Scheduling with alfred and new syntax.

[MrYawe]

Example

launch_cron.js

require("./client");

const CronWorkflow = require("./Workflows/CronWorkflow");

await new CronWorkflow().schedule("* * * * *")

Same thing for tasks.

[geomagilles]

@MrYawe let's talk about the output, I'm not sure to understand it fully.

• id: what's this?
• intentId: this should not the exposed naming (scheduleId)
• targetType: why not a type in target? (and btw always use lowercase 'workflow' - we have to be consistent)
• why a name?
• it seems to me that canonicalName should be "CronWorkflow"
• a date is probably missing "scheduledAt"

[MrYawe]

@geomagilles

id: what's this?

it's the id of the resource we just created (a schedule)

intentId: this should not the exposed naming (scheduleId)

I'm not sure to understand but today all requests made by the lib to the agent have an intent_id. I think we should do the same between the lib and alfred. You just want to "hide" this field to the client ?

targetType: why not a type in target? (and btw always use lowercase 'workflow' - we have to be consistent)

👍

why a name?

I think it may be useful to be able to choose the name of a schedule as an option (if you have multiple schedule for the same workflow for example). We can't pass this option from the lib today (we will wait for the new syntax ?) but this option is available and ready server side. Today the name of a schedule is an optional parameter and by default it's the name of the target (workflow or task).

it seems to me that canonicalName should be "CronWorkflow"

👍

a date is probably missing "scheduledAt"

yes. I will add the inserted_at and updated_at in the GraphQL response.

[pylebecq]

Seems good to me.

The only question I have is why do you uppercase the programming language?

[MrYawe]

@pylebecq
Because programming_language is a GraphQL enum type so it must be an uppercased string.