max / makenotwork

Park the CSS seal and fold the duplicated copy and toast helpers The seal: hand-written CSS is 340 KiB against a sealed 337, and the test asserts equality, so it fires on a decrease as well as an increase. Mid conversion that means re-baselining every batch, which measures nothing. Parked with #[ignore] rather than deleted, CSS_KIB_HIGH_WATER kept as the record of where the conversion started, and the exit condition written down: the conversion finished and the duplicated CSS gone. the_two_lists_agree stays active, since bytes moving to an unweighed sheet is a different failure. Max, 2026-08-18. Six copy buttons hand-rolled the label revert that core/clipboard.ts already does through its resetMs parameter. They now call it through the legacy bridge. Durations preserved per site, including the one that waits 2000ms where the others wait 1500: which is right is a question for the semantic-timing work, not for a consolidation pass. The error toast in htmx-glue carried a verbatim copy of toast.ts's dismissal timers. Extracted as autoDismiss(toast, lifetimeMs), with the lifetime still the caller's: 3000 for an ordinary toast, 6000 for one carrying a Retry button. The audit's fourth claim does not hold: there is one debounce helper (core/tabs.ts), not four. The other four are timer handles named debounce, used inline. Left alone, since folding them would add a global to a bridge whose purpose is to lose entries. GoingsOn mnw-server 2d01900f, 033c722f.

Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>

Author: Max Johnson <me@maxj.phd> · 2026-08-18 18:16 UTC

Signed with PGP, not checked

Commit: 4c2fd4e2fe125f703a4c1d597b09d8acb56eae9f

Parent: 6d9f5ea

6 files changed, +76 insertions, -27 deletions

M server/static/actions-tabs.js +6 -3

			@@ -97,7 +97,7 @@
97	97		};
98	98
99	99		window.copyKeyCode = function (code) {
100		-	navigator.clipboard.writeText(code); this.textContent = 'Copied!'; setTimeout(() => this.textContent = code, 1500);
	100	+	window.copyWithFeedback(this, code, 'Copied!', 1500);
101	101		};
102	102
103	103		window.toggleContextMenuBtn = function (itemId) {
			@@ -130,11 +130,14 @@
130	130		};
131	131
132	132		window.copyElementText = function (id) {
133		-	navigator.clipboard.writeText(document.getElementById(id).textContent).then(() => { this.textContent = 'Copied'; setTimeout(() => this.textContent = 'Copy', 1500); });
	133	+	window.copyWithFeedback(this, document.getElementById(id).textContent, 'Copied', 1500);
134	134		};
135	135
	136	+	// 2000ms here rather than the 1500 every other copy button uses. Preserved as
	137	+	// found: which of the two is right is a question for the semantic-timing work,
	138	+	// not for a consolidation pass.
136	139		window.copyFeedUrl = function () {
137		-	navigator.clipboard.writeText(document.getElementById('feed-url').value).then(() => { this.textContent = 'Copied!'; setTimeout(() => this.textContent = 'Copy URL', 2000); });
	140	+	window.copyWithFeedback(this, document.getElementById('feed-url').value, 'Copied!', 2000);
138	141		};
139	142
140	143		// --- user_media.html ---

M server/static/synckit-tabs.js +2 -10

			@@ -29,21 +29,13 @@
29	29		btn.className = className;
30	30		btn.textContent = 'Copy';
31	31		btn.addEventListener('click', function () {
32		-	navigator.clipboard.writeText(text).then(function () {
33		-	btn.textContent = 'Copied';
34		-	setTimeout(function () { btn.textContent = 'Copy'; }, 1500);
35		-	});
	32	+	window.copyWithFeedback(btn, text, 'Copied', 1500);
36	33		});
37	34		return btn;
38	35		}
39	36
40	37		window.syncKitCopyKey = function (event, key) {
41		-	navigator.clipboard.writeText(key).then(function () {
42		-	var btn = event.target;
43		-	var orig = btn.textContent;
44		-	btn.textContent = 'Copied';
45		-	setTimeout(function () { btn.textContent = orig; }, 1500);
46		-	});
	38	+	window.copyWithFeedback(event.target, key, 'Copied', 1500);
47	39		};
48	40
49	41		window.syncKitRegenKey = function (appId) {

M server/static/tab-item-embed.js +1 -5

			@@ -5,11 +5,7 @@
5	5		var itemTitle = cfg.dataset.itemTitle;
6	6
7	7		window.copyEmbed = function(btn) {
8		-	const code = btn.previousElementSibling.textContent;
9		-	navigator.clipboard.writeText(code).then(() => {
10		-	btn.textContent = 'Copied!';
11		-	setTimeout(() => btn.textContent = 'Copy', 1500);
12		-	});
	8	+	window.copyWithFeedback(btn, btn.previousElementSibling.textContent, 'Copied!', 1500);
13	9		};
14	10
15	11		window.switchCardLayout = function(layout) {

M server/tests/frontend_payload.rs +43

			@@ -7,6 +7,14 @@
7	7		//! and that is currently the largest standing cost of the conversion, paid on
8	8		//! every page load by every visitor including signed-out ones.
9	9		//!
	10	+	//! # Parked, 2026-08-18
	11	+	//!
	12	+	//! The ratchet below is `#[ignore]`d by Max's ruling until the conversion is
	13	+	//! finished and the duplicated CSS is gone. Everything this doc argues still
	14	+	//! holds and none of it is retracted; see the test's own doc comment for why the
	15	+	//! enforcement was suspended anyway, and for what turns it back on.
	16	+	//! `the_two_lists_agree` is unaffected and still runs.
	17	+	//!
10	18		//! # Why this seal exists rather than a note saying the same thing
11	19		//!
12	20		//! The conversion's payoff is the hand-written sheet shrinking. Without a
			@@ -83,7 +91,42 @@
83	91		(each.iter().map(\|(_, bytes)\| bytes).sum(), each)
84	92		}
85	93
	94	+	/// Parked by Max on 2026-08-18, and the module doc above argues against parking
	95	+	/// it, so the argument gets answered rather than ignored.
	96	+	///
	97	+	/// What the doc says: a guard people edit rather than read is not a guard, and
	98	+	/// without a number "we converted a screen" and "we removed its cost" are the
	99	+	/// same claim. Both points stand. What they do not survive is the assertion
	100	+	/// pair below: this test fails on any decrease as well as any increase, so mid
	101	+	/// conversion it fires on work going in the direction it exists to reward, and
	102	+	/// the number gets re-baselined every batch. A seal that is re-baselined on
	103	+	/// schedule measures nothing, and it teaches exactly the editing habit the doc
	104	+	/// warns about.
	105	+	///
	106	+	/// So the number is kept and the enforcement is suspended, rather than the seal
	107	+	/// being quietly lowered batch by batch. `CSS_KIB_HIGH_WATER` stays at 337
	108	+	/// because it is the record of where the conversion started, and it is what the
	109	+	/// finished conversion gets measured against.
	110	+	///
	111	+	/// WHAT BRINGS IT BACK: the conversion complete and the duplicated CSS removed.
	112	+	/// At that point the number stops moving in both directions, and the decision
	113	+	/// deferred with it becomes worth taking: seal the minified size rather than raw
	114	+	/// bytes, which is what Max ruled earlier the same day. `lightningcss` is
	115	+	/// already a dependency of this repo, so that costs no new dependency.
	116	+	///
	117	+	/// STATE AT PARKING, measured 2026-08-18:
	118	+	///
	119	+	/// static/style.css 324,648 (322,102 at the 2026-08-15 baseline)
	120	+	/// static/wizard.css 13,459
	121	+	/// static/media-player.css 8,953
	122	+	/// static/no-js.css 1,212
	123	+	/// total 348,272 = 340 KiB against the sealed 337
	124	+	///
	125	+	/// Red on the grow assertion by 3 KiB, which is what prompted the ruling.
	126	+	///
	127	+	/// GoingsOn mnw-server `2d01900f`.
86	128		#[test]
	129	+	#[ignore = "parked by Max 2026-08-18 until the conversion is finished and the duplicated CSS is gone: see this test's doc comment"]
87	130		fn hand_written_css_does_not_grow() {
88	131		let (total, each) = measure();
89	132		let kib = total / 1024;

M server/frontend/src/core/htmx-glue.ts +4 -5

			@@ -4,7 +4,7 @@
4	4
5	5		import { resolveHtmxLoadingButton } from './loading.ts';
6	6		import { csrfHeaders } from './net.ts';
7		-	import { showToast } from './toast.ts';
	7	+	import { autoDismiss, showToast } from './toast.ts';
8	8
9	9		interface HxDetail {
10	10		elt: HTMLElement;
			@@ -78,10 +78,9 @@
78	78		toast.appendChild(closeBtn);
79	79
80	80		container.appendChild(toast);
81		-	setTimeout(() => {
82		-	toast.classList.add('fade-out');
83		-	setTimeout(() => toast.remove(), 300);
84		-	}, 6000);
	81	+	// 6000 rather than the 3000 an ordinary toast gets: this one carries a Retry
	82	+	// button, so it has to outlive a glance.
	83	+	autoDismiss(toast, 6000);
85	84		});
86	85
87	86		// Loading button on request start; restore on every terminal event.

M server/frontend/src/core/toast.ts +20 -4

			@@ -9,6 +9,25 @@
9	9		type?: string;
10	10		}
11	11
	12	+	/** How long the fade-out animation runs before the node is removed. Matches the
	13	+	* `.fade-out` transition in the stylesheet, so shortening it here drops the
	14	+	* toast mid-animation. */
	15	+	const TOAST_FADE_MS = 300;
	16	+
	17	+	/** Fade `toast` out after `lifetimeMs`, then remove it.
	18	+	*
	19	+	* Extracted because `initHtmxGlue`'s error toast had its own verbatim copy of
	20	+	* the same two nested timers. The lifetime stays the caller's business: this
	21	+	* renderer's toasts live 3000ms and an error toast lives 6000ms, and which of
	22	+	* those is right is a question for the semantic-timing work rather than for a
	23	+	* helper that only exists to stop the dismissal being written twice. */
	24	+	export function autoDismiss(toast: HTMLElement, lifetimeMs: number): void {
	25	+	setTimeout(() => {
	26	+	toast.classList.add('fade-out');
	27	+	setTimeout(() => toast.remove(), TOAST_FADE_MS);
	28	+	}, lifetimeMs);
	29	+	}
	30	+
12	31		/** Fire a toast. Rendered by the delegated listener installed by `initToasts`. */
13	32		export function showToast(message: string, type = 'error'): void {
14	33		document.body.dispatchEvent(new CustomEvent<ToastDetail>('showToast', { detail: { message, type } }));
			@@ -28,9 +47,6 @@
28	47		toast.className = 'toast toast-' + (detail.type \|\| 'info');
29	48		toast.textContent = detail.message \|\| 'Action completed';
30	49		container.appendChild(toast);
31		-	setTimeout(() => {
32		-	toast.classList.add('fade-out');
33		-	setTimeout(() => toast.remove(), 300);
34		-	}, 3000);
	50	+	autoDismiss(toast, 3000);
35	51		});
36	52		}