Changed the UI to better match the different abstractions - #6
Merged
Conversation
This was referenced Aug 19, 2026
Member
|
Naming note for the new UI: the reserved role names are |
devel brings the merged stack and the phase-1 role rename (#9). Per the rule in #7, every conflict resolves toward `inference`/`learning`: the pool cards keep their new table+graph body and say 'Pool: inference' / 'Pool: learning', `poolHistory` and the lane colour checks follow, and the page version supersedes both sides as 0.7.0. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Three gaps from the first live run of this UI:
* the header counted `w.tasks.size`, and live tasks are ephemeral --
created, terminal within a second, dropped after the fade -- so it
read `0 tasks` while the pool graphs pulsed. It now counts a
cumulative `taskTotal`.
* the recent-tasks tables filtered the same ephemeral map and never
populated. A rolling `recent` list (the task objects themselves,
capped at 64) feeds them now, so a finished task keeps its row. The
columns changed with it: DT / Task / State / Age -- what the wire
actually carries. The old Kind / Agent / Investigator columns were
faked from a uid hash, which reads as data on a live page; component
attribution returns when the service sends per-task metadata (#8).
* the 'visible to broker' subtitle drew into the Channels title at
narrow widths; it now yields when there is no room.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Member
|
Pushed two commits on top of the rework:
Verified against a live broker with two twins (screenshot in the DT session log): pool graphs show the sensor cadence and training windows, tables populate, header counts. Remaining for #8: per-task component attribution, and the learner card has no convergence/metrics display in this design yet. |
The card layout dropped whatever no longer fit below an expanded card (`+N more`, easy to miss), so on a small window expanding the first twin hid the second entirely. Every twin now keeps at least its collapsed strip: an expanded card is clamped to the height that leaves the strips below it room, and its content clips inside the card. The demo-era hard cap of two cards goes up to twelve while at it; 0.7.1. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
The pane was the one part of the page that is a variable-height nested
tree with per-node collapse -- which a browser already knows how to lay
out, scroll and hit-test. It is now a scrollable <div> positioned over
the canvas frame; the canvas keeps everything else (header, channels,
pools, arcs) and draws only the pane's frame and empty-state.
* ~320 lines of hand-rolled card/agent/sparkline drawing and
hit-testing replaced by markup + CSS; sparklines are inline SVG fed
from the same fake-data functions (still placeholder until #8).
* the visibility contract survives: the pane publishes each visible
card's rect onto `tw._rect` in canvas coordinates, null when
scrolled out -- same meaning it had when the layout dropped cards.
* constraints held: no jumping (scrollTop survives rebuilds, plain
birth order, gone twins fade in place), no auto-scroll on a new
twin, newly-seen twins still start collapsed.
* full twin id rides the title tooltip; state pill, initializing
pulse, stream-pulse ring and the closed marker all carried over.
Verified against a live broker with four twins across two sessions at a
1280x620 viewport: expand, scroll to the last strip, collapse -- and the
recording tests still pass (12). VERSION 0.8.0.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
* 'collapse all' / 'expand all' chip in the pane header (twin level);
an 'agents' chip per expanded card toggles that twin's agent blocks
at once. Agents now start collapsed, like the twins themselves.
* scroll cues: a thin styled scrollbar plus a subtle edge gradient at
whichever end has more content -- shown only when there is any.
* 'Endpoint Pools' is 'Resource Pools', and each pool title is a
link to the task dispatcher's surface (`.../task_dispatcher/pools`;
its Explorer page once one exists) -- the pools are its, after all.
Verified live: four twins, bulk toggles both directions, cues appear
only on overflow, the pool link lands on the dispatcher listing.
VERSION 0.8.1.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
The all-twins chip reads `twins ▸/▾`, same as the per-card agents chip. Native scrollbars are overlay-style on current desktops and hide themselves, so the pane hides them for real and draws its own 3px position thumb -- always present when there is overflow, wider on hover, draggable. VERSION 0.8.2. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Moved into the gutter between the cards and the Digital Twins frame, and brightened -- flush on the cards' right edge it read as part of their border. VERSION 0.8.3. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
The twin pane is DOM: scrollable, no twin can vanish
Member
|
#10 merged in per Andre: the twin pane is now a scrollable DOM overlay (expand/collapse per twin and per agent, bulk toggles, scroll cues + position thumb), pane title 'Digital Twins', pools renamed 'Resource Pools' with titles linking to the task dispatcher surface. Page version 0.8.3, verified against a live broker throughout (four twins, two sessions). This PR is the single review unit now; merges to devel after your look. |
The link target is the new Explorer deep link (radical.orbit#116, `#plugin/broker/task_dispatcher`): inside the Explorer the hash change navigates in place, from the standalone page it is a redirect. No more raw JSON landings. VERSION 0.8.4. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
I modified the UI to better capture the multiple digital twins, agents, tasks, and so on. It captures the changes that #1 will bring (using pools).
See for yourself here: https://codingcando.com/research/digital.twins/index.html
Note: the UI right now is using completely fake data. The plugin will need to be updated to reflect this new UI.