Re: hah

On Wed, 2005-04-20 at 16:08 +0200, Henning Brauer wrote:
> * Jeroen Massar <[email protected]> [2005-04-20 15:51]:
> > Or do you want M$ to include BSD code
>
> yes. that's why we write it, for everybody beeing free to use it.

The protocol or the code? The code in this way unless somebody reverse
engineers it for you, makes a small mistake and thus makes an
incompatible version. joy...

> [no CARP spec]
>
> why don't you help and write one then instead of just whining

Oh sorry that I request the specification of a protocol when you are
whining about the fact that some company did write their documents and
filed them ;)

Why should I reverse engineer and document your code?
Most people I know actually first write a specification before they
randomly start putting code into their stuff.

But this clearly shows why IANA would not allocate you a proper protocol
number, no docs -> no protocol number (*1)

You could have maybe read what I wrote, but I'll abstract it for you
again:
 - a spec allows people to say:
    'we implemented that document and followed those rules'
 - thus no ambiguity
 - thus allows for interoperable implementations

But you do also know most likely that organizations don't like to rely
on a single vendor for an implementation. And for CARP there is only 1
and that single 'vendor' does not even have any support, coder goes on a
nice long holiday to hawaii and code bugs, no support... you can't have
that in a product now can you? You need docs, you need a spec.

Btw to quote: http://www.ietf.org/proceedings/03nov/193.htm
8<-----------
The conclusion of the
discussion was  that an internet-draft needs to be prepared in order to
seriously consider  CARP in the context of this working group.
----------->8

Apparently writing a draft how it actually works, there is not even a
real presentation about it available, is too hard. But then again those
people are whining of course. Wonder why you then took the effort of
presenting the thing in the first place, without a draft... "hah"? :)

I wonder what you would have done if there was no BGP spec around, would
you have reverse engineered it in a clean room and then implemented
OpenBGPd ? I wonder how many different assumptions you would have made
about it in those cases ;) Given you have the code in this case, but

Greets,
 Jeroen

*1 the thing you most likely read:
http://www.iana.org/cgi-bin/protocol.pl
"We need some documentation on your proposed use of a protocol number."

For system port numbers they are being silly and demand that something
is an IETF standard indeed, how can something become an IETF standard if
it is not tested yet and with what do you test it ? :)
Also IANA is most likely the slowest bureaucratic organization in the
list of internet related organizations but that should not hinder the
above.

[demime 1.01d removed an attachment of type application/pgp-signature which had a name of signature.asc]