Re: Fwd: buggy sorting function
Thomas Davie <[email protected]> Fri, 16 Sep 2005 16:21:41 +0100
| Newsgroups | gmane.comp.lang.haskell.hat |
|---|---|
| Message-ID | <[email protected]> |
On 16 Sep 2005, at 16:09, Malcolm Wallace wrote: > Thomas Davie <[email protected]> writes: > > >> lappybob$ ./hellSort >> [6,4,2,9,10,9,3] >> >>> [2,3,4,6,9,10,9] >>> >> ^CKilled >> > > This looks like a problem straight-away. The "Killed" message > indicates that your OS is killing the program and truncating the trace > file, whereas the ^C signal ought to be caught nicely, allowing the > program to finish flushing the trace file to disc. > > >> hat-detect (error): file hellSort.hat is too short >> > > Yup, looks like the trace file was incompletely written to disc. > > In the given example program, you don't actually need to ^C it - just > give it a blank line as input and it should terminate normally. Oops... dumb bob... > Given that, I was able to use hat-observe and hat-trail successfully, > but the current CVS versions of hat-detect and hat-check are slightly > broken (they don't deal correctly with DoLambda 0x5). I'm working with my new version of hat-detect (although not of hat- check). I knew it would be something obvious, thanks people :) Bob