The comforting thing about one-sided attacks is that they come from both sides!
Rationalists emphasise rational and hence necessary truths and may take the view that all truths are necessary.
Leibniz, for example, does make a clear statement of the necessary/contingent distinction, but his principle of sufficient reason has been held to entail that all true propositions are analytic, and he also held that all analytic propositions are necessary, whence he or we might conclude that all true propositions are necessarily true.
Empiricists emphasise the importance of experience in the origins and justification of knowledge, and are therefore likely to regard all important truths as contingent.
John Stuart Mill believed that all knowledge including logic and mathematics is a posteriori.
He does seem to have admitted some purely verbal truths, but does not count logic and mathematics among these, so, if there are necessary truths they are not of much consequence.
Attacking the line|
Quine takes front stage here.
Quine's holistic position is an update on Mill's empiricist standpoint which argues that experimental confirmation or refutation applies only to complete theories, including the logical system in which they are embedded, rather than to specific propositions.
When a system is refuted the repair job could result in a change to the logic, and so the logic is just as contingent as any other part of the theory.
Quine figures again by presenting arguments about the untenability of the analytic/synthetic distinction based on the "indeterminacy of translation" of natural languages.
Indeterminacy undermines the synonymy relations on which analyticity is supposedly based.
If analyticity makes no sense then our grounds for believing any proposition necessary are greatly weakened.