digital-garbageandClaude Opus 5
090669e950
v4's Chain Sampler became a hub: 8183 lines, roughly seventy optional inputs, a _sample_chunk that grew six ALG arguments, and a hardcoded list of knobs silently inert on one model family. Every technique that needed the loop was added to it, so it collected all of them. This one picks an algorithm and a schedule from ComfyUI's own registries -- so a sampler it gains appears here without an edit -- and runs them. Three things stop it becoming a hub again. It never learns what a model-side modifier is. Sampler-side ones reach it through ONE call site, chain.process(ctx, latent), which cannot grow because the step travels in a context object rather than in the signature. And it declares the hook points it offers, so core hands it only what fits and it never names a modifier. Some modifiers genuinely cannot ride on a model: they need the step index, the schedule, or a second pass. Those attach here instead, and one asking for a hook this sampler does not offer is absent and said so, not half-running. Verified against ComfyUI's own guider path with real weights: bit-identical output. That took correcting the method, not the code -- see below. MEASUREMENT, corrected: sampling on this Mac's MPS backend is NONDETERMINISTIC. The same node with the same seed and the same inputs differs run to run: 2.42 by max latent difference, 0.0008 by mean pixel gap. Comparing by MAX said this sampler diverged from core by 14.2 and it was about to be "fixed"; by MEAN against a measured noise floor it is 0.00000 while core against itself is 0.00183. Every comparison here now runs a control twice first. Re-checked the Detail modifier the same way: 145x the floor, so that claim stands. Also matched what core's sampler actually does and this did not: fix_empty_latent_channels using the ratios our own empty-latent node reports, the preview and progress callback, the move back to the intermediate device, and dropping the ratio keys so nothing rescales twice. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
ComfyUI-FunPack v5 — "Cunning Cranberry"
FunPack v4.0.0 is going to be the latest release of classic FunPack. Next update is another milestone update, or shall I say, a full rework.
Maintenance of current FunPack became a mess — forgetting to add or remove buttons from/to UI, finding functions that don't do anything but still present in the code absolutely randomly, trying to keep the UI neat while it has no strict boundaries on how it should look like, and of course — making it easy for users to understand how FunPack operates and how you can operate it — this is all time, money and token-consuming.
General idea behind the FunPack v5 rework is writing it from scratch, with a new architecture where a feature can't be forgotten, where a wire can't be missed, where one failing module never stops whole process from executing. Loaders and features would tell the UI by themselves what sorts of inputs they need and where to put it — and UI will compose them according to design guidelines it would have. Samplers would become independent from features — if a feature is available and enabled, sampler would call it. If it's not — it gets skipped and doesn't even appear in the UI so it won't confuse you.
Most of this comes from a FunUI backend draft that got shelved after understanding on how much time has to be invested into polishing it. Without a community support and wide usage it would become another dead project only one person in the world needs but no one else.
Come to think of it, FunPack already became something alike, with an exception — this project is more than alive, and while I can, I will support it.
Final goal is to make FunPack an ultimate answer for a novice's question "I don't know how to generate — what do I install and how do I start?" and for a professional's question "I need a UI with all the cutting edge techniques available on day 0 — which one?"
Preparation for the v5 had started.
FunPack v5.0 "Cunning Cranberry" — coming fall 2026.
This branch is a clean start. Classic FunPack lives on main.