[Xastir-dev] N/S and E/W bit

[Owen DeLong]

> I have no major problems with this.  If the group wishes this, it's
> relatively easy to change.
>
> However, what breaks if we take the opposite tack, taking the few
> that are case-insensitive and make them case-sensitive?  That might
> be a short-term hardship for some, but better in the long run,
> correct?  Again, what breaks?
>
I disagree.  In my opinion, the controlling sentiment for such situations
should be "Be generous in what you accept and conservative in what you
send."  This should be the first principal of ANY protocol development.
As such, I would agree we are broken if we send lower case, but, I think
we should accept lower case and encourage others to do so.  Strict protocol
enforcement for the sake of strict protocol enforcement belongs only in
test suites and in the controls of what you transmit.  Receivers should
accept anything that has a clear and deterministic valid intent.

> I assume we're seeing packets out there with the wrong case.  Who's
> generating them and how do we make them stop?
>
Well... I would propose we first figure out who they are, then, for at least
some sampling try to determine the software in question.  Then, contact the
authors of that software and ask them to fix what they are sending.  I think
we should encourage liberal acceptance and conservative sending.  Certainly,
if we reflect a packet, we should send uppercase even if we received lower
case.  Is case defined in the standard, or, was upper case just assumed to
be the requirement by lazy coders?

Owen
KB6MER

-- 
If this message was not signed with gpg key 0FE2AA3D, it's probably
a forgery.
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 186 bytes
Desc: not available
URL: <http://xastir.org/pipermail/xastir-dev/attachments/20040120/0d38b863/attachment.sig>