----- Original Message ----- From: "Frank Ellermann" <nobody@xxxxxxxxxxxxxxxxx> To: <ietf@xxxxxxxx> Sent: Tuesday, February 21, 2006 3:57 PM Subject: Re: 'monotonic increasing' > Marshall Eubanks wrote: > > > a RFC-2119 type RFC to define mathematical terms ? > > Maybe more like some glossaries (Internet, security, > I18N, ...), informational RFCs. But I think that's > unnecessary. There are online math. dictionaries, > authors can provide references for unlear terms, or > say what they mean. > > > Otherwise this thread is unlikely to do much to > > change the situation. > > It highlights why "clear" terms in RFC are good, > defined by reference or inline. In some groups > saying 'header' instead of 'header field', 'byte' > instead of 'octet', or 'charset' instead of IIRC > 'encoded character repertoire' is enough to start > a thread. And 'monotonic increasing' instead of > 'strictly (monotonic) increasing' is apparently a > similar issue. > Bye, Frank > What I see from this thread is that there are two common interpretations to the phrase 'monotonic increasing', either a sequence in which each number is greater than or equal to its predecessor, or one in which each number is strictly greater than its predecessor, with the former meaning having somewhat the greater support (at least amongst those with access to text books): which, of itself, makes it a risky term to use in a specification. I still think that it is sometimes used in RFC and I-D in a third sense, of a sequence of integers increasing by one each time, not a meaning anyone has supported. But only the editor can know what is really intended. So, the next time I see it used, perhaps in a Last Call of a pkix, kink or secsh I-D, I will seek further clarification. Tom Petch _______________________________________________ Ietf@xxxxxxxx https://www1.ietf.org/mailman/listinfo/ietf