You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
The pagination is a new feature to help speed up page response times. We noticed in the old UI, since everything is streamed whenever you click to expand the task, that if the task had a lot of output, it could get very slow and make the entire page sluggish.
I see your use case though. Let's see if there's a way we can solve this. Is it always the latest response you'd need? I could see about adding that as a separate field in the browser script function definition. That way it wouldn't break any of the current browserscripts, but would allow you to see the latest response. I can see a lot of cases like this where there's some sort of streamed status and the completion state being in the final section.
In the meantime, one way to fix this would be to flip the order of your messages. So, your first user_output message would give the file_id (or they all could) and you could determine if the task is done not by looking for that specific message, but by checking task["completed"] as true or even task["status"] as either "error" or "completed", both of which would mean that the task is done.
In this case having the latest response would work great since I'm using it to give updates in the UI and really only care about the current one. Alternatively, if there was a way to access the additional pages that'd work too.
I've currently got my browser script checking for task status so that would work. For now I'll probably just remove the status updates and use static messages.
I'm not sure if this is a bug or a misunderstanding on my part about how the task output is intended to work.
My download task returns
user_output
with every chunk that includes the current chunk and total_chunks.{ "current_chunk": "1", "total_chunks": 17}
However, if more than 10 responses are sent, only the first ten are available to the browser script.
It took me a bit but I found that the additional responses are recorded but the browser script only sees them if you go to the next page.
In the old UI all of the responses are available as there isn't any pagination.
Is there a way via a browser script in the new UI to always get the latest response regardless of which page of output the UI is set to?
I'm using a slightly modified version of the example download browser script where I added console.log for task/response at the top.
The text was updated successfully, but these errors were encountered: