Commit Graph
3 Commits
Author SHA1 Message Date
jaap-janandClaude Opus 5 16e0051e89 Draw the tags that have been storable and invisible since the domain landed
ci / build and test (push) Successful in 1m22s
ci / android head (push) Failing after 5s
ci / api image (push) Successful in 52s
`Tag` has been a full item kind for three commits — a table, a migration, a
codec, a merge, a cipher — and `HostSecret.TagIds` has merged per tag so two
people tagging one host both keep theirs. Nothing drew a chip. The tags a
client could store were ones nothing here could see.

Chips on host rows, both heads, from names resolved through the tag list rather
than ids: a tag that does not resolve is left out rather than drawn, because it
means the tag was deleted elsewhere or belongs to a vault this session cannot
read, and a host with one chip fewer is the honest answer where a host wearing
a GUID is not. The id stays on the host, so the chip comes back if the tag does.

The picker is chips that toggle, matching the chips on the row behind it. A
list of names to tick would make the user match an entry to a chip they can see
two inches away. The box under it creates a tag and puts it on straight away,
because that is when a tag is usually wanted — while tagging a host and finding
it does not exist yet. Unlike every other field in that editor it writes to the
keychain immediately, since a host can only name an id that exists; cancelling
therefore leaves the tag behind, which is honest rather than hidden. A name
that already exists is used rather than repeated: two tags called "staging" are
storable and must stay storable, because two people creating one offline is how
it happens, but typing it into a box beside a chip of the same name is a slip.

Renaming and deleting needed a home, or the picker fills with names nobody uses
and never empties. That home is a TAGS category on the keychain screen, where
every other item kind is managed — and renaming is the whole reason a tag is an
item rather than a string repeated inside twenty payloads: it is one write, and
no host is touched. The delete confirmation counts the hosts wearing it, which
is the difference between a tidy-up and losing a filter somebody relies on.

The desktop host editor now scrolls, and that is not a tidy-up. A picker's
height is a chip per tag in the keychain, wrapped, so somebody with fifteen
tags has an editor half again as tall as somebody with three; no fixed height
holds that, and trimming other fields to buy room only moves the failure to
whoever has sixteen. The layout suite caught it the moment its seeder grew tags
— which is why the seeder now creates ten rather than three, enough to drive
the pane onto its cap so the capped shape is what gets measured rather than one
no real keychain produces. The cost is named where it is paid: the harness
skips anything inside a ScrollViewer, so from here it certifies that pane fits
the column rather than that every field in it does.

Two smaller things fell out. Five buttons overflowed the keychain header by a
few pixels, so GENERATE lost the word KEY — its tooltip carries what the word
did. And TotalItemCount had been counting keys and credentials while ALL showed
four kinds; it counts all five now, because a number under a chip that
disagrees with the rows it opens is worse than no number.

An adversarial review of this change found two defects it had introduced, both
green against the full suite. NewTag filed into the "new items go to" picker
while the tag list only ever holds the active vault's — so with a team vault
selected a tag would be created, queued for push, reported as added, and then
invisible, with no row, no count, no picker entry and nothing able to rename or
delete it, because there is no active-vault switcher to go and find it with.
The comment on the host editor's own create path states that exact rule; this
was the one place that broke it, and NewObjectStore, whose list is likewise
active-vault-only, already ignored the picker. And the tag editor was the only
one of five that did not disarm a pending deletion when it opened, so arming a
key's deletion and then pressing + TAG left a live DELETE for an item the user
was no longer looking at, directly above the boxes they were typing into. Both
are fixed, both have a test, and the first was checked against the broken
version before being kept.

The same review caught a doc comment that had been inserted between
SnippetRowViewModel's summary and its declaration, silently taking it over.

Verified by the whole suite on a clean build: 1413 tests over nineteen
projects, none failing. Both heads build. The rectangles the layout suite
cannot reach are phase 9 of docs/manual-checks.md.

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
2026-08-03 12:42:43 +02:00
jaap-janandClaude Opus 5 5593f337b6 Give the phone the second design, and both heads the palette it arrives with
The Android v2 design is what this head draws now: four destinations in a bottom bar — Hosts,
Terminal, Keychain, More — with snippets, SFTP, S3, logs and preferences one tap deeper behind the
last. The first design's four had nothing behind them, which is what made a hub worth building.

The palette moved from green-black to blue-black, and it moved in the shared project because that is
where it lives and the desktop v2 specifies the same seventeen tokens. One colour changed meaning
rather than value, and it is the only semantic change in the file. Green used to *be* the accent, so
Ellipse.dot.live filled with Accent and "the thing to press" and "a shell is open on this host" were
the same colour by construction. v2 makes the accent blue and keeps a green for status alone, which
finally separates them: Live is that green and nothing merely interactive may use it. The accent is
also two colours now — Accent fills, AccentText writes — because a row of chips in the fill colour is
a row of things that all look like the primary action.

A palette is not one file, which is the part worth knowing before the next one. Nine hex literals
lived outside it: the nav bar's own label colours, the accessory keys and their Ctrl-latched state,
two scrims, the window background Android paints before Avalonia has a frame, and the launcher
vector. The two C# sites now resolve from the dictionary by name rather than restating it. The
renderer's page cannot — it is served to a WebView over a loopback socket — so terminal.css and
terminal.js keep hand-copied values and say so at both sites.

ShellScreen gained More and Buckets, appended rather than slotted in. SFTP and S3 are one screen over
one TransfersViewModel differing only in which picker they offer, and the kind is set by the button
that navigates rather than on arrival — doing it in OnScreenChanged made every arrival at Transfers
force the picker back to hosts, including the desktop's own rail arriving at a screen with a bucket
already open. It refuses to change kind while a session is live, because there is one session behind
both destinations and switching under it would title a screen S3 while it listed an SFTP host.

What the design draws and this does not, on the usual grounds. The FORWARDING screen: nothing here
forwards anything, so every toggle would be a control with no effect — it is a paragraph on the hub
naming the absence, for the reason the desktop keeps TEAMS in its rail. The terminal's `23 ms · fwd
5432`. An ED25519 badge and a SHA256 line on keychain cards, which need an algorithm field and a
fingerprint the item type does not have. An `agent` chip, for an agent that does not exist. Snippet
run history and exit codes. The Logs FOLLOW pill, which claims a live tail over records that are
written once at close and read when the screen opens, and the severity filter, which has nothing to
count — that chip row is spent on the real choice, which of the two logs. S3 bucket totals and
lifecycle. And the + on HOSTS, which would open a host editor this head has not got.

SFTP is browse, open and delete. Both transfer commands work, and what they work against is the local
pane: QueueDownloads writes to Path.Combine(LocalPath, name), and LocalPath starts at
SpecialFolder.UserProfile, which on Android is the application's own private directory. A download
would have reported success and left the file where the person who asked for it cannot open it, which
is worse than not offering it — a refusal is visible and a file in /data/user/0/ is not. The queue is
not drawn either, since nothing here can put anything in it. Both return with the document picker.
The foreground service still counts zero transfers, and the reason moved rather than went away.

Four defects worth naming, because three of them are the kind that compile. A Button as a ListBox
ItemTemplate swallows the pointer press before the list sees it, so the files listing selected
nothing and every command reading the selection did nothing — the row is a Border now and the
phone-only single-tap-to-open is a Tapped handler, which also keeps a desktop single click from
walking into directories. Avalonia type selectors are exact, so TextBlock.fingerprint never matched
SelectableTextBlock and every fingerprint on this head rendered proportional and unwrapped: that was
breaking the never-truncated rule on the host-key sheet already. The new two-level hierarchy had no
handler for the system back gesture, so back left the application from a log screen. And the tab's
close cross had shrunk to a 30x32 target flush against the select target, which is the one control
here that ends a shell with no confirmation and no undo.

Fingerprint unlock is raised on arriving at the lock screen rather than waiting for its button, which
is still there. Only at launch: a lock the user asked for is not answered with an immediate request
to unlock, which makes LOCK look inert and trains the reflex of authenticating at a prompt nobody
asked for. And once, because a declined gesture leaves the passphrase box exactly where it was and a
prompt that came back after being dismissed would be a modal you cannot get out of to type into it.

Two fixes fall on the desktop. Its file listing coloured directories with Info and executables with
Accent, which was blue against green and is now two steps of one blue; an executable is Live now.
And a bucket's folders were drawn with a 0001-01-01 timestamp, because a prefix has no modification
time — blank now, for the reason a directory's size is blank.

Verified by the whole suite: 1309 tests over nineteen projects, none failing, including the layout
suite that stands up real Avalonia and parses every desktop screen. Both heads build. Not verified on
a device — nothing in this head ever has been; see docs/android-port.md.

Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01AZE3u99BNt6LzgTC5jhbz2
2026-08-02 18:23:53 +02:00
jaap-jan 7a3a521c59 Give the phone the rest of its screens, and a way in
ci / build and test (push) Failing after 2s
ci / android head (push) Failing after 1s
All seven screens of the design, plus the two it does not draw because it starts at an
enrolled phone: naming a server, and choosing a passphrase.

The five states docs/android-port.md worried about losing at 360dp are all here and none
of them softened. The changed-key refusal is a full-screen panel rather than a bottom
sheet, because a sheet is swipe-to-dismiss by convention and that screen must have no way
forward. The recovery code raises FLAG_SECURE for its own state and lowers it afterwards,
so the sentence about screenshots is true rather than decorative. The delete
confirmations keep their counts and replace the row in place.

Signing in works, and the seam it needed is worth more than the implementation:
IAuthorizationCallback now sits between OidcClient and the loopback listener, so the two
heads differ in where the response arrives and in nothing else. PKCE, the state check,
discovery, the token exchange and the key binding stay one implementation — a second OIDC
client would be a second place for a security bug to live. The phone registers a
private-use scheme with the system rather than binding a loopback port, which on a shared
device any other app can do first.

The accessory key row needed TerminalWorkspace.SendInputAsync: ordinary typing goes from
the renderer straight down the socket, and there was no way in for the keys a software
keyboard does not have. Ctrl latches, because one thumb cannot chord, and the latch is
drawn — a modifier that is on and does not look on is how somebody sends ^L to a database
prompt believing they typed an l.

597 client tests green, including two new ones for the input path and one for the
terminal surface command. Nothing has run on a device.
2026-07-31 21:43:11 +02:00