Conversation with Mia

0 replies · 7 marks on Screenwalk: 6 answered by Mia, 1 resolved · 1 mark on ScreenwalkRecords: answered by Mia

William's takes are recordings, and recordings stay outside the room for readers who do not write in it. The replies Mia voiced are shown.

Guillaume marked Witnessing on Screenwalk2026-10-20 03:28 UTC

Wouldn't that be one more state machine in betweenWhen it hasn't been actually added to the system

Guillaume2026-10-20 03:29 UTC
Are you replying to me here
Mia2026-10-20 03:33 UTC
Yes, I answer here now. Until tonight no agent listened for marks, so your marks waited; I am wiring the episode's Mia seat to wake on them like a spoken take. Your question: yes. While Witnessing, the witness reads each named session one by one, and that has its own life (named, read, cited in the closing) that the system does not hold yet: WitnessedSession is in the data drawing but has no machine. That would be the machine in between.

Guillaume marked an area on Screenwalk2026-10-20 03:30 UTC

M,I,,,ilihghilijl9

Mia2026-10-20 03:33 UTC
This one reads as an accidental entry (M,I,,,ilihghilijl9). I resolve it; reopen it if it meant something.

Guillaume marked NameSessions on Screenwalk2026-10-20 03:31 UTC

Is it the moment where I get published on YouTube?

Mia2026-10-20 03:33 UTC
Not yet. NameSessions is earlier: while recording, you name the agent sessions the witness should read, and the walk goes from Recording to Witnessing. YouTube is Upload, after StopRecording: Recorded to Published. I light it on the board.

Guillaume marked NotStarted, Recording, Witnessing and 3 more on Screenwalk2026-10-20 03:32 UTC

That part of the work is where we want to get ready with the observation of water correct on the screen and also the internal use out how we are coming to use this

Mia2026-10-20 03:33 UTC
Agreed: NotStarted, Recording and Witnessing are the live part, before and while you record. The machine has no step yet for checking that what is on the screen is right before StartRecording; a Preparing state before Recording would hold it (the screen, the sessions, the episode chosen). The internal use is the other end: Chronicled and InConversation, where we come back to ask about it.

Guillaume marked Upload and Published on Screenwalk2026-10-20 03:39 UTC

Yes the recording stops then I click the recording to upload it into YouTube and I start adding on these are very boring element such as the title description is it public in which list doesn't go in which location is it locatedIs the community set community member for a week to contribute and check box to AI generated content can you imagineVery boring transition

Mia2026-10-20 03:39 UTC
Right: the machine draws Upload as one arrow, and on YouTube it is a whole form: title, description, visibility, playlist, location, who may contribute, the AI-generated box. Most of it is already known before you upload: the witness closing and the episode say what the recording is. So that boring transition could become a state of its own, Uploading, where the details arrive drafted from the closing and you only check them and publish. In the data drawing it would be one more record beside Review: the upload details (title, description, visibility, playlist, AI-generated). Do you want me to draw both?

Guillaume marked Recorded, Upload and Published on Screenwalk2026-10-20 03:40 UTC

There's a step because adding on the video and play some manual operation some notes taken

Mia2026-10-20 03:41 UTC
Drawn. Recorded now goes by Upload to a new state, Uploading, and Publish takes it to Published. Uploading holds the manual work you describe (title, description, visibility, playlist, location, contributors, the AI-generated box, your notes), and its prompt asks an agent to draft those details from the witness closing so you only check them. The machine validates.

Guillaume marked Recorded, Upload, Uploading and 2 more on Screenwalk2026-10-20 03:42 UTC

OK then the review documentCould also be in multiple states I suppose

Mia2026-10-20 03:42 UTC
Yes. Here the review is one arrow, CreateReview, and it lives a life of its own: requested from the YouTube address as an internal source, transcribed, a first version, new versions when you ask, then the academic fields (step 2) and the internal usage (step 3), then held in the episode. That is a second machine, Review, beside Screenwalk in this system, tied by the Review record already in the data drawing; Screenwalk would only wait at Reviewed until it is ready. Shall I draw it?

Guillaume marked Screenwalk, WitnessClosing and Review on ScreenwalkRecords2026-10-20 03:43 UTC

What does the whole diagramRepresenting by the way

Mia2026-10-20 03:44 UTC
It is what a screenwalk leaves behind: the records kept about one. An episode holds its screenwalks; each keeps when it was recorded, its YouTube address and where it stands; it witnesses the agent sessions you name; it ends with versions of the witness closing, each with your correction; it is read as one review and teased by shorts. The machine says what happens; this drawing says what is kept. I am playing a short walk through the three drawings now.