Especially when dealing with browser automations, various lags and errors lead to rather messy Process level design. To counter surprising attach or timeout issues, Process level flowchart tend to be filled with actions surrounded with blocks, resume/recover stages and retry loop structures. (Yes, you could clean the process by submerging retry logic into object level, but in the end, it's a same mess, but under a different carpet.) My suggestion is to add a Try-Catch tool into a Process Studio toolbox. Idea is to place an action inside of Try-Catch area and handle exceptions the action might cause. Visually it mimics a Block tool, but adds to its design
properties: number of retries action within a Try-Catch will be retried (or possibly a timeout after x seconds?)
two exit nodes, one for positive and one negative end (means despite of retries, object faced an error) - error node could be a Recover stage and it's up to a developer if the error is rethrown or suppressed by resume
to visualize which retry round is going on, add a caption to stage that signals the retry round for developer
The trickiest part is to create a nice design for a error node, but I hope you'll get the idea!
... View more