Re: [tlug] Novel embraces Microsoft

[stephen@example.com]

Keith Bawden writes:

 > I'm not sure if writing a tool to pull some meta data is going out of
 > ones way to "screw up the kernel development process".

He was specifically warned by Bitmover and asked by Linus to cease and
desist.  The consequences of continuing were known.  He was definitely
violating the license under which he received the software.

 > I saw a demonstration on how Andrew Tridgell "reverse engineered"
 > bitkeeper's proprietary protocol to create his tool. He used telnet to
 > connect and then typed help. He then used this output find out which
 > commands he could issue to pull out the meta data data he needed.
 > 
 > I'm not a developer, but is the above really reverse engineering

Yes.  Building git to provide bitkeeper-like behavior is forward
engineering.  Figuring out how bitkeeper works from how it behaves is
reverse engineering.

I also suspect that there was more to the reverse engineering than
simply accessing the stream.  Ie, it presumably needed to be decoded
into commands relevant to version control operations.

 > or unethical in any way?

That's not my place to decide for anyone else.  Reverse engineering
per se is clearly not unethical, it's just debugging with somewhat
different emphasis.  Reverse engineering without a license is surely
fair use in copyright law.

Distribution of the product of reverse engineering is clearly beyond
the pale, though.

 > After all isn't the meta data also the property of the person who
 > owns the data it is related to?

No.  If you have access to an American passport, look at the first
page.  It is not the bearer's property, it is the property of the
U.S. government.  I'm sure this is similarly true for most other
countries and it is also true for credit cards and ATM cards issued in
the U.S. (ie, they're property of the issuer).  You will definitely
get busted if you try to access the bank's metainformation about your
account.

Also, look up "Westlaw index".  The law in the U.S. is public domain.
The Westlaw indexing numbers (which are required for citation in some
cases) are proprietary.

 > If it is then should one be allowed ready access to this?

If it were already your property, then when it comes out of the telnet
program it would still be your property, and there is no problem.
There are precedents for *not* having ready access to personal
information (eg, the blind trusts that US officials put their wealth
into, at least at the cabinet secretary level), but these are
obviously special.

 > Especially if the mechanisms are already made available to do so
 > via using something as simple as telnet?

Shoplifting is a simple mechanism for satisfying various needs, and
typically things like candy bars and cans of beer are left lying loose
on store shelves, yet it is widely considered unethical to pick them
up and leave the store without paying for them.