Path: utzoo!utgpu!news-server.csri.toronto.edu!cs.utexas.edu!sdd.hp.com!decwrl!ucbvax!A.ISI.EDU!PADLIPSKY
From: PADLIPSKY@A.ISI.EDU (Michael Padlipsky)
Newsgroups: comp.protocols.tcp-ip
Subject: Re: Network Temperature Protocol
Message-ID: <12608289024.36.PADLIPSKY@A.ISI.EDU>
Date: 24 Jul 90 19:20:09 GMT
References: <9007240048.AA00668@bel.isi.edu>
Sender: daemon@ucbvax.BERKELEY.EDU
Organization: The Internet
Lines: 28

Jon--

Since you did request comments, after all, I'd observe that it would be
more appropriate to offer Fahrenheit readings via TCP and UDP, reserving
Celsigrade/Centius for TP4 and CLNP whenever X.500 can furnish the necessary
SAPpiness.  (Perhaps Kelvin should be furnished via TP0.)  I mean, every
time we try to be international about things it only leads to difficulty,
as witness the subsequent messages re what 22C is....

Or should we be spec-ing a CF "gateway", to be contemporary?

Other than that, though, it seems to be one of the better protocols to come
along in some time, except for one technical omission, I believe.  Unless,
of course, the Host Requirements RFC specifies that UDP checksums will be
used for all protocols that don't say not to use 'em, since if one cares
enough about the temperature to ask, one presumably cares enough to want to
get it right.

So I urge the IAB not to progress the TQS to DIS [*] status without changing
it to Fahrenheit for both TCP and UDP, and explicitly requiring UDP checksums
(except for the UNIXtm "man" page, naturally, where total recall of all
remotely relevant context is always assumed).

   cheers, map

[*] Draft Internet Silliness
    (not to be confused with Draft International Sanctimoniousness)
-------