Research notes on every problem, and a library of the papers behind them. Built from the open erdos repository.
Updated
Statement
The Dickman function (p. 42) is the continuous function with for and for , .
Theorem 2 (p. 43). Let and be real numbers with and . Let and be positive integers with , where exceeds a number effectively computable in terms of and . Then there are distinct positive integers and distinct non-negative integers with
Here is the greatest prime factor of .
Remarks of the paper (pp. 43--44). The minimum defining is attained, and for , so (7) improves by a power on the trivial bound given by , . For , Buchstab's lower bound for gives . The bound (7) also holds with , the number of distinct prime factors, in place of . In the introduction (p. 39) the authors state that it follows from Theorem 2 that, even for of the form with , the bound (2), (p. 38), cannot be replaced by for every and .
Conjectures stated (pp. 39 and 44). For and every the authors conjecture that some , have for (p. 39); and that there is no positive real with arbitrarily large , , admitting distinct positive and distinct non-negative with (p. 44).
Proof pointer
Pp. 47--48. Lemma 6 (p. 46) is the analogue of Lemma 3 with the smooth-number count taken from Dickman's asymptotic (Lemma 5, p. 46): applying Lemma 1 to the -smooth integers up to gives about integers with every smooth. The proof of Theorem 2 takes , a point where is minimal, and .
Read depth
Claims checked: the definition of , the statement and the remarks above were read clause by clause on the page images of the print. The proof was followed for the outline above and is not independently verified.
Dependencies
- Lemma 1 (p. 39), through Lemma 6 (p. 46).
Source. P. Erdős, C. L. Stewart and R. Tijdeman, Some diophantine equations with many solutions, Compositio Mathematica 66 (1988), 37--56; the edition read is named on the source card.
Bears on
No problem page of this corpus.