Talk:Serial number arithmetic

Originality of the 'General Solution' Section

This section contains no references, makes some impossible claims, and then contradicts itself (betraying the involvement of at least two hands). However, it doesn't look like absolute nonsense. What to do? Preceding unsigned comment added by 93.96.235.0 (talk) 15:39, 10 May 2009 (UTC)[reply]

Doesn't that section simply go to greath lengths in order to claim that if you make use of the negative integer range in two's complement comparisons you gain some margin against wraparound as compared to just using positive integers? That's not a 'General Solution' (by any means), merely a postponement of the problem. — Preceding unsigned comment added by 79.168.138.50 (talk) 03:50, 11 January 2017 (UTC)[reply]

RFC

The RFC referred to is very specifically about the serial numbers used in the Domain Name System. This article however seems to allude to its contents applyting to serial number in general. I don't think that that holds. — Preceding unsigned comment added by Sejtam (talk • contribs) 08:16, 10 May 2012 (UTC)[reply]

I agree that the RFC specifically says that serial number arithmetic and serial numbers are used in the Domain Name System. Also, bits are used in the Domain Name System.

However, as far as I can tell, exactly this serial number arithmetic is used for packet sequence numbers in every sliding window protocol system I've ever seen, not only for the Domain Name System. Also, bits are used in systems other than the Domain Name System.

If you think the serial number arithmetic described in RFC 1982 only applies to the Domain Name System, what do you call the other arithmetic(s) used for sequence numbers in other sliding window protocol systems such as Ad hoc On-Demand Distance Vector Routing, B protocol, High-Level Data Link Control, Generic Routing Encapsulation, TDMoIP, TIPC, OSCAR protocol, RapidIO, sequenced packet exchange, etc.? --DavidCary (talk) 05:31, 6 February 2015 (UTC)[reply]