 
> > I tried the most recent darcs version of binary, and the problem
> > appears to have been fixed, presumably by the patch replacing
> > inlinePerformIO with unsafePerformIO.
> 
> It turns out that, although the code I posted earlier in this thread
> now works, the larger piece of code which I am actually interested in
> running does not...
> 
> I guess this is the price you pay for using libraries which make use
> of RULES pragmas and unsafePerformIO and friends.

I'm keen to identify what your problem is, however, Duncan is on
vacation, and I'm writing up my dissertation. Possibly Lennart can track
this down if its a Data.Binary issue. If its a ByteString issue, Duncan
or I can investigate it soonish.

However, we'll need a test case, so more info please! :-)

Also, the first place to look would be Data.Binary (and also try out the
stable branch of Data.ByteString).

-- Don


More information about the Libraries mailing list