-
Notifications
You must be signed in to change notification settings - Fork 0
Open
Description
Many times IPFS will just fail for a request or two and then resume working. I assume this is due to the relatively volatile network that we are relying on at the moment. In any case, just waiting a bit and retrying should be a suitable work-around for this case. We should work out a nice approach for queuing these retries and design the data flows such that it can handle such an interruption. Perhaps using promise chains is not the most suitable approach here, as we are doing currently.
Reactions are currently unavailable
Metadata
Metadata
Assignees
Labels
No labels