Here's something I forgot to mention - the code uses TMR1 in "Oscillator"
mode, i.e. it uses a 32kHz crystal to generate the TMR1 clock signal. This is
done using - rather predictably - RB6 and RB7. Could this be the cause of all
this trouble?
I certainly didn't expect XWisp to die the way it did - surely it should have
ignored the PIC's device ID, forced the PIC into PGM mode and programmed it
as if it were a 16F628 (because that's what I specified on the command line)?

Thanks.
--
Phil.                              | Acorn Risc PC600 Mk3, SA202, 64MB, 6GB,
philpem@dsl.pipex.com              | ViewFinder, Ethernet (Acorn AEH62),
http://www.philpem.dsl.pipex.com/  | 8xCD, framegrabber, Teletext
... The current limits placed are based on resistance

--
http://www.piclist.com hint: The list server can filter out subtopics
(like ads or off topics) for you. See http://www.piclist.com/#topics