aboutsummaryrefslogtreecommitdiffstats
path: root/src
Commit message (Collapse)AuthorAgeFilesLines
* A post's run of pictures goes into one grid instead of a screen each. "on a ↵Adam Soltys62 min.2-0/+34
| | | | | | | note with lots of images like this could we put them into a masonry grid layout within the post kinda like we do on our images feed so they don't take up so much real estate as we scroll?" (Adam, 2026-11-18). Two or more pictures with only blank space between them now share one box of justified rows, laid out like the Images feed: 140 px rows, 100 px on a phone, each tile at its own proportions and cropped to fill. A thirteen-photo post that ran thirteen screens long now takes a few rows. A tap still opens the picture in the pager, with the post's others a swipe away. A lone picture is shown as before. feed-picture-pager-test (two pictures side by side at one height, a lone one not in a grid, pager unchanged), feed-media-views, compose-preview and features-smoke pass. Co-Authored-By: Claude Opus 5.5 <[email protected]> Claude-Session: https://claude.ai/code/session_01VMan6TV6sjEZSpqqBen9T5
* PAFF songs and albums play in a bar that stays while you go about the app. ↵Adam Soltys11 hours3-1/+132
| | | | | | | | | | | | | "Can you work with PAFF again to make songs embed nicely in nostr posts?" … "From the start please" (Adam, 2026-11-17). A paff.club song or album link in a post (/song/<id>, /album/<id>, also under /play/) is a slim card: cover art, "Song · Artist" and its line from the page's preview, a ▶ and a ↗ out to PAFF. The tap doesn't play it in the post. PAFF's 64 px bar player (/embed/song|album/<id>?bar=1, the link's query kept) loads into a music bar pinned above the bottom nav, or under the feed column on desktop. The bar lives in the document body outside render(), so the feed's windowing never drops it: the music carries on through scrolling, threads and other tabs. The bar is never moved, because a moved iframe reloads. Another song changes its src, and × removes it. The card shows "Playing below" with an equaliser while its song is in the bar. The Post button, the new-posts pill and the page's bottom padding move up while the bar is open. link-preview-test (stand-in paff.club): the song card has its cover and title with no frame; a tap opens the bar above the nav at /embed/song/abc123?bar=1; the bar is still there on the Wallet page; × closes it. Checked against the real paff.club: the bar player for Daft Punk's Harder, Better, Faster, Stronger played in a frame and answered a pause command. features-smoke and feed-when pass. Co-Authored-By: Claude Opus 5.5 <[email protected]> Claude-Session: https://claude.ai/code/session_01VMan6SwkivduFDDkMmnXjZ
* A PAFF film pauses when its post scrolls off screen and carries on when it ↵Adam Soltys14 hours1-1/+38
| | | | | | | comes back. "can you ask paff to do the pause command?" (Adam, 2026-11-17). PAFF's embed now takes {source:'paff', command:'pause'|'play'|…} and reports {source:'paff', event:'playing'|'pause'|'ended'}. Each loaded film box is watched: once it's fully off screen while playing, it's told to pause. When it's back at least half in view, it's told to play, but only if we paused it; a viewer's own pause stays paused. link-preview-test runs a stand-in paff.club (self-signed TLS, host-resolver-rules) whose player speaks the protocol. In it the film pauses when scrolled away and plays when scrolled back (commands logged: pause, play). features-smoke passes. Co-Authored-By: Claude Opus 5.5 <[email protected]> Claude-Session: https://claude.ai/code/session_01VMan6SwkivduFDDkMmnXjZ
* A PAFF film link plays inside the post. "What would it take to embed a PAFF ↵Adam Soltys14 hours3-0/+71
| | | | | | | video in a Nostr feed, or to have it play in line like other videos and YouTube videos do?" … "Yes please" (Adam, 2026-11-17). PAFF now serves a bare player at paff.club/embed/movie/<slug> (and /embed/show/<slug>/sNeM), and its query string carries over from the share link (q quality, c copy, t start). A paff.club movie or episode link, with or without /play, shows as a 16:9 box: its backdrop and title from the page's own preview, a play button and a ↗ out to PAFF. The tap swaps in PAFF's player with autoplay, fullscreen and picture-in-picture allowed, and no repaint touches the box after that. It loads only on a tap, never on scroll, because a film starts a torrent swarm in the viewer's browser. link-preview-test (a film link is a player with backdrop and title, no frame until the tap, and the tap loads /embed/movie/1382-hackers-1995?q=1080p&t=4910), compose-preview and features-smoke pass. Co-Authored-By: Claude Opus 5.5 <[email protected]> Claude-Session: https://claude.ai/code/session_01VMan6SwkivduFDDkMmnXjZ
* The scroll date chip blends between posts and goes the moment you stop. "I'm ↵Adam Soltys16 hours2-9/+21
| | | | | | | good with blending. Can you make the chip appear and disappear faster? I want it gone as soon as I'm stopped to look at something" (Adam, 2026-11-17). The time now moves between the post at the top of the screen and the next one, in step with how far through that post you've scrolled. The clock rolls with the scroll instead of jumping from post to post, and runs quickly through a gap between posts. The chip leaves 250 ms after the last scroll event (was 1.2 s); a coast keeps sending events, so it stays while the page is still moving. It fades in over 80 ms and out over 120 ms. feed-when-test (blended time between the two posts, rolls with a few px of scroll, gone half a second after stopping) and feed-new-posts pass. Co-Authored-By: Claude Opus 5.5 <[email protected]> Claude-Session: https://claude.ai/code/session_01VMan6SwkivduFDDkMmnXjZ
* Scrolling the feed shows when you are. "When I'm scrolling the feed, could ↵Adam Soltys17 hours2-0/+72
| | | | | | | we have some sort of little counter that shows the date and time of posts where I'm in the timeline, so I can kind of roughly gauge how far back I've scrolled in time?" (Adam, 2026-11-17). Past the first screenful, a small chip under the top bar shows the date and time of the post at the top of the screen, and how long ago it was (Mon 16 Nov · 09:57 · yesterday). It updates as the feed or a profile's posts move and fades 1.2 s after scrolling stops. It is drawn straight into the page each frame, outside render(). New feed-when-test passes, along with feed-new-posts and features-smoke. feed-stability fails at the same step with and without this change. Co-Authored-By: Claude Opus 5.5 <[email protected]> Claude-Session: https://claude.ai/code/session_01VMan6SwkivduFDDkMmnXjZ
* A draft's Preview shows a link as its card. "i added metadata to paff pages, ↵Adam Soltys20 hours1-0/+13
| | | | | | | was hoping links to movies would render with a nicely formatted thumbnail" (Adam, 2026-11-17). The registrar already reads paff.club's movie pages (title "Hackers (1995) · from 1:21:50", description, backdrop), and a posted note warms its links' cards before it shows; the composer's Preview ran the same renderer but never asked for the page, so a link not already cached stayed a bare URL. It now asks once the typing settles (half a second), then repaints into the card. link-preview-test (a draft's preview shows the movie card, asked once) and compose-preview-test pass. Co-Authored-By: Claude Opus 5.5 <[email protected]> Claude-Session: https://claude.ai/code/session_01VMan6SwkivduFDDkMmnXjZ
* People search looks among the people you follow first. "derek ross leaked ↵Adam Soltys20 hours2-2/+50
| | | | | | | his key and started a new npub. i tried searching him up and found this … it says no posts" … "Did we get our search to bring up his new account … as the top result?" (Adam, 2026-11-17). Search for "derek ross" came up with an empty impostor key: Primal's user search (our main source) doesn't index his new key yet and ranks by its own trust score, and only one NIP-50 relay has it, 34th. Adam already follows the new key. Now a follow whose name or NIP-05 matches the query ranks with coinos usernames, above the indexes; the chat feature hands its follow list and the profiles it holds to the search, and fetches follows' profiles it has never seen once a session, the first time a search asks. Rows found while those are still loading aren't cached, so the next search sees them; older cached results are dropped (search cache v3). New search-follows-test (follow first, impostor still listed after, renamed old key not matched, NIP-05 match, partial answers not cached) passes, with search-scope, search-avatar-loading and features-smoke. Co-Authored-By: Claude Opus 5.5 <[email protected]> Claude-Session: https://claude.ai/code/session_01VMan6SwkivduFDDkMmnXjZ
* The balance breakdown folds away under the total. "let's hide the breakdown ↵Adam Soltys2 days4-8/+30
| | | | | | | by default but give some way to expand to see it" (Adam, 2026-11-16). On the one balance card, the Ark / On-chain split and the Gifts line now sit behind a small "Breakdown ▾" under the total (only when there is a breakdown to show); "Hide breakdown ▴" folds it again, and the choice is remembered per device. Lines about money on its way (Pending, Moving) still always show. The desktop side panel follows the same choice. signup-quick-test (folded by default, opens to the Gifts line, remembered), unified-balance and features-smoke pass. Co-Authored-By: Claude Opus 5.5 <[email protected]> Claude-Session: https://claude.ai/code/session_01VMan6SwkivduFDDkMmnXjZ
* Balance lines say Ark, On-chain and Gifts. "let's make the balance titles ↵Adam Soltys2 days2-2/+3
| | | | | | | "Ark" "On-chain" and "Gifts" rather than Instant or Locked in gifts" (Adam, 2026-11-16). The lines under the one balance's total now read Ark (was Instant), On-chain, and Gifts (was Locked in gifts). The gift list keeps "Locked in gifts" as an unclaimed gift's status. unified-balance, signup-quick and gift-build tests pass. Co-Authored-By: Claude Opus 5.5 <[email protected]> Claude-Session: https://claude.ai/code/session_01VMan6SwkivduFDDkMmnXjZ
* One balance: Spending and Savings on the same seed are one account. "Do you ↵Adam Soltys3 days4-34/+185
| | | | | | | think we could combine the ARC and savings Bitcoin accounts into a single account? So if someone receives into a on-chain Bitcoin address, would hold it there, but they'd have a combined history, and it would select which coins to spend based on where you're paying out of, but the balance would be unified" … "Yeah go for it" (Adam, 2026-11-15). Step 1 of two: the wallet shows one balance card (the total, Ark and on-chain, with money mid-board still counted; an Instant / On-chain split appears only when there's some in both). History is one timeline, and a move between the two halves is one row priced at its cost: a board's on-chain Sent and its Boarded become "Made instant" (or "Making instant" while it confirms), and an offboard to our own address becomes "Moved on-chain". An offboard to someone else reads Sent. Receive leads with the payment address, with a Bitcoin address button for on-chain payers; the address stays on-chain and is all that's offered offline. Payments start from the Ark coins, and an on-chain address is paid from on-chain coins when there are any, leaving Ark only when there aren't. Accounts lists one row, Settings → Wallet has one balance row (Spending's coin settings plus the public key), and a per-device "Show Spending and Savings separately" brings the two-card carousel back as it was. Sign-ups are now this one account instead of Spending-only with Savings added later. Older Spending-only accounts are converted the same way, and the Add savings row only shows when the balances are shown apart. Not yet (step 2): covering a shortfall by boarding or offboarding automatically, and the "available now" split on the amount screen. New unified-balance-test (folded rows, their costs, the separate Spending timeline, on-chain-first routing) passes, along with signup-quick (one balance, Bitcoin address on Receive, separate toggle and back), balance-settings-row and send-enter-keys (both now opt into separate; send-enter-keys also enters through the seed-phrase door since Create became a sign-up), features-smoke, onboard-backup, gift-profile-setup, footer-support, pos, chat-layout, account-delete, lock-wallet, padlock-flow, profile-actions, gift-build, ark-ln-savings-small, carousel-flick, amber-dm-upgrade, send-lnaddr-advance, account-switch-leak and ark-offline-init. feed-choice-per-wallet fails at its second wallet's import, and fails the same way on 5554c70. Co-Authored-By: Claude Opus 5.5 <[email protected]> Claude-Session: https://claude.ai/code/session_01VMan6SwkivduFDDkMmnXjZ
* One door to a signer app: the Android-intent "Sign in with a signer app" ↵Adam Soltys3 days1-7/+6
| | | | | | | button is gone. "Are the two signer buttons redundant?" … "Yeah remove" (Adam, 2026-11-15). On Android both opened Amber: Connect a signer app over NIP-46 (link or QR, any signer, any device, reads messages), and the NIP-55 intent, which needed no relay but made a login that couldn't decrypt DMs (Messages already sent those users back through Connect) and a trip through the app per signature. Only Connect is offered now, at sign-in, reconnect and Ways back in. Accounts that came in by intent keep working: their sessions still resume, their pending requests still show the waiting card, and amber-dm-upgrade-test (their move to Connect) passes. amber-login-test drove the removed button and is retired; features-smoke, nostr-login, onboard-backup and signup-quick pass Co-Authored-By: Claude Opus 5.5 <[email protected]> Claude-Session: https://claude.ai/code/session_01VMan6SwkivduFDDkMmnXjZ
* Ask for a way back in once, and only when there's money to lose. "Let's only ↵Adam Soltys3 days2-3/+39
| | | | | | | prompt to add a way back in after there's some balance and only if we haven't prompted before" (Adam, 2026-11-15). A sign-up's wallet no longer carries the reminder line from its first minute. The first time it holds anything (Spending, Savings or an incoming payment) and has no way back in yet, it shows one notice: 'You have bitcoin in here now. Add a way back into your account, in case you lose this device.' with Add one (→ Settings → Wallet) and Not now. That's recorded on the account as soon as it shows (directory and synced snapshot too), so it stays for that visit and never comes back, here or on another device. Seed wallets opened by sign-in keep their old phrase reminder. signup-quick-test: nothing on an empty account; given a Spending balance it asks; Not now; not again after a reload Co-Authored-By: Claude Opus 5.5 <[email protected]> Claude-Session: https://claude.ai/code/session_01VMan6SwkivduFDDkMmnXjZ
* Receive → Amount: ask for a sum with the point of sale's pad. "Add an ↵Adam Soltys3 days4-14/+29
| | | | | | | 'Amount' button to receive that basically brings up the PoS screen" (Adam, 2026-11-15). Under the payment address on Spending's Receive (and under the Ark address where there's no name), Amount opens the till's amount screen as "Request an amount": the same pad, a What it's for note, Create invoice → the QR and amount → paid, then Done back to Receive. It leaves the till out: no tip prompt, no tip settings, no day's sales, and its invoices are marked so the till's sales list doesn't count them. Also: the HISTORY heading under Receive/Send no longer stands over an empty wallet (the last change hid the 'No transactions yet' card but not its heading). pos-test (request mode: no tips, straight to the invoice, Done, kept out of the sales) and signup-quick (Amount on Receive, opens the pad, Back; no History heading) pass, with footer-support and features-smoke Co-Authored-By: Claude Opus 5.5 <[email protected]> Claude-Session: https://claude.ai/code/session_01VMan6SwkivduFDDkMmnXjZ
* No footer once a wallet is open, no empty history, no default name on a lone ↵Adam Soltys3 days2-3/+12
| | | | | | | balance. "Actually those can go into settings too. And let's not show an empty history section just keep it hidden until we have history. I don't think we need the spending account name by default let's leave it blank unless the user supplies one" (Adam, 2026-11-15). The language and currency pickers now sit at the top of the Settings hub; the footer under every page is gone while a wallet is open (a visitor reading the public feed, with no wallet and so no Settings, still has them there; the wizard keeps its language picker). The wallet's history shows nothing at all until there's a first payment, instead of a 'No transactions yet' card. A single-balance account (a sign-up's Spending) shows no SPENDING over its amount unless its owner named it; the two-face carousel keeps Spending/Savings. footer-support (no footer with a wallet, pickers in Settings, About), signup-quick (no name on the lone face), chat-layout, onboard-backup, balance-settings-row, feed-window-dom and features-smoke pass Co-Authored-By: Claude Opus 5.5 <[email protected]> Claude-Session: https://claude.ai/code/session_01VMan6SwkivduFDDkMmnXjZ
* The footer is just the language and currency; everything else moved to ↵Adam Soltys3 days2-40/+35
| | | | | | | Settings → About. "Let's move all the footer stuff to an about page or something except for the language and currency toggles" (Adam, 2026-11-15). A new About tile (an info mark) on the Settings hub opens a page with the credits (made by Adam Soltys, source on git.coinos.io), How it works, Show your support (once a wallet is open), Install app (when the browser offers it), the Dark/Light mode switch and the running commit. footer-support-test now checks the footer holds only the two pickers and drives About (credits, theme switch, support → hat shop); onboard-backup, chat-layout, features-smoke and signup-quick pass Co-Authored-By: Claude Opus 5.5 <[email protected]> Claude-Session: https://claude.ai/code/session_01VMan6SwkivduFDDkMmnXjZ
* Balance settings rows: each balance in its own tinted tile, with room around ↵Adam Soltys3 days2-22/+28
| | | | | | | it. "Icons and padding need work here" (Adam, 2026-11-15, phone screenshot of Settings → Wallet): a gear sat against the left border of the Spending row and a small grey coin beside Add a savings account. Now Spending wears the purple bolt and Savings (and Add a savings account) the orange bitcoin disc, each in a 38px rounded tile, the rows padded 14px 16px with a 14px radius and 8px apart, the › or + in a fixed column on the right; the hover tint only where there is a mouse. signup-quick-test (with a dark-mode screenshot hook), balance-settings-row and features-smoke pass Co-Authored-By: Claude Opus 5.5 <[email protected]> Claude-Session: https://claude.ai/code/session_01VMan6SwkivduFDDkMmnXjZ
* Ways back in: Passkey auth, Google auth, Nostr auth, Pen and paper. "Let's ↵Adam Soltys3 days3-14/+23
| | | | | | | call recovery phrase 'Pen and paper' with a suitable icon and no need to show it twice just make it similar to the other buttons. Instead of saying Add... for everything let's try 'Passkey auth', Google auth, Nostr auth" (Adam, 2026-11-15). The card's buttons drop Add…; the recovery phrase is the fourth button (a pen over a sheet, ✓ once the words have been seen) instead of a row with Show, and the separate Recovery phrase card under it appears only when Pen and paper asks for it (scrolled to, words masked first). Accounts without the card (watch-only, imported keys) keep the phrase card as before. signup-quick-test, onboard-backup, balance-settings-row and features-smoke pass Co-Authored-By: Claude Opus 5.5 <[email protected]> Claude-Session: https://claude.ai/code/session_01VMan6SwkivduFDDkMmnXjZ
* The "You're all set" page has no Back. "I don't think we need an on screen ↵Adam Soltys3 days1-1/+2
| | | | | | | back button here" (Adam, 2026-11-15, phone screenshot): the name and the look are done by then, and both change anytime from the profile; Enter coinos is the one way on. signup-quick-test and onboard-cover pass Co-Authored-By: Claude Opus 5.5 <[email protected]> Claude-Session: https://claude.ai/code/session_01VMan6SwkivduFDDkMmnXjZ
* Signing up is a name and a look: no seed phrase to copy or check, no Savings ↵Adam Soltys3 days6-27/+354
| | | | | | | until you want it, and ways back in you add afterwards. "i think i'd like to move away from requiring people to backup and verify a seed phrase to sign up. we should just have them claim a name and pick their avatar/cover and they should be in. we should also not create a savings account by default but let them add it later if they want to. adding additional auth methods like passkey, nostr, or google should be possible once they've created their account so that they can get back into it with one or all of those methods or by seed phrase." (Adam, 2026-11-15). Where a name can be claimed (mainnet, mutinynet, with Ark and the registrar), the front door's "Create an account" makes the account at once (seed saved to the device, never shown) and goes Claim your name (no Not now) → Choose your look → in. The account is Spending alone; Settings → Wallet → Balance settings has "Add a savings account" (same seed, account 0), and on-chain money or history turning up adds it by itself. Settings → Wallet opens with "Ways back in": add a passkey, Google or a Nostr account, each of which gets a copy of the seed encrypted to its own key (kind 30078, marked recovery), and signing in with it on another device opens this account as itself, no identity adopted; the recovery phrase sits in the same list with Show. Until one exists the wallet carries a quiet "Add a way back into your account" line. The shape and the list survive the session (directory) and sync between devices (snapshot). Regtest and builds without Ark keep the classic phrase flow, and so does the /migrate door's Create. tools/signup-quick-test.js (mainnet, mocked registrar, virtual passkey): the whole sign-up, Add savings, a reload, then a wiped page signs back in with the passkey to the same account and identity. balance-settings-row, send-lnaddr-advance and onboard-backup reach their seed wallets through the phrase door or on regtest; features-smoke, onboard-cover, onboard-migrate-door, nostr-login, gift-profile-setup, passkey-lock and seed-cache pass Co-Authored-By: Claude Opus 5.5 <[email protected]> Claude-Session: https://claude.ai/code/session_01VMan6SwkivduFDDkMmnXjZ
* Chat bubbles start at the left edge, and zaps sit quietly in them. "Let's ↵Adam Soltys3 days2-6/+12
| | | | | | | remove the border and BG for zaps. Also remove the zap icon at the right of every message or adds too much clutter we'll just have it in the click menu or tapping an existing zap count can zap too. Also there's wasted space on the left of each message bubble can we make them flush to the left side?" (Adam, 2026-11-15, phone screenshot of the Coinos community). In a group chat the avatar now sits beside the name and the bubble runs from the left edge under both, not indented a whole avatar column; a zap total in a bubble is just the amber bolt and sats, without the tinted pill or the ring it had once you'd zapped; the one-tap ⚡ beside every bubble is gone from group chats and DMs (Zap is in each message's tap menu, and tapping a zap total still zaps). Live-stream chat keeps its ⚡. chat-layout, gif-picker and features-smoke pass Co-Authored-By: Claude Opus 5.5 <[email protected]> Claude-Session: https://claude.ai/code/session_01VMan6UKGJyuGyZFxddYRxS
* The clip-moment toggle reads when it's on, in dark mode too. "The contrast ↵Adam Soltys3 days1-1/+3
| | | | | | | here is no good" (Adam, 2026-11-15, phone screenshot of the reaction sheet): the on state was --accent-ink text on --accent, which in the dark theme is #e7e9ee on white. It now takes the accent's own text colour (--accent-fg): dark on white in the dark theme, white on near-black in the light one; the off state (dashed, muted) is unchanged Co-Authored-By: Claude Opus 5.5 <[email protected]> Claude-Session: https://claude.ai/code/session_01VMan6UKGJyuGyZFxddYRxS
* A zap to you gets one receipt, not one per device. Helen's 88-sat zap on ↵Adam Soltys3 days1-0/+15
| | | | | | | Adam's comment (2026-11-15 05:33) had two identical kind 9735 receipts a minute apart; Adam guessed "probably my phone and laptop". It was: the zaps waiting for their receipt live in the synced wallet state and every device holds the offer key that signs them, so each device published one when it saw the payment complete. coinos's notifications fold receipts by invoice (two rows were two zaps), other clients may not. Now a device skips a receipt it already sent (zapDone, synced, last 100) or that the relays already carry (one query by the offer key and recipient for that bolt11), and drops it from the pending list either way. noffer-test: the receipt goes out once however often the pass runs, and a second device holding the same pending zap finds it on the relays and stays quiet; nwc, notif-app-zap and features-smoke pass Co-Authored-By: Claude Opus 5.5 <[email protected]> Claude-Session: https://claude.ai/code/session_01VMan6UKGJyuGyZFxddYRxS
* A reply, zap or reaction is about the moment of a clip only when you say so. ↵Adam Soltys3 days6-15/+47
| | | | | | | "i zapped and commented on a post from helen yrmom and didn't intend to tie it to the video timestamp but i guess it did by default. maybe we can revisit the ui/flow there so it doesn't do that by default?" (Adam, 2026-11-14). Anything done to a post whose clip had played took the clip's current time: the reply opened as "Comment at 0:12" (an × to drop it), and a zap's or reaction's sheet had nothing showing it at all. Now the time is offered as a toggle, off until tapped — "⏱ About 0:12 in the clip", ✓ when on — in the reply box, the reaction sheet, and both zap sheets (Ark and Lightning); a one-tap zap, which has no sheet, is never tied. Ticked, it goes out as before (the media tag, "[0:12]" before a comment, its mark on the clip's bar). media-moments-feed-test ticks the toggle where it used to rely on the default and checks it starts off; media-timeline, features-smoke and the zap tests pass Co-Authored-By: Claude Opus 5.5 <[email protected]> Claude-Session: https://claude.ai/code/session_01VMan6UKGJyuGyZFxddYRxS
* Search inside someone's posts: their profile puts them in the search field, ↵Adam Soltys3 days4-29/+233
| | | | | | | and "franca wildwood" is understood. "i found him by searching franca but then i was wanting some way to search his feed for something … i was wanting to search 'franca wildwood' but knew it wouldn't work … we couldn't just have a clever way of guessing what the user is trying to do?" (Adam, 2026-11-14). No second field: on a profile the header field (and the phone's search page opened from a profile) carries the person as a chip, "Search their posts, or anyone…"; Enter searches only their posts, people still show below for picking someone else, and × or Backspace on the empty field searches everywhere again. Words that begin or end with someone's name (first/last one or two words, matched at the start of a word of their name or address) are offered as "“wildwood” in Franca Yibisi's posts" at the top of the results; when that's someone you follow or whose profile you opened this session, it's the Enter choice (↵), otherwise only a row ("bitcoin wallet" is not Bitcoin Magazine's posts), and never when the whole query is someone's name. The search is a feed like any text search, /feed?q=wildwood&p=npub…: the NIP-50 relays asked with the author, plus the author's own recent posts (home, archive and their own relays, up to four pages of 400 until ten match) checked here word by word, since few people are in a search index; what their profile already loaded shows at once. tools/search-scope-test.js (real relays): chip on fiatjaf's profile, "relay" → only his posts, every one saying it; "fiatjaf relay" guessed and taken by Enter; × and Backspace; phone width. profile-stream, post-search, search-avatar-loading and features-smoke pass Co-Authored-By: Claude Opus 5.5 <[email protected]> Claude-Session: https://claude.ai/code/session_01VMan6UKGJyuGyZFxddYRxS
* A withdraw link says when a claim couldn't be delivered, and while one is ↵Adam Soltys3 days2-1/+28
| | | | | | | still being paid. "As you can see, I made a transaction for 1,000,000 sats using the Withdraw link method, but they never arrived" (a user via Adam, 2026-11-14). The registrar couldn't route the claim to the wallet that asked, so nothing was sent and the 1M went back on the link, but the link's page said only "1,000,000 sats left", the same as an unclaimed one. Now the latest claim shows on the link's page: "on its way" / "couldn't find a Lightning route yet, still trying for a few minutes" (the registrar now retries for up to ten minutes, coinosv3-server a29af6f), looked at again every 10 s, or "couldn't be delivered … nothing was sent and the sats are still in this link", with the QR and Take back still there; the list row reads "claim failed · N left". lnurl-withdraw-links-test claims a second link with no route: retried, given back, both messages shown, and the same link pays once a route turns up Co-Authored-By: Claude Opus 5.5 <[email protected]> Claude-Session: https://claude.ai/code/session_01VMan6UKGJyuGyZFxddYRxS
* Typing a new search from a profile shows its results. "when i search ↵Adam Soltys4 days2-3/+18
| | | | | | | something and then click on a user to bring up their profile/feed then type into the search again i don't see any results, it's like the search box stops working. if i click enter it does a full text search" (Adam, 2026-11-14). The search is kept when a result is opened (Back from the profile lands on it), and the desktop header field only cleared the page on top when there was no search yet: from then on each keystroke updated a results page hidden under the profile, while Enter (openTextSearch) still worked. Now every keystroke there steps the profile, thread or live page aside and keeps the query and rows (clearFeatureNavKeepSearch, userSearchQuery); Back returns to the profile. A visitor (no wallet) typing there was sent to the wallet screen, which derived receive addresses from no seed ("Invalid mnemonic"); the wallet screen is only chosen with a wallet open. profile-stream-test types a second query into the 1300px header field from the opened profile: no results and no Back before, both now, and no page errors Co-Authored-By: Claude Opus 5.5 <[email protected]> Claude-Session: https://claude.ai/code/session_01VMan6UKGJyuGyZFxddYRxS
* A profile's posts stream in as the relays send them, with post-shaped ↵Adam Soltys4 days3-20/+90
| | | | | | | placeholders until the first arrive, and search warms its top two results first. "when i open someone's profile that i'm visiting for the first time it can take several seconds for any posts to show up. there's no indication that it's loading posts. then suddenly a whole feed appears. can we stream them as they arrive … maybe it could start fetching posts for the top one or two search results before we even click them?" (Adam, 2026-11-13, Ben Arc from search). fetchNotes looked up the author's relay list first, then waited for every relay (4.5 s cap), then for every post's media (8 s cap) before painting anything. Now it asks the home and archive relays at once and the author's own relays as soon as their list is known (queryStreamOn), paints the first post the moment it's verified and a batch every 300 ms after, gives the rows about to be on screen 1.2 s to size their media rather than 8, and merges each batch into the page as before. With nothing in hand the page shows three faint breathing post shapes (aria-busy, "Loading posts…"; still no spinner) instead of an invisible empty line. User search used to queue a profile prefetch for every row of every letter typed, two at a time, so the one tapped waited behind a dozen; now only the top two of the current answer are warmed, ahead of anything queued. Measured on the real relays (tools/profile-stream-test.js, search "fiatjaf", tap the top result): first post 5.8 s / >10 s before, 0.35–0.9 s now, placeholder within 20 ms. profile-open, -feed, -replies, -repost, -pager, feed-custom and features-smoke pass Co-Authored-By: Claude Opus 5.5 <[email protected]> Claude-Session: https://claude.ai/code/session_01VMan6UKGJyuGyZFxddYRxS
* Leaving a group from its own screen no longer breaks the app. "Tried to ↵Adam Soltys5 days1-1/+6
| | | | | | | leave but got this" (Adam, 2026-11-13): ReferenceError "Cannot access 'Q' before initialization" in the group screen. wnView declares `const wn = gClient(g)` for the group on screen, which shadows the module's wn for the whole function; the "groups still loading" test above it (`!wn || … || !ng || ngLoading`, widened for relay groups in 3197c70) ran when the group was gone — just left — and read the inner wn before it was set. The test now lives outside the function (groupClientsLoading). nip29-group-test leaves the dead-relay group from its own screen on the second device: "Something broke" without this, back to Messages with it Co-Authored-By: Claude Opus 5.5 <[email protected]> Claude-Session: https://claude.ai/code/session_01VMan6UKGJyuGyZFxddYRxS
* Relay groups with nothing said wait at the bottom of Messages, and a dead ↵Adam Soltys5 days3-3/+14
| | | | | | | relay says so. "They shouldn't be at the top if they have 0 members and no recent activity" (Adam, 2026-11-12), of four groups.0xchat.com groups that came in on his kind 10009 list (saved by Armada) and topped the phone's list. A group's place was its last message, else the moment this device first recorded it, so a list adoption counted as brand new. Now a relay group without messages sorts by when it was joined on this device (joinedAt, set by Join) and one that only arrived on the list sorts last. groups.0xchat.com no longer resolves: a group whose relay can't be reached and that has shown nothing reads "<host> isn't answering" on its row ("On <host>" while it is still quiet) instead of "0 members"; the mark clears with the relay's first event and on every start. nip29-group-test joins gone.invalid'deadgroup on the second device and reads the row Co-Authored-By: Claude Opus 5.5 <[email protected]> Claude-Session: https://claude.ai/code/session_01VMan6UKGJyuGyZFxddYRxS
* A relay group joined on one device shows up on the others. "Not seeing the ↵Adam Soltys5 days1-5/+36
| | | | | | | pyramid runners chat I joined and messaged in on my laptop appear on my phone" (Adam, 2026-11-12). The joined groups live in the kind 10009 list, but it was read once, when Messages started (a phone open in the background since before the join never looked again), and written only to the four home relays and the user's own, of which relay.coinos.io refused kind 10009 (allowed since, desk write-policy) and nos.lol and damus kept nothing: the laptop's list sat on primal and ditto alone, beside a stale copy on purplepag.es. Now the list is followed live (a limit-1 subscription on the home relays; a newer copy is adopted and its groups join the list here), read again whenever Messages resubscribes after an absence (at most every 30 s), always read deep (home, index and own relays) before a write, and published to all of those, so the newest copy is where any device looks. Groups dropped from the list elsewhere are still kept here. nip29-group-test: the same twelve words open a second browser, which follows the list live and shows the group once the laptop's signed list arrives on that subscription Co-Authored-By: Claude Opus 5.5 <[email protected]> Claude-Session: https://claude.ai/code/session_01VMan6UKGJyuGyZFxddYRxS
* The top bar stays on screen while the page scrolls. "I'm thinking maybe ↵Adam Soltys6 days1-0/+13
| | | | | | | topnav with the Coinos logo search/settings and profile avatar should stick around on screen even as you scroll the feed" (Adam, 2026-11-12). .brand-header is sticky at the top of every page that has it (feed, wallet, settings, profiles), as a frosted strip of the page background the width of the column, pulled out by the padding it adds so nothing moves at rest, clear of a notch by the safe-area inset. html scroll-padding-top keeps a post scrolled into view from landing under it. It sits at z-index 30, under every overlay (40 and up) and level with the bottom nav; nothing fixed lives inside it, so its backdrop-filter traps nothing. Checked on a 390px phone and a 1300px desktop against the live public feed scrolled 1400px; chat-layout, features-smoke and account-access tests pass Co-Authored-By: Claude Opus 5.5 <[email protected]> Claude-Session: https://claude.ai/code/session_01VMan6UKGJyuGyZFxddYRxS
* Settings → Wallet lists Spending and Savings, each opening its balance ↵Adam Soltys6 days2-1/+21
| | | | | | | card's settings. "can we make the spending wallet settings accessible through the main settings -> wallet page somehow?" (Adam, 2026-11-12), after Nils couldn't find the Spending settings: the gear on the Spending card was the only way in. A "Balance settings" card under the recovery phrase has a row per balance the wallet shows (Spending only once it's set up; a single-purpose wallet shows its one), each opening the same page as that card's gear (coins, renewals, custody and exit for Spending; lock, password, recovery phrase and public key for both). tools/balance-settings-row-test.js (mainnet, only the Ark server's info let through) Co-Authored-By: Claude Opus 5.5 <[email protected]> Claude-Session: https://claude.ai/code/session_01VMan6UKGJyuGyZFxddYRxS
* Messages joins NIP-29 relay groups: New group → Relay group → paste an ↵Adam Soltys6 days3-10/+418
| | | | | | | address → Join. "can i join pyramid runners from coinos v3? … add it" (Adam, 2026-11-11), for fiatjaf's pyramid.fiatjaf.com'Tnq7x2ZTgrPZWFrC. A NIP-29 group lives on one relay, which keeps its members and signs its metadata (39000 name/about/picture, 39001 admins, 39002 members); messages are plain signed events with an h tag. src/nip29-client.js speaks it behind the same interface as the White Noise and Cordn clients, so the group screens drive all three (g.nip29; gClient/gOn/findGroup/wnGroups learn it): live kind 9/7/5 (and 9000/9001/9021/9022) on the group's relay with NIP-42 AUTH where the relay asks (never an interactive signer unasked), metadata read only from the group's own relay, replies sent as NIP-C7 q tags (shown through an added e), NIP-29 previous tags from the relay's recent events, reactions and own-message delete as before. Join sends kind 9021 and holds the group either way (an open group reads before the relay's yes; a refusal is said in a toast); leave sends 9022. Addresses: host'id, wss://host'id, an naddr of the 39000, or a client's link (chachi.chat/<host>/<id>, flotilla's /spaces/<host>/<id>). The joined groups are the NIP-51 kind 10009 other clients keep: read at start (groups joined elsewhere appear here), written back on join and leave after reading the newest copy (deep on a first write, like the mute list), and a group left this session can't be brought back by a list read still in flight. Messages and metadata kept per identity in the marmot IndexedDB store (|nip29, last 200 per group; wiped with the wallet). The member panel says it isn't encrypted, shows about, and copies the address. No attachments, no creating groups or admin actions yet. tools/nip29-group-test.js reads the real group (relays: 'read'; join, list, message and leave publishes answered in place) Co-Authored-By: Claude Opus 5.5 <[email protected]> Claude-Session: https://claude.ai/code/session_01VMan6UKGJyuGyZFxddYRxS
* Post search shows a bot fleet's newsletter once, not forty-six times. ↵Adam Soltys6 days1-0/+30
| | | | | | | "searching "fiatjaf" surfaced this same post over and over i guess from different people all boosting it" (Adam, 2026-11-11). Not boosts: about a dozen keys each post the NostrMag "Quiet Coup" issue every few hours, reworded a little each time (an emoji, "BREAKING", the picture moved), so the exact-repeat spam check never matches. In a search feed a post whose words and links overlap half or more (Jaccard) with one already shown is dropped; posts under 8 words are left alone. Measured on 120 live hits for "fiatjaf" from the three search relays: newsletter copies overlap 0.53 and up, distinct posts 0.45 at most; 46 copies → 1, nothing else dropped. post-search-test adds three reworded copies from three keys Co-Authored-By: Claude Opus 5.5 <[email protected]> Claude-Session: https://claude.ai/code/session_01VMan6UKGJyuGyZFxddYRxS
* Search: Enter on typed words (no suggestion picked) opens a feed of the ↵Adam Soltys6 days3-5/+42
| | | | | | | notes and replies that say them. "if i submit just a string without selecting one of the suggestions i'd like it to do full text search on kind 1 posts and replies. so if i search fiatjaf i should find posts that are mentioning him without necessarily tagging him" (Adam, 2026-11-11). A session feed with def.search asks NIP-50 relays ({ kinds: [1], search }) — relay.ditto.pub, search.nos.today, nostr.wine, the three that answered a kind-1 search probe (nostr.band and primal answered nothing) — keeps replies as rows of their own, pages back with until, and has no live subscription. Enter in the search page's field or the desktop header field opens it, and the search page lists a “words” row ("Posts and replies that say this") above the topic and people rows. Its address is /feed?q=… so it can be shared. tools/post-search-test.js answers the search relays in the page Co-Authored-By: Claude Opus 5.5 <[email protected]> Claude-Session: https://claude.ai/code/session_01VMan6UKGJyuGyZFxddYRxS
* Settings → Wallet → Delete account: the name is given up, the servers ↵Adam Soltys8 days6-1/+136
| | | | | | | forget the account, and it leaves this device. "someone's asking for a way to delete their v3 account. we allowed that in v2 … we can expose a button for it in settings somewhere" (Adam, 2026-11-10). The page first asks the registrar what stands in the way (coinosv3-server f8c9efa, POST /account/delete with check): sats still on their way here or held in a withdraw link block it, and say so. It lists what goes (the name, free for anyone after; the server's records: payment address, notification registrations, scheduled posts, hats, point-of-sale and receipt records; the wallet on this device) and what does not: the coins, which only the recovery phrase controls, and what was published to nostr relays. Coins still here are named in sats with a warning, and the box to tick then says "I have moved my funds or saved my recovery phrase". On the tap the registrar deletes (signed by the wallet's key, which also answers for a login identity's name it manages), the profile drops a lud16/nip05 that named the freed address (addressGone; publishProfileFields learns `remove`), and the wallet is wiped from the session, the saved vault and its caches. A vault behind a real password not entered this session asks for a lock and unlock first. tools/account-delete-test.js (a stand-in registrar checks the NIP-98 signature) Co-Authored-By: Claude Fable 5.2 <[email protected]> Claude-Session: https://claude.ai/code/session_01VMan6V74NSFWZtNtgCfokN
* A community's admin taps a message and removes its author, with everything ↵Adam Soltys8 days5-6/+104
| | | | | | they said. "can i have some control as a chat group admin to click on a message and see an option to remove the user and all their messages from the group?" (Adam, 2026-11-10). The message sheet gains "Remove <name> and their messages" for the owner on anyone else's message; a first tap asks ("for everyone? Tap again"), the second does it. It is the owner's ban list (CORD-04 vsk 4): one owner-signed edition naming every removed key, which the app, the public /chat page and other Concord clients already fold: those authors' messages, old and new, drop out and they leave the member list. The list is replaced whole, so setBanned asks the relays for the control stream first and builds on the newest edition (banlistHead/makeBanlist chain version and hash), stopping if the relays hand back nothing. The member list shows the admin who was removed, each with Restore. Offered only where the app holds the control stream's signing key (the built-in community and ones founded here). Removed people no longer light a room's unread dot. And the control fold now lets only the owner's editions compete for a head: any member holds the control stream's key in a public community, and a higher-versioned edition of their own used to shadow the owner's ban list, name or channels (dropped, leaving nothing). tools/chat-remove-member-test.js runs it against a relay inside the test Co-Authored-By: Claude Fable 5.2 <[email protected]>
* The public chat page's button opens the chat for someone who has a wallet ↵Adam Soltys8 days4-5/+24
| | | | | | here, and stops telling them to sign in. "andy's having some trouble with the public chat … I can see the feed etc fine, so I am logged in" (Adam relaying Andy, 2026-11-10): /chat read "sign in to join the conversation" to a signed-in reader, and its "Open coinos" button went to the wallet, leaving him to find the chat. With an account on the device the line now reads "open the chat in Coinos to join the conversation" and the button "Join the chat", which goes to /?chat: the app takes the marker off the address and, once the wallet is open, opens the coinos community's room. A saved wallet counts as an account too (a tab opened from a link has no session yet). And /chat, /pos and /feed no longer paint the static profile shell (a card named "chat") while they load. tools/public-chat-join-test.js Co-Authored-By: Claude Fable 5.2 <[email protected]>
* A zap answered over NWC waits on far fewer round trips, and a phone with ↵Adam Soltys8 days5-30/+155
| | | | | | | | | Coinos behind another app answers from its worker. "I'm on V3 … I find nwc zaps a lot more slower than usual" (398ja, 2026-11-10). Measured from the notifier's wakes and the ASP's log over four days: a closed app took about six seconds from the push to its cosign (the Lightning part after that is one and a half), an Android phone fifteen at the median, up to thirty. Where it went: every call to the ASP and the notifier was two round trips (no Access-Control-Max-Age, so Chrome re-asked the CORS preflight after five seconds; a Cloudflare response-header rule now sets 7200 on OPTIONS for ark.coinos.io, ark-staging and nwcpush), and the worker made them one after another: answered?, GetArkInfo, Handshake, claim, a save, the chain tip, then the payment, then status ("pending"), the tip again, the held status call, and status once more for the preimage it had just been handed. Now: what the mirror alone decides (invoice, network, budgets, coins, keys) comes first with no network; the claim (which now also says "already answered", coinosv3-server 3c5e9c3) goes out beside the manager's start, whose info call, handshake and tip run together (init, warmTip/_tipFresh); a send just initiated returns to its caller instead of asking at once, and awaitLnPay settles from the held call's own answer. A worker-style send on regtest: eight calls, six in sequence, where there were thirteen. The app's own answer no longer waits a second on a phone that is on screen (only one behind another app holds back for a computer). Android freezes an app behind another one, and the worker's nudge to it sat unread until it was opened: on a phone the worker now hands a request only to a window on screen (as it already did on iOS), nudges a hidden one in case it is awake, and answers itself; the request's lock decides between them. For that the mirror is no longer written over the worker's hand: the worker stamps its writes (workerAt), the app writes under the worker's own lock (nwc-bg-work) and folds the worker's state in first (never mid-payment), carries the stamp, absorbs it when it comes back on screen (arkBgAbsorb), and refreshes the mirror as it goes behind another app. Tests: ln-pay-double-wake (the one held call settles; a failed or mismatched answer doesn't), sw-push-quota (an Android phone's hidden window), nwc-sw-test's notifier stand-in answers /claim as the real one; tools/ark/ln-e2e-test's pay half against regtest captaind 0.8 + CLN Co-Authored-By: Claude Fable 5.2 <[email protected]>
* A YouTube clip under a mouse takes its own clicks, and YouTube's full-screen ↵Adam Soltys8 days2-2/+29
| | | | | | button leaves our full screen. "The fullscreen icon on this youtube video worked to make the video fullscreen but not toggle it off … clicking the central pause button didn't pause, it fullscreened. it should just pause" (Adam, 2026-11-10). YouTube's newer player puts its pause button in the middle and its full-screen button above the bottom bar, both under the layer that takes a tap on the picture to full screen: so either press took the box full screen, and once there (the layer aside) the same button asked for full screen again from inside the frame, which shows nothing and leaves no way out. Now the layer is off while a mouse is over the box (by-mouse, set from the pointer's type; a finger gets it back), so the player's buttons do what they say, and a mouse click on the still plays it where it is. A touch's tap still goes full screen with sound; there, when the frame becomes the full-screen element in place of its box (the page sees that much), it's read as "out" and both steps are left. youtube-tap-test covers the touch runs with a finger, the button leaving full screen, and a mouse Co-Authored-By: Claude Fable 5.2 <[email protected]>
* A scanned LNURL opens as what it is even while Spending is still connecting. ↵Adam Soltys10 days1-1/+9
| | | | | | | "I refreshed and scanned it again and it put the lnurl in the recipient field and gave me a zap button" (Adam, 2026-11-08): zaps' matchSendText turned everything down until canLnPay, which is false for the first moments after a refresh, so his own withdraw link fell through to the send form's "Zap LNURL…" affordance. A withdraw pays us and needs nothing spendable, and only the GET says which kind an lnurl is, so a scanned or pasted lnurl is taken first and resolved; a pay one meets the pay checks where it's paid. lnurl-withdraw-test covers the scan with nothing spendable Co-Authored-By: Claude Opus 5.5 <[email protected]> Claude-Session: https://claude.ai/code/session_01VMan6Ts5Tktega8VbLg31q
* A Lightning receive steps past a payment hash the server says was canceled, ↵Adam Soltys10 days1-3/+5
| | | | | | | as it already did for one paid or merely issued. Claiming Adam's own withdraw link (2026-11-08) failed before the claim was sent: "StartLightningReceive: bad user input: the invoice for payment hash c2d7… was canceled" — the preimage counter was sitting on an index whose invoice an earlier cancel retired (another device's, or a state that hadn't heard), and createLnInvoice only treated "already been paid" / UNIQUE constraint answers as spoken-for, so it threw instead of taking the next index. Matched on the server's wording ("payment hash <hex> was canceled"), not a bare "cancelled", so an aborted request still fails as itself Co-Authored-By: Claude Opus 5.5 <[email protected]> Claude-Session: https://claude.ai/code/session_01VMan6Ts5Tktega8VbLg31q
* A zap on your Zapstore app reads "zapped your app PAFF" in Notifications and ↵Adam Soltys11 days2-3/+64
| | | | | | | opens the app's Zapstore page. "Clicking this franzap zap notification keeps bringing me to his profile not a note" (Adam, 2026-11-07): franzap's 10k "Love it" was a zap on the PAFF listing (kind 32267, a-tagged in the zap request), but Zapstore's receipt carries only p and P, so the row read as a zap on nothing and fell through to the zapper's profile. Rows now take a zap's address from the receipt or, failing that, the zap request inside it (zapAddrOf; streams get that too); a 32267 one is an app zap, named from relay.zapstore.dev once per session, opening zapstore.dev/apps/<id>. A row saved before reads its receipt by id on the first tap and goes where it says; a relay handing the receipt again heals it as well. tools/notif-app-zap-test.js Co-Authored-By: Claude Opus 5.5 <[email protected]> Claude-Session: https://claude.ai/code/session_01VMan6Ts5Tktega8VbLg31q
* Prepaid Lightning withdraw links: Send → ⚡ Withdraw link makes an LNURL ↵Adam Soltys11 days3-2/+230
| | | | | | | (QR + copy) that anyone can withdraw sats from, like a voucher, with this wallet asleep. "Do we have lnurl withdraw links somewhere? can we add them?" → "Prepaid" (Adam, 2026-11-07); redeeming them already worked (38ce0b4), making them didn't. Amount per withdrawal, how many withdrawals (1–1000), an optional description; the whole amount is paid up front from Spending over Lightning to the coinos registrar (coinosv3-server 912944d, names/withdraw.js), which pays each claim out of it, so it works with any Lightning wallet while the phone is off; "Your withdraw links" lists them (what's left, used up, closed) and a link's page has Take back {what's left}, paid to a fresh Lightning receive of this wallet, which closes the link. Links belong to the wallet's nostr key (NIP-98), so every device with the seed sees them. Lives with the gift link (deferred gifts chunk, ui.giftMode === 'lnurl'). tools/lnurl-withdraw-links-test.js: makes one against the real service behind a stand-in server with a fake node, claims it through our own redeem flow, takes the rest back Co-Authored-By: Claude Opus 5.5 <[email protected]> Claude-Session: https://claude.ai/code/session_01VMan6Ts5Tktega8VbLg31q
* A flick through the feed no longer drags background work through the middle ↵Adam Soltys11 days1-18/+71
| | | | | | | of it. "When I flick to scroll, it's not getting up to speed very quickly, and clicking around in the app feels a bit sluggish at times" (Adam, 2026-11-07). Measured on his Pixel 9 in the installed PWA: idle is quiet (0.16 s of script in 10 s), but three real flicks through Following spent 1.7 s of 9 on the main thread, frames up to 200 ms, and the work was the avatar thumbnailer finishing faces as they scrolled in: a synchronous canvas.toDataURL WebP encode per face (229 ms), the profile cache JSON.stringified twice per half-second flush (once just to measure it, up to ~2 MB with thumbnails), and a whole-page repaint per finished face that changes nothing on screen (feed rows keep the face they were admitted with). Now: the crop, resize and encode happen off the main thread (createImageBitmap with resizeWidth/Height, OffscreenCanvas.convertToBlob, FileReader), with the old canvas path as the fallback; a finished face is recorded once the page settles; scheduleRepaint holds while a finger is down or the page scrolled in the last 200 ms (at most 1.5 s), and the profile flush likewise (at most 5 s, forced on an account switch) and sizes the blob by summing rows instead of stringifying it. Tests: the create flow's heads-up / passphrase / check steps from earlier today broke every test that made a wallet by waiting for the words and clicking "Skip verification"; tools/puppeteer.js exports seedWords() and skipSeedCheck() and the 96 tests use them. room-faces and avatar-profile fail the same way on 8858591 (row.hover, scrollIntoView), before today's changes Co-Authored-By: Claude Opus 5.5 <[email protected]> Claude-Session: https://claude.ai/code/session_01VMan6Ts5Tktega8VbLg31q
* The passphrase gets a page of its own in the create flow, between the words ↵Adam Soltys11 days2-21/+64
| | | | | | | and the check. "The add passphrase button is too small … maybe it should get its own separate prompt screen with explainer of what it is and letting people know it's optional but recommended. we don't need a 'show previous screen' link on verify page because there's already a back button. 'Have an existing seed or entropy to import?' can go away once the seed is displayed" (Adam, 2026-11-07). New wallets go heads-up → words (Copy, Next) → "Add a passphrase?" (an extra secret on top of the 12 words; optional but recommended; write it down with the seed, nobody can reset it), with the field, its confirmation once typed, and one button that reads "Continue without one" until something is typed, then Next; a mismatch is caught there and clears once the fields agree → "Which one is word #5?". The page's own Back now steps through those (check → passphrase → words → heads-up, keeping what was typed) on the first-run and "Sign into another account" pages, so the check drops "Show the words again"; the classic card, which has no Back of its own, keeps one. "Have an existing seed or entropy to import?" shows only on the heads-up (and its "create instead" twin on the import page). The import pane's Add passphrase is a full-width button. passFields() split out of optionsPanel Co-Authored-By: Claude Opus 5.5 <[email protected]> Claude-Session: https://claude.ai/code/session_01VMan6Ts5Tktega8VbLg31q
* "Sign into another account" opens the way the start screen does: "Add an ↵Adam Soltys11 days2-3/+74
| | | | | | | account" with Create a new wallet / Sign in to an existing account, then Sign in with Seed phrase, Passkey, Google, Nostr. "Can you update the 'sign in to another account' screen to match the initial start screen?" (Adam, 2026-11-07). It used to open on the old Create new / Import existing tabs with the sign-in buttons stacked underneath. addAccountScreen (unlockScreen while ui.fromWallet && ui.addFlow, set by signInAnother) draws the front door's two steps with the same buttons (frontDoorSignin, short labels, the same loginWith, so a sign-in still adds beside the open accounts); Nostr opens its signer card in place; a choice leads to the front door's seed pages (no tabs, titled Set up your wallet / Enter seed phrase, Back to where the choice was made), and Back on the first step returns to the wallet. The coinos.io migrate door stays on the first-run screen only. The Accounts "new wallet" kind picker keeps the classic card (addFlow false). account-access and padlock-flow tests walk the new screens Co-Authored-By: Claude Opus 5.5 <[email protected]> Claude-Session: https://claude.ai/code/session_01VMan6Ts5Tktega8VbLg31q
* A new seed phrase comes in three steps: a heads-up, the words, then one word ↵Adam Soltys11 days4-96/+125
| | | | | | | to pick. "Reintroduce the seed phrase preamble telling users to write it down but don't give it a warning icon … an intermediary step before we display the seed phrase … instead of verify and skip buttons just have a next and another intermediate step that asks for just the 5th seed word and gives a few options to click and also a skip button" (Adam, 2026-11-07). Creating a wallet (front door, Accounts, signInAnother) now opens on a plain note, "Next you'll see your seed phrase: 12 words … Write them down on paper, in order … make sure no one is looking", with Show seed phrase; the words page has one Next; the check asks "Which one is word #5?" with four of the phrase's own words as buttons (a wrong one turns red and says "Not that one"), Skip, and Show the words again. The gift claim's new-wallet backup gets the same three steps in place of its warning box and typed three-word check, and a sign-in's "Write down your recovery phrase" shows its explanation first and the words behind Show seed phrase (Later still on both). createStep starts at 'intro'; pickConfirm returns { index, options, wrong } Co-Authored-By: Claude Opus 5.5 <[email protected]> Claude-Session: https://claude.ai/code/session_01VMan6Ts5Tktega8VbLg31q
* Under the "Sign in" heading the ways in just name themselves: Seed phrase, ↵Adam Soltys11 days3-9/+13
| | | | | | | Passkey, Google, Nostr. "I think we can remove 'Sign in' from all the sign in buttons since it's at the top in the heading" (Adam, 2026-11-06). frontDoorSignin takes { short } and the sign-in step asks for it; the coinos.io move page (under "or") and the sign-in sheet in Messages keep the full "Sign in with …" labels, where nothing above says it. "Migrate a coinos account" is unchanged. Tests wait for the list by its last door Co-Authored-By: Claude Opus 5.5 <[email protected]> Claude-Session: https://claude.ai/code/session_01VMan6Ts5Tktega8VbLg31q