What reads these clips

[Paul Iadonisi]

On Sun, 2005-02-06 at 19:42 -0800, Dan Hollis wrote:
> On Sun, 6 Feb 2005, Jeff Spaleta wrote:
> > On Sun, 6 Feb 2005 18:56:27 -0800 (PST), Dan Hollis <goemon at anime.net> wrote:
> > > isn't samba reverse engineered too?
> > funny you should mention this...
> > http://www.groklaw.net/article.php?story=20050205010415933
> 
> Well, we know for a 100% fact that drivers/net/forcedeth.c in the current 
> kernels is reverse engineered.
> 
> http://www.hailfinger.org/carldani/linux/patches/forcedeth/

  Last I knew, reverse engineering most certainly was *not* illegal, at
least in the US.  Some companies and individuals would like us to think
so, but even the notorious and much maligned (justifiably) DMCA has a
reverse engineering provision.
  Having reverse engineering provisions in EULAs is, IMNSHO, a scare
tactic.  Of course, IANAL, so take that with a grain of salt.
  There's always the one-group-does-the-reverse-engineering-and-docs and
second-group-follows-docs-and-does-coding trick, (IOW: clean room) if
you want to be extra cautious.
  The biggest problem with some of the aforementioned technologies is
not so much reverse engineering (though I hear the Sorenson codec was
b**ch to figure out, which is why it took so long), but patents.  Also,
another problem with reverse engineering, as Andrew Tridgell brings up
in the Groklaw article above, is that it usually yields bad code.
Better to have to docs in hand.

> IIRC the adaptec scsi drivers are reverse engineered as well. though i 
> think adaptec eventually came to their senses and started helping years 
> later.

  Don't know about the history, but you can download the GPLed source
for the driver for at least aacraid cards.  It is a year old, I think,
and it *appears* that what's in the kernel is a little older, given the
version number and some of the significant differences, but the one in
the kernel also appears to be actively maintained.  Still, from some
things I've read googling around, I wonder if rolling in some of the
updates from Adaptec would be a good idea.  But I digress.
  Nevertheless, the point is that, yes, Adaptec seems to have a clue
when it comes to releasing GPLed drivers (not just specs) for their
controllers.

-- 
-Paul Iadonisi
 Senior System Administrator
 Red Hat Certified Engineer / Local Linux Lobbyist
 Ever see a penguin fly?  --  Try Linux.
 GPL all the way: Sell services, don't lease secrets