Re: Coyotos : A restatement

[Jonathan S. Shapiro]

All:

Mr. Fortaine has been running around stirring up trouble among several
senior researchers recently, sharing quotes and statements back and
forth without context and wasting a great deal of time.

Concerning his latest post here, he has offered two links:

 [1] http://www-1.ibm.com/linux/news/semiconductor.shtml
 [2] http://www.computationallogic.com/reports/files/028.pdf

The first was offered in response to my statement that there are zero
example of high-robustness or high-security monolithic systems today. I
think his point is that the Linux systems have been running there for 5
months. Just for the record, 5 months of uptime is not considered "high
robustness", and the article says absolutely nothing about security.

The second link is a reference to Bill Bevier's dissertation, which is
entitled "Kit: A Study in Operating System Verification". Matt Kaufmann
suggested this link to Fortaine at about 4pm EST yesterday, but Fortaine
obviously hasn't read it.

First, let me say that the Kit work is brilliant. It was a major step
forward in the state of the art in operating systems work, and it
remains an excellent piece of work today. In fact, I spent a bunch of
time speaking with Bill Bevier about the work about two years ago.

While Kit is a terrific piece of work, Bevier is very straightforward
about acknowledging that the Kit system is a long way from anything that
could be considered a "production capable" kernel. Kit shows that
verification on a more capable operating system kernel may be possible
with a lot of work. It does not show that it has been done.

Yesterday, Fortaine sent a note to Matt Kaufmann that quoted me out of
context and created a serious miscommunication. As a result of this and
other actions, I have asked him NOT to forward any other mail of mine
(which request, please note, he has just ignored). One day later,
without waiting for my response to Kaufmann, Fortaine is now attacking
me here.

Fortaine closes with:

> For me, this man and his project are only a bad joke ... Be more 
> serious, guys :-) !
>
> Yes, in fact, I believe if you want to see the HURDNG boot one day, we 
> will need to build our microkernel and choose our language ....

As I have said in a previous, extensive note, building a kernel may in
fact be the right thing for the HurdNG team to do. I've also given some
ideas about how this should be decided.

But please keep in mind that Fortaine has no idea what he is talking
about, and that his "conclusions" are established by a series of
discussions involving out-of-context quotes on topics that he doesn't
really understand.

Also, keep in mind that Fortaine has absolutely no knowledge of the
financial investment that is going into Coyotos and the progress that
has been made on its implementation. In fact, he hasn't even *asked*
about this!

Regards,


Jonathan Shapiro