I’m not sure it is possible, I’m afraid to say. Although [Step 1 linked Automation log] is available in automation formulas, I don’t know of any way to extract it using markdown, sorry.
One workaround would be a create a new to-one relation field to save the “Most recently linked Automation Log”, and create a Rule that triggers whenever an Automation Log is linked, and saves the newly linked Automation Log entity to this “Most recently linked Automation Log” field.
Indeed, this is workable, but I believe that there is a non-zero possibility that you could ‘lose’ events if they happened in quick succession, e.g.
two items are linked, and by the time the automation checks the to-one relation, the first value has been overwritten by the second linked item, causing the first link event to be lost
There will be an automation triggered for every linked entity, and if this automation’s action is to update a to-one relation with the linked item, any secondary automation which is triggered by this to-one relation being updated will also be triggered twice.
But any secondary automations will sit in the queue waiting to execute, and by the time they run, the update value of the to-one field may not be what you expect.
As an example, imagine this sequence:
To many
To one
empty
empty
User links two things
To many
To one
ThingA, ThingB
empty
Automation 1 (that triggers on linked item) is triggered twice.
First execution takes place - it updates the to-one field
To many
To one
ThingA, ThingB
ThingA
Automation 2 (which triggers on the to-one field being updated) is triggered, but will only execute when the queue of automations is empty
Second execution of automation 1 takes place.
To many
To one
ThingA, ThingB
ThingB
Automation 2 (which triggers on the to-one field being updated) is triggered again, since the value of the to-one field has changed.
The first execution of automation 2 takes place, and sees that the to-one field contains the value ‘ThingB’.
The second execution of automation 2 takes place, and sees that the to-one field contains the value ‘ThingB’.
Ideally, yes, but as @YvetteLans has discovered, markdown does not support referencing [Step 1 linked item] unlike formulas, so I can’t think of a way of making a comment that references the linked item all within a single automation rule.
I don’t think so. I mean, if I drag and drop an item into a column on board view (where the columns are entities in another DB) and then I drag and drop another item into the same column, their relative rank depends on whether the second drop is above or below the first, not a function of the order in which I dropped them.
I can do that, but I don’t think I will then be able to create a comment in the ‘automation error log’ entity/thread view That’s the ultimate goal and I can’t create a comment in another entity as far as I know…
It’s not a real must have → we also have a view with all automation logs. But it’s not very user friendly. Since the user then needs to check that view once in a while to find out if something went wrong.
The thread view is an awesome way to inform a user without a complete inbox overwhelm.
We have a lot of integrations and some of our bigger clients have 1000+ automations actions per week.
If a third party server is down, they will get an error notification of each and every automation that fails if I use the normal notify automation in Fibery.
That’s not helpful. There will be an auto retry every hour for those failed automations. I only want to auto inform them in the meantime about the problems without overloading their inbox.
Thanks, awesome! I have more use cases for this; will put them in the feature request
My approach to centralized error tracking has been to create a special “Error” DB.
When an error is detected anywhere in a Workspace script, it is reported by creating a new Error entity with all the details.
The drawback to this approach is that javascript is needed if the Error DB is not linked to the subject DB. But it is fairly easy a script. The script can also assign the new Error entity to appropriate user(s).
However, the errors that we log are not caused by scripts within the workspace, but in automations outside the workspace.
The external integrations run via Node-RED (Google Calendar, Calendly, ActiveCampaign, Order automations, etc.)
When the automation start, we create an entity in the automation log database and include the payload. If the automation run without errors, we update the entity status to finished. When an error occurs, Fibery will trigger the external automation again (and includes the original payload).
As a Fibery partner, we handle the errors so no need to assign a user. Our clients can see the status of external automations in a view. And we would like to inform them about the errors so that they don’t have to open the view for that.
If I want to use the thread view for that, I only have 1 entity plus comments.
Let’s assume I also create a special DB for errors only. Would it then be possible to create the comment in the automation error channel via script you think?
If so, I don’t fully understand why it is possible via the error DB and not possible via the current log database
Thanks! My script skills sucks but I think the challenge is that I want the URL of the error (either in a separate error database or the current automation log database) and put that entity in the comment of another entity (the channel entity where the thread view is based on).
So it’s won’t work if a script simply just add the comment; that will create a comment in the origine database instead of the channel entity I think?
That’s the ultimate goal and I can’t create a comment in another entity as far as I know…
