New n8n Version

I have been updating my n8n version (0.202.1b) for a long time.
“Start” has been dropped by a click trigger.
Now I have noticed a few things that have fallen away:
There is no feedback anymore which node is currently running. Only when the workflow is complete there is an overall feedback.
Starting from a single node often has empty data as input. (If e.g. the node before has not been executed yet). Before n8n automatically knew which path to run.

Will this function stay like this forever or will it be upgraded again. Or will they be removed completely?

I’ve noticed some funky UI behaviour as well. Not sure if it is the same as you describe, but sounds like it.
I am guessing it is some kind of bug or setting we are missing.

Hey @lublak,

What browser are you using? I have just updated my install to 0.202.1 and it is showing me which node is running (I am using Brave).

The start node being dropped is by design and it was replaced with 2 other nodes “Manually” and “When called by another workflow”.

The single node part I think was raised internally so I will need to take a look and see if I can find out what is going on with that.

Hey @Jon,

i currently use the Edge Browser.
So i started via click handler:


You can see that it is still running. But you can no longer see which node is running. Before, a spinner was displayed.
Some other strange issue:
image
This happens when I have not run it before but now start a single node.
Sometimes it runs through anyway.
Here is some other issue:
if create something like this and you execute the last node twice you also get something strange:

Here all examples as workflows:
For the first two:

For the last:

Pin Data for Code1

[
  {
    "test": 1
  }
]

Hey @lublak,

I am just not seeing that first issue sadly.

image

For the second issue it looks like it works the first time but not on runs after that but matches what we have seen internally. I am not sure if we created an internal ticket for it though so I will create that now.

@Jon i created a gif for it:
chrome_SQznfgwkph

Hey @lublak,

Cheers for the gif, Do you get it with other browsers and does it also happen when using private browsing mode?

@Jon has issues:

  • Chrome: true
  • Chrome (private): true
  • Edge: true
  • Edge (private): true
  • Firefox: true
  • Firefox (private): true
  • Firefox (developer edition): true

@lublak that is very odd, I don’t get the first issue in Brave, Safari, Chrome or Firefox. Normally if I see Edge I assume Windows but as it can be installed on almost anything now… What OS are you using? If it is windows we can see if @marcus or @MutedJam is able to reproduce that one.

@Jon I use windows as a client also as a server.
n8n is installed via npm.

might also be a issue with an anti virus. had some funky behaviour with bitdefender here. I am now adding n8n environments to exclusion list to fix some of them.

The funny thing is that I have not changed anything on the server but only updated n8n. Only after that the problems occurred. Windows Defender is currently running on the server.

Would be something in your front-end not the server side.

Issue 1 works for me on Win10 with Edge and Edge (Private).

Also works for me in Chrome and Firefox including private.

I even tried it on several computers.Even directly on the server:

Is there anything in the browser console?

image

And yes i love n8n :stuck_out_tongue:

Even with diagnostics enabled to get the same rudderlabs message I don’t get the same issue, Very odd.

I think i have something similar. n8n self-hosted behind reverse proxy.

I will try it tomorrow and report.

Can you share the javascript code inside Code node to try it on my instance?