That went into effect a month ago. Claude prose quality has been a disaster at least since Opus 4.7

That makes no sense. Watermarking only became a thing in the past month. Claude has been spewing unreadable slop for much longer than that.

ChatGpt/Codex is nowhere near the level of sloppy vomit that Claude generates, so that theory doesnt really hold up.

Claude writing quality got unbelievably bad with Opus 4.7, with no improvement in Fable. Opus 4.6 was fine. Im starting to see it as a security risk - my brain just can't process its word vomit, so just tell it to go on, implement whatever

You can tell Opus 5 or Fable to route answers through 4.6 for rewrite

Telling Opus 5 or Fable to route answers through Opus 4.6 usually works pretty well. 4.7 really was the version where the writing style became horrible. I also have a Codex subscription in addition to to Claude 20x, that i use mainly for rewrites of Claude doc vonit and explaining Claude's plans

F80 would have been the perfect autofocus film camera if only it could meter and stop down manual lenses. F100 is just too heavy

You really need a specific focusing screen for manual lenses though, metering or not. The rangefinder in the AF Nikon F/D cameras is absolutely no good for accurate focus unless you really stop down hard (f/5.6+ usually). Decided to just use the AF-D lenses at the time.

Sadly it seems a lot of Nikon lenses were not designed with reducing focus shift as a priority. That completely wrecks them near their minimum focus distance on film bodies. Phase-detect AF only sees light coming from the f/5.6 direction so AF will systematically miss its target at any aperture with a different focal plane. There is no live view on film, obviously, so no on-instrument way to correct it.

I am going through this right now... I really wanted an 85mm 1.4G but according to reviews it will put focus completely behind the subject from f/1.8-f/5.6 when closer than head+torso distance. In headshots if you target the eyes it will focus somewhere on the temples. I have no idea if it really matters in real life. So I am stuck with the 1.8 G.

Edit: A focus screen with some kind of prism mechanism would let you manually focus, at least, by holding down DOF preview. Richard Haw describes cutting down and shimming off-body focus screens so it's possible. Actually now that I look it up, whoa, an off-the-shelf one exists on aliexpress ("Dual 45 degree Split Image Focus Focusing Screen for Nikon D1 D1H D1X F100 PR129")

Yeah that's exactly what happens. The 85mm 1.8G is really good enough though! I wouldn't want to fudge with focus screens - I managed to cock up a D750 doing that once. Took ages to get it right after.

Really I stopped shooting film in 2022 because of the arrival of CT scanners at airports. Fed up of arguing with people as I do most of my photography out of my home country. And it's expensive. Sold it all, bought a Nikon Z50ii basic kit in 2024 and that was so good and tiny that I sold my D850 and the rest of the F mount lenses. AF on the newer Z cameras is something special, even on the crap bodies. Haven't felt the need to buy an FX Z yet as the DX ones are so good.

Want the best software for editing photos? No thanks, I'd rather use inferior open source solutions. Go Gimp

Upvoting you, because I actually use Gimp. While it has its headaches, for photo editing it's still the best after Photoshop.

(Yes, still better than Krita).

But this is all a digression, as Lightroom/Darktable are not Photoshop/Gimp replacements. Not even the same category.

> still better than Krita

??? Krita is painting/raster image creation software, why would you use it for photo editing?

You shouldn't. But I point it out because people keep posting "Ditch Gimp. Use Krita instead."

Krita's better for painting. Gimp's better for photo editing.

Pretty sure they said "go gimp" ironically

Darktable is top class.

Yeah who needs the best software for hobbyist editing? $20/mo to Adobe forever until they jack it up even more, no thanks.

I have added a skill to make Opus 4.8 use Opus 4 6 to translate its answers into teadable English, sine it had much higher quality prose. Im seriously considering switching from Claude 20x to Codex because I just cant read more of of Opus or Fable's walls of text

In my case, since "new session" Claude couldn't understand it either, I assume it means it's going to contribute significantly to context rot. In your case, it probably results in a lower quality context.

Hey Anthropic, give agents/skills the ability to modify previous context! This could be used to truly clean reasoning and task definition mistakes, corrected by the user, and strip output to more fundamentals, which should strengthen/purify the context!

CLAUDE_REWRITE_HISTORY=true ;)

Editing agents, and my own, messages is still how I get the best results for high reasoning planning, in chat clients. I remove the mistakes, because I naively assume that if your context has mistakes, now your next token prediction will be biased to a space where the AI makes mistakes! It would be interesting to write some benchmarks around this, but I've seen significant differences when I do this for less capable models.

I haven't been impressed by 5.6 SOL extra. Typically I asked to do a summary of our session, and it went to do a summary of the repo. Tried different things and I just gave up.

Opus 4.8 did the same at first and after it blended better the repo knowledge with the session one.

I have both $20 as I am less a power user these days and I can't justify a 100 let alone 200 but each model as its irks.

I stopped reading Claude's ridiculous distracting nonsense walls of text months ago. Honestly, I prefer Codex being boring and basically ignoring any human utterances I make which are unrelated the work to be done.

Adobe