Talk:Two Generals' Problem

Archives

This page has archives. Topics inactive for 90 days are automatically archived 1 or more at a time by if there are more than 4.

Diagram for section "Illustrating the Problem"

I believe a diagram would help better visualize the problem then just plain text ShiberToast (talk) 18:15, 12 February 2020 (UTC)[reply]

Issue with "Proof by Symmetry" subsection in section "Proof"

Proof by symmetry assumed that the attacker can send fake messages, but in the problem, the attacker can merely intercept messages. Should the subsection be removed? ShiberToast (talk) 18:38, 12 February 2020 (UTC)[reply]

(Good) source for "first proved unsolvable"

"The Two Generals' Problem was the first computer communication problem to be proved to be unsolvable" is a very strong but questionable statement. What about the various halting problems? Church, Turing, Godel, etc were working on unsolvable problems in the 1930s and 40s. Without an explanation and a great reference, I think we should drop that statement. --Drpixie (talk) 00:55, 2 May 2022 (UTC)[reply]

agreed Titaan123 (talk) 08:29, 24 May 2022 (UTC)[reply]

Yes, the halting problem was proved unsolvable long before the two generals' problem.

However, the halting problem is not a "communication problem".

I added a reference that says the closely-related, more complicated, Byzantine generals problem was "one of the first" computer communication problems proven to be unsolvable. It seems likely to me that two generals' problem was solved before that.

It would be nice if we found a reliable source that specifically named the first few computer communication problems proven to be unsolvable. (If it turns out that the two generals' problem wasn't proven until much later, what Wikipedia article would be best for mentioning that source?). --DavidCary (talk) 22:20, 31 August 2022 (UTC)[reply]

Engineering approaches

I dont think this sentense is right, but it's hard to say with all the negation:

There is no algorithm that they can use (e.g. attack if more than four messages are received) that will be certain to prevent one from attacking without the other.

There seems to be a trivial algorithm they can use to guarantee no one attacks without the other: Simply don't attack. In other words, the article does not guarantee that the algorithm is required to make progress. — Preceding unsigned comment added by ~2025-41753-41 (talk) 10:27, 19 December 2025 (UTC)[reply]