We’ve started using gumloop in production behind a small tool using the gumloop API.
It’s super dangerous for me right now to touch anything. Things break way too easily in gumloop.
I wish we could browse through revision, name some specific revisions, and point our API calls to a specific revision so that we can work on timed releases.
Appreciate any feedback.
Ed
Totally valid feedback — this is definitely on our priority list, and we want to make sure we get it right. Really appreciate you flagging this.
On the note of things breaking easily, do you have any specific examples or issues you’ve run into? I’d love to help if I can.
Yes.
If I change some things in nodes, I’ll loose the link to the downstream node and/or the referenced input.
This new way of using “pills” to insert upstream node output looks cool but if:
You can’t visually know that your node is actually broken unless you re-referenced the upstream input.
I don’t know if I’m clear enough. But basically, a node should visually tell me how many inputs and outputs are used. And showing a “ghost” link (a link with no data stream set up yet) makes more sense to me than just suppressing the link and letting the user remember where that node used to plug into…
Yes, that makes complete sense to me.
We’re working on more improvements to the new “badge/pill” UX — first, to make input/output naming immediately clear, and second, to ensure node connections aren’t unexpectedly broken in certain edge cases.
This is definitely high up on our priority list!