Wiki
Wiki

Research notes on every problem, and a library of the papers behind them. Built from the open erdos repository.

Updated


Claim. For every graph HH with m≥1m\ge1 edges and no isolated vertices,

R(K3,H)≤2m+1≤3m,R(K_3,H)\le2m+1\le3m,

so in the notation of Problem 569 c1≤3c_1\le3. The one-edge graph K2K_2 is eligible and R(C3,K2)=3R(C_3,K_2)=3, so c1≥3c_1\ge3, and hence c1=3c_1=3. Goddard and Kleitman proved the same theorem independently (Goddard and Kleitman 1994); the same theorem is the case k=3k=3 of Problem 570, recorded on its claim page there, whose account of the statement's sources this page follows. A comment of 27 March 2026 in the site's discussion thread credits the case k=1k=1 to this paper and to Goddard and Kleitman.

Covers. The case k=1k=1, c1=3c_1=3. Nothing about k≥2k\ge2.

Depends on. Nothing in this wiki; the result rests on the cited paper, and the one-edge endpoint is elementary.

Acceptance. Refereed: A. F. Sidorenko, The Ramsey number of an nn-edge graph versus triangle is at most 2n+12n+1, J. Combin. Theory Ser. B 58 (1993), no. 2, 185--196, in the July 1993 issue, the month this page is dated by; the day is a placeholder. The site labels the problem OPEN, so its pages are not acceptance.

Read depth. The paper is not held. The statement, with its hypotheses (nn edges, no isolated vertices), is read in the zbMATH Open review of the paper (Zbl 0794.05090): "We prove the conjecture of Harary that for any graph GG with nn edges and without isolated vertices, r(K3,G)≤2n+1r(K_3,G)\leq 2n+1". Theorem 1 of the 2026 preprint of Cambie, Freschi, Morawski, Petrova and Pokrovskiy attributes the same statement to it. Nothing is independently reviewed in this corpus.