Re: heatbugs hot!

[Nelson Minar]

>    THEY WORK!

yay!

>    2. Rebuilt GCC using GAS (this turned out to be crucial).

hmm, I was afraid of that. It sounds like maybe gcc under hpux just
doesn't work at all without gas? Or maybe linking gcc/gas code with
gcc/nongas code is causing trouble?

>        a. replaced "rint" with "floor" in Diffuse.m

This yields incorrect behaviour, but I can't think offhand how much
it's going to matter for right now, probably just make things
marginally colder than they should be. I think floor(x+0.5) is going to
be correct in this particular case: the value is always positive.

>        b. added stdlib.h to ProbeMap.m and Probe.m to correct implicit
>           declarations of "free" and "malloc".
>        c. added BondingBlocks.h to CGrid2d.m ("cannot find method,
>           getCoordinate defaults to id")
>        d. didn't fix RandomBit.m: method isKindOf not implemented by
>           protocol (couldn't figure out how to fix this....)

No need to worry about those, sorry for the confusion. There are a few
warnings left in the code - b is there because probe shouldn't really
be calling free/malloc, c is unfinished code, d is a really annoying
gcc misfeature I haven't hacked around.