TTP_condsetarea_merge_test

Being straight about this one: it's not called out separately anywhere in the pack's README - the "Condition Merge Node" section describes the concept once, without distinguishing this variant from the plain TTP_condsetarea_merge, and it's got zero search impressions and no community discussion I could find. The _test in the class name is the author's own signal that this is the experimental build, not the one to default to.

What it is. Functionally identical to TTP_condsetarea_merge - same conditioning_batch and coordinates inputs, same strength (0–10, default 1), same single CONDITIONING output that stamps a batch of per-tile prompts to their matching regions and merges them for the sampler. The one addition is group_size (integer, default 1, minimum 1). Given the name and the fact that it's the only thing separating this node from its non-test sibling, the reasonable read is that it lets you cluster multiple entries in the coordinate/conditioning batch into a shared group before merging - for example, group_size=2 applying one merge pass per pair of tiles instead of one per single tile - but that's inference from the parameter, not a documented behavior, so treat it as a hypothesis to verify against your own output rather than a confirmed feature.

Which one should you actually use? If you don't have a specific reason to need whatever group_size does, reach for the plain TTP_condsetarea_merge instead - it's the one the README actually documents, and "test" in a node's class name is generally the author's own hint that it's less battle-tested than its sibling. Come to this one specifically if you've hit a case where you need to batch tiles in groups rather than one-to-one with coordinates, and are prepared to verify the exact grouping behavior yourself before relying on it.

Install. ComfyUI Manager, search "Comfyui_TTP_Toolset", or cd ComfyUI/custom_nodes && git clone https://github.com/TTPlanetPig/Comfyui_TTP_Toolset, then restart. Same as its sibling - no models, no extra dependencies, pure conditioning-data manipulation.

Common issues. With no community reports and no dedicated README section to lean on, I don't have sourced failure modes to give you here, and I'd rather say that plainly than invent plausible-sounding ones. The practical approach: build a small test graph with two or three tiles, a group_size of 1 first (to confirm it behaves the same as the plain merge node with that setting), then bump it and watch how final_text-style debug outputs or a quick sample change - the same "verify before you build on it" approach worth taking with any node this thin on documentation and traffic.