This topic was automatically generated from Slack. You can find the original thread here.
Hi.
This doesn’t seem right. On a flow that I have used successfully in the past, suddenly I get errors. I started simplifying it to identify the source of the issue. And there it is. A string that shows perfectly fine in the hovering preview, but that can’t be returned or printed to the console.
In the second screenshot the data looks suspicious. It is formatted like a string.
Error details:
at Object.run (file:///tmp/__pdg__/dist/code/1bc8bf7733aa0c62f48da28423b7e969eeb53a3e5ab72b319e9e7e319c537ecb/component.mjs:10:21)
at null.executeComponent (/var/task/launch_worker.js:242:53)
at MessagePort.messageHandler (/var/task/launch_worker.js:748:28)
I haven’t used Pipedream in a while and I noticed there have been changes to workflows. It’s an old workflow, not yet assigned to a project
Hi Anton, thanks for the screenshots.
Pipedream will automatically unserialize data to and from JSON between steps.
So if the upstream step is not properly exporting a JS object, but instead is exporting a string, then that will break downstream steps.
Is it possible that the step that produces this data is exporting the data in an incorrect format? I’m also noticing that the keys change between your screenshots.
Hi Pierce. Thanks for your quick response.
Guidance from you may prove a bit difficult as I am not a programmer, just someone who has messed around with code a bit. 
So I am struggling to understand what you are trying to tell me here.
As I said, the workflow was working before and I know that the events still come in the same format.
In the previous screenshots you can also see that data is not formatted as a string there.
As you can see the data is coming straight from the event. It’s not produced by a previous step.
What do you mean by “the keys change”?
It’s maybe hard to see in the screenshots, because it is highlighted in a very very light blue, but I tried to show the preview of the different elements.
So the first one is the preview of the text and the second one of the data in text.
So you can see that on the upper level it still looks alright and then it doesn’t.
Could you tell us more about the source of the data for this workflow and the steps prior to this failing one?
I just want to rule out that there’s a data formatting issue, since the code hasn’t changed and since Pipedream hasn’t changed how it processes data between steps, my inclination to is rule out that the source of your data is intermittently changing it’s format.