Two rolls of HP5 at box speed, Rodinal 1+100, one hour with a single agitation at the halfway mark. Shadows came up nicely, but the sprocket holes left faint streaks in the sky on three frames. Next attempt: gentle inversion at 20 and 40 minutes, tank tapped hard twice after each.
The thermometer was reading two degrees high. Everything I developed "at 20" was really at 18, so every roll was roughly 15% underdeveloped. A cheap second thermometer would have caught it in May. Lesson recorded, thermometer replaced.
Check the tools before blaming the film.
It works if you split the bag by volume and accept some variation, but the clearing test is mandatory: drop a snipped film leader in the fixer and double the time it takes to go clear. My batch cleared in 45 seconds, so 90 seconds minimum, I give it three minutes.
Fibre curls, takes forever to wash and dries with a slight cockle. It also looks noticeably better than resin-coated paper, which is annoying. Drying between blotters under a stack of old atlases is my current compromise.