* Ronald Schmidt
  http://www.tu-chemnitz.de/~ronsc/tree/2_Projects/0_c++/lm78/index.html
  This one is very ugly. It is the one Phil started with. Alex wrote about
  it:
    [This] code is based on a windows program for reading the lm78 by
    Pedro Ortigao. It doesn't look very nice, you can tell from the variable
    names (like m_SMBBA) that they have been generated by VC++... Also it
    uses floating point which is a no-no in the kernel (fp state is not
    saved, but i assume you know that).
* Alexander Larsson
  http://www.lysator.liu.se/~alla/
  The one we are working with right now. 'nuff said :-)
* Erik Hendriks (Beowulf project)
  ftp://ftp.vu.union.edu/pub/users/hagopiar/lm78-0.1.tar.gz
  Frodo has examined it, and wrote the following about it:
    Well, this seems to be a real nice implementation. Personally, I do not
    really like the multitude of files, though they may be nice if you want
    to write scripts using these values. Perhaps both a more user-friendly
    display and the rather low-level display of this module could be combined.

    Another thing I miss is a conversion of voltage inputs to real voltage
    levels. It is nice to know that 'in1' is at '176', but that is not what
    I want to know; I would rather see that 'VCore' is at '2.85V' (conversion
    not accurate here).
    Finally, not very surprisingly, there is again a lack of CPU temperature
    readings.
* ????
  http://pweb.uunet.de/ganzfix.duew/
  Another implementation recently found. On first sight, this one seems
  rather extensive - it may even be what we are trying to reach. It is
  partially in german, though. Not yet examined.

