Hey, devs. Fix the prompt enhancer please.
Congrats on adding the prompt enhance button.
It seems it wasn’t tested properly, though, because prompts over 1000 characters get abruptly cut off.
Obviously it should respect the character limit of whatever model is selected and generate freely up to that 👍
Whilst we’re here - personally I’m genuinely liking some of the direction. However, just a thought:
- I don’t care about ‘characters’, I care about bugs and workflow page hassle.
- I don’t care about ‘then and now’, I care about bugs and workflow page hassle.
- I don’t care about ‘apps’ when they’re not on explore mode, I want to be able to integrate my own custom third party steps into workflows so I’m not left waiting for basic stuff to be implemented that I could just do myself if it wasn’t locked down.
And for the love of god please oh please fix the unbelievably annoying thing of entire workflows just silently stopping because more than two nodes tried to start at once.
Finally, here’s a feature idea:
It’d be great if we could programmatically change node characteristics, and use logic gates. Eg even something as simple as setting the generation length, via an input. This wouldn’t be such a problem if it wasn’t such a PITA re-jigging nodes around.
_________
I’m not sure who your target market is. Is it professionals? If so, please consider opening up the walled garden so people can implement their own nodes - because it doesn’t seem you’re realistically going to be able to keep up with larger players in terms of usability whilst also trying to do the other magic stuff you’re doing.
The model selection alone isn’t special any more - we can literally use your models on other platforms (though let’s be real everyone is just using seedance right, lol), and those platforms keep integrating useful features and UI you don’t have.
Anyone can vibe code now. Why not use that to your advantage by making workflows ‘moddable’.
Essentially: comfyUI depth plz.