Add open_url/close_url notebook APIs
#10602
Conversation
|
Web viewer built successfully. If applicable, you should also test it:
Note: This comment is updated whenever you push a commit. |
There was a problem hiding this comment.
Can you add the workflow in the PR description. I see it now
If we do
url_a = <>
url_b = <>
viewer = Viewer(<>)
viewer.open_url(url_a)
# wait to show a
time.sleep(<>)
viewer.open_url(url_b)
# wait to show b
viewer.open_url(url_a)Does this do the right thing by showing a, then b, then a? I don't know the next level deeper enough to know how opening the same url twice behaves. If we can have a bunch of things open my first attempt in updating my code would be to open ALL my urls, then call open a second time to bring a preloaded url to the main view.
There are no uniqueness constraints on data sources, so this will open A, then B, then A again, resulting in 3 open recordings. It will fetch A's data twice.
We don't have a way to activate a recording by its URL. If you want the exact same behavior as before, you'd do viewer = Viewer()
current_url = None
# ...
current_url = dataset.partition_url(...)
viewer.open_url(current_url)
# ... later
viewer.close_url(current_url)
current_url = dataset.partition_url(...)
viewer.open_url(current_url)But there should be no harm in leaving the older partitions open, no? So you'd have this instead: viewer = Viewer()
viewer.open_url(dataset.partition_url(...))
# later, open another one
viewer.open_url(dataset.partition_url(...)) |
cc @ntjohnson1 as this replaces
set_active_partition_url- do we have to update some notebooks?Also reworked the
_data_queue+_table_queueinto a single_event_queuewhich all calls tosendgo through now. This means not only rrd/table data, but all events will be buffered in the same way, and in the right order relative to each other.