This Day In History For 2027-09-06
1 Year Ago (more context)
|
2026-09-06
| ||
| 22:54 | • Edit [0cd3610f2a]: Add "ci-pass" with value "db2906e8a700375e1470186d61722642". artifact: 9a450f26be user: quarry | |
| 22:54 | • Post: Pin zit 0.3.7. Four read verbs now cost what the image already paid for: grep over a file's history, timeline, info NAME, and reads on caches written by older builds; a forge rendering a page from many info calls sees each at the process-start floor. No artifact, tag or sync change. The 0.3.7 Linux tarball is staged on the host and verifies; the full suite is green locally against zit 0.3.7. artifact: 3d1a6e955f user: zeitgeist | |
| 22:54 | Pin zit 0.3.7 for the Linux build (grep, timeline and info read what the image already holds) closed check-in: 0cd3610f2a user: dmestas tags: zit-0.3.7 | |
| 22:52 | • Edit [5668b43c41]: Add "ci-pass" with value "e5480927dfead084a102d2f27e816edb". artifact: a64c14208c user: quarry | |
| 22:52 | land chat-letter check-in: 5668b43c41 user: dmestas tags: trunk | |
| 22:52 | • Edit [5668b43c41]: Add "landing" with value "56de6ba0c5f51d8ba7634cf9d2f039da9bc7ebe9306f39b9a0621896ca72f0e1". artifact: 2904247544 user: quarry | |
| 22:52 | • Reply: Access hole: the chat page rendered the whole channel for any reader the repository was visible to, and the stream served anyone without a session, while the API op behind both refused readers lacking the chat letter. Seen after switching to an account with no role, which our settings give tickets only. When a capability plane answers, the page now needs the letter (a visitor with no session is sent to sign in, a signed-in reader without it is refused), the stream refuses without ever redirecting, and the Chat tab hides on the same answer, since a tab to a refusal is a wrong tab. A forge with no capability plane keeps its old behaviour. Tests cover the tab, the page and the stream on both sides of the letter; the rail test now expects the refusal rather than an empty rail. artifact: fee3eccc4b user: dmestas | |
| 22:52 | • Edit [c380d073a9]: Mark "Closed". artifact: 2db922e1e1 user: quarry | |
| 20:17 | • Edit [c380d073a9]: Add "ci-pass" with value "f7de9430f953b4dea71e22531a9ea969". artifact: 6cadf0ca1a user: quarry | |
| 20:16 | • Post: Access hole: the chat page rendered the whole channel for any reader the repository was visible to, and the stream served anyone without a session, while the API op behind both refused readers lacking the chat letter. Seen after switching to an account with no role, which our settings give tickets only. When a capability plane answers, the page now needs the letter (a visitor with no session is sent to sign in, a signed-in reader without it is refused), the stream refuses without ever redirecting, and the Chat tab hides on the same answer, since a tab to a refusal is a wrong tab. A forge with no capability plane keeps its old behaviour. Tests cover the tab, the page and the stream on both sides of the letter; the rail test now expects the refusal rather than an empty rail. artifact: 56de6ba0c5 user: zeitgeist | |
| 20:16 | Chat takes the chat letter on the page, the stream and the tab, as the API op already did closed check-in: c380d073a9 user: dmestas tags: chat-letter | |
| 20:01 | • Edit [aa4f7edd77]: Add "ci-pass" with value "6f935dffbc6870b2123ae7d2e2252aef". artifact: e9238e2611 user: quarry | |
| 20:01 | land ticket-inline-edit check-in: aa4f7edd77 user: dmestas tags: trunk | |
| 20:01 | • Edit [aa4f7edd77]: Add "landing" with value "fc62cbe30e9225a3349d804258de4d8f8d33eaf237ac4325c19133ddde085ccb". artifact: 7810810a5a user: quarry | |
| 20:01 | • Reply: timeline: never serve a stale leaf set; a merged check-in must not keep its ring artifact: 8a2047c82c user: dmestas | |
| 20:01 | • Edit [7f9399b497]: Mark "Closed". artifact: c0422756a8 user: quarry | |
| 20:01 | • Edit [7f9399b497]: Add "ci-pass" with value "5ef8d1cbf427d6964f8ecf0cbd29b94b". artifact: 41a4c45b3f user: quarry | |
| 20:00 | • Post: timeline: never serve a stale leaf set; a merged check-in must not keep its ring artifact: fc62cbe30e user: dmestas | |
| 20:00 | timeline: never serve a stale leaf set; a merged check-in must not keep its ring closed check-in: 7f9399b497 user: dmestas tags: ticket-inline-edit | |
| 19:58 | • Edit [d4ea112ef4]: Add "ci-pass" with value "569407b5aa5557534aab0180af98cade". artifact: dcdccc068b user: quarry | |
| 19:58 | • Reply: Two things on the Branches panel of repository settings. Layout: the rule rows sat flush against the card frame with the editor's fields ragged; rows now carry the card's inner padding, the editor's fields align on one label column (stacking on a phone), the fold gets a scoped focus ring instead of a full-width outline, and the Remove control has room. Verified on a capture of the live page. Truth: the panel warned that nothing had reported the quarry check in 90 days while the forge's own runner reports it hourly, because the seen-checks query read only external statuses. It now unions the forge's own CI runs by workflow name, which also makes those names appear in the check-name suggestions. Test covers a recent run counting and an ancient one not. artifact: b186d14859 user: dmestas | |
| 19:58 | land protect-editor check-in: d4ea112ef4 user: dmestas tags: trunk | |
| 19:58 | • Edit [d4ea112ef4]: Add "landing" with value "a444abfb56e99fa05a3f64246540ed091e563cf78829740cfca5441a59fbff71". artifact: cb01290f9e user: quarry | |
| 19:58 | • Reply: Two things on the Branches panel of repository settings. Layout: the rule rows sat flush against the card frame with the editor's fields ragged; rows now carry the card's inner padding, the editor's fields align on one label column (stacking on a phone), the fold gets a scoped focus ring instead of a full-width outline, and the Remove control has room. Verified on a capture of the live page. Truth: the panel warned that nothing had reported the quarry check in 90 days while the forge's own runner reports it hourly, because the seen-checks query read only external statuses. It now unions the forge's own CI runs by workflow name, which also makes those names appear in the check-name suggestions. Test covers a recent run counting and an ancient one not. artifact: a60c9b2251 user: dmestas | |
| 19:58 | • Edit [b871e04414]: Mark "Closed". artifact: 0ac300d6cf user: quarry | |
| 19:58 | • Edit [b871e04414]: Add "ci-pass" with value "e358b24704d119bde33b2bb4ad9f1888". artifact: 46ebe2bb82 user: quarry | |
| 19:57 | • Post: Two things on the Branches panel of repository settings. Layout: the rule rows sat flush against the card frame with the editor's fields ragged; rows now carry the card's inner padding, the editor's fields align on one label column (stacking on a phone), the fold gets a scoped focus ring instead of a full-width outline, and the Remove control has room. Verified on a capture of the live page. Truth: the panel warned that nothing had reported the quarry check in 90 days while the forge's own runner reports it hourly, because the seen-checks query read only external statuses. It now unions the forge's own CI runs by workflow name, which also makes those names appear in the check-name suggestions. Test covers a recent run counting and an ancient one not. artifact: a444abfb56 user: zeitgeist | |
| 19:57 | Protected-branch rules: rows get their padding and an aligned label column; a required check the forge's own CI reports is no longer called stale closed check-in: b871e04414 user: dmestas tags: protect-editor | |
| 19:54 | • Edit [4501f0db9b]: Add "ci-pass" with value "ac8ababa6fd5b557d5d662dbda9b8525". artifact: ae5e06fb6f user: quarry | |
| 19:53 | land ticket-inline-edit check-in: 4501f0db9b user: dmestas tags: trunk | |
| 19:53 | • Edit [4501f0db9b]: Add "landing" with value "8171fd76431fc2c848d4317ee99d5d5f5955a35ab06cb3567ead63b9286b50f0". artifact: 500c289a6c user: quarry | |
| 19:53 | • Reply: landing: drop the full-diff link; the page already pages in every line artifact: a48bf0b6fe user: dmestas | |
| 19:53 | • Edit [0bc0c3a6ef]: Mark "Closed". artifact: 33c1387b6a user: quarry | |
| 19:52 | • Edit [0bc0c3a6ef]: Add "ci-pass" with value "0765c23c7d7529b9bfcfd9a77df04c10". artifact: 929a0291f6 user: quarry | |
| 19:52 | • Post: landing: drop the full-diff link; the page already pages in every line artifact: 8171fd7643 user: dmestas | |
| 19:51 | landing: drop the full-diff link; the page already pages in every line check-in: 0bc0c3a6ef user: dmestas tags: ticket-inline-edit | |
| 19:49 | • Edit [5f79616ff0]: Add "ci-pass" with value "6295157bbda37e9c47e4cfbccbe63557". artifact: 96524a46b3 user: quarry | |
| 19:48 | land ticket-inline-edit check-in: 5f79616ff0 user: dmestas tags: trunk | |
| 19:48 | • Edit [5f79616ff0]: Add "landing" with value "7bd64cfdeb39debb8ce7173b1718d9556e35a713db6e2fc849029a409dab415b". artifact: 7b196af388 user: quarry | |
| 19:48 | • Reply: ticket: edit the title and description in place, where they are read artifact: 1502c5b4cf user: dmestas | |
| 19:48 | • Edit [5660ca9083]: Mark "Closed". artifact: db86931fc0 user: quarry | |
| 16:25 | • Edit [5660ca9083]: Add "ci-pass" with value "afca7c69eafc6f7425895c7b8c56d711". artifact: 51c06d093e user: quarry | |
| 16:24 | • Post: ticket: edit the title and description in place, where they are read artifact: 7bd64cfdeb user: dmestas | |
| 16:24 | • Edit [5660ca9083]: Move to branch ticket-inline-edit. artifact: cd92d4f0e0 user: dmestas | |
| 16:24 | ticket: edit the title and description in place, where they are read check-in: 5660ca9083 user: dmestas tags: ticket-inline-edit | |
| 16:10 | • Edit [7433973031]: Add "ci-pass" with value "cb3bbd832aa262dbe99236b0fa07227b". artifact: 1f845038a2 user: quarry | |
| 16:07 | land landing-log-overflow check-in: 7433973031 user: dmestas tags: trunk | |
| 16:07 | • Edit [7433973031]: Add "landing" with value "f3c7834975305aa963257a355cc65fc040d807df7d0e53ae9756a35914307829". artifact: 03de7669e7 user: quarry | |
| 16:07 | • Reply: On a landing page the running check's log ran off the card to the viewport edge. The .l-check flex rule, written for the waiting row (a p), also hit the check's details element; a flex details puts its body in an internal slot the stylesheet cannot reach, so the pre grew to its longest line (4,900 characters here) and nothing above it could shrink. The flex rule is now scoped to the waiting row and the details is a block, so the pre's own overflow rule works. Reproduced and verified against a capture of the live page at 1400px: the log scrolls inside the card. artifact: 7582ee86e2 user: dmestas | |
| 16:07 | • Edit [1f5553d3e4]: Mark "Closed". artifact: 01cc8794bf user: quarry | |
| 16:06 | • Edit [1f5553d3e4]: Add "ci-pass" with value "5044a9db2fc609d2b462e6145edc7437". artifact: 82d6c855f3 user: quarry | |
| 16:05 | • Edit [b0e0b307c6]: Add "ci-pass" with value "e7ea4988c435501ae57a990624f67519". artifact: fbf8bc316b user: quarry | |
| 16:05 | • Post: On a landing page the running check's log ran off the card to the viewport edge. The .l-check flex rule, written for the waiting row (a p), also hit the check's details element; a flex details puts its body in an internal slot the stylesheet cannot reach, so the pre grew to its longest line (4,900 characters here) and nothing above it could shrink. The flex rule is now scoped to the waiting row and the details is a block, so the pre's own overflow rule works. Reproduced and verified against a capture of the live page at 1400px: the log scrolls inside the card. artifact: f3c7834975 user: zeitgeist | |
| 16:04 | A landing's run log scrolls inside its card: the check details is a block again, only the waiting row is a flex line closed check-in: 1f5553d3e4 user: dmestas tags: landing-log-overflow | |
| 16:04 | • Edit [54ff2a1ebd]: Add "ci-pass" with value "5e9b79a119d4d2846334cc1ec8ec2e40". artifact: 4f379fe16e user: quarry | |
| 16:04 | land feed-landings-chip check-in: b0e0b307c6 user: dmestas tags: trunk | |
| 16:04 | • Edit [b0e0b307c6]: Add "landing" with value "309a3bb9cc6381ad7b82f56ee68b6309270bd235f7b983bf117ce961a5320ba6". artifact: 63716d7d11 user: quarry | |
| 16:04 | • Reply: The dashboard feed showed landing events but its kind chips had no landings entry. me.feed now counts the landing record and the replies on it as one landings chip; the l filter already reached both in the index. Test folds a landing thread with a reply and expects one chip counting two. artifact: b2978f5973 user: dmestas | |
| 16:04 | • Edit [54ff2a1ebd]: Mark "Closed". artifact: 91d86abfa4 user: quarry | |
| 16:03 | • Edit [764e467a4d]: Add "ci-pass" with value "abaff577cf1243fc465864426b4fd2c0". artifact: 3e3bca01b2 user: quarry | |
| 16:03 | land ci-run-page check-in: 764e467a4d user: dmestas tags: trunk | |
| 16:03 | • Edit [764e467a4d]: Add "landing" with value "7151ce2f14d51b830bced1fb88eb81253c806fe61a8f47fadc85144135116931". artifact: cb6e405302 user: quarry | |
| 16:03 | • Reply: ci: the run page reads like a run — commit subject, log filter, expand, wrap, raw artifact: 61dced9738 user: dmestas | |
| 16:03 | • Edit [ed072e1bdb]: Mark "Closed". artifact: 36ead5e555 user: quarry | |
| 16:03 | • Edit [ed072e1bdb]: Add "ci-pass" with value "216fec362e76ea21ca396998b65eea1a". artifact: 1d34b205c0 user: quarry | |
| 16:02 | • Post: ci: the run page reads like a run — commit subject, log filter, expand, wrap, raw artifact: 7151ce2f14 user: dmestas | |
| 16:02 | • Edit [ed072e1bdb]: Edit check-in comment. artifact: fb85896650 user: dmestas | |
| 16:02 | ci: the run page reads like a run — commit subject, log filter, expand, wrap, raw closed check-in: ed072e1bdb user: dmestas tags: ci-run-page | |
| 15:59 | • Post: The dashboard feed showed landing events but its kind chips had no landings entry. me.feed now counts the landing record and the replies on it as one landings chip; the l filter already reached both in the index. Test folds a landing thread with a reply and expects one chip counting two. artifact: 309a3bb9cc user: zeitgeist | |
| 15:59 | The activity feed offers a landings chip, counting records and replies together closed check-in: 54ff2a1ebd user: dmestas tags: feed-landings-chip | |
| 15:53 | • Edit [e5d92f969c]: Add "ci-pass" with value "236b2b5953f174c6e24c23df7e3da4fc". artifact: 73225b6e3a user: quarry | |
| 15:52 | land zit-0.3.6 check-in: e5d92f969c user: dmestas tags: trunk | |
| 15:52 | • Edit [e5d92f969c]: Add "landing" with value "27fd67e616cb4d015480988aaa73c7732b4ec4a444aff1b31d97cf4ce78a72dc". artifact: 40fcea427b user: quarry | |
| 15:52 | • Reply: Pin zit 0.3.6. It reads the Content-Type media type rather than the whole header (zit#179), so servers that answer with a charset parameter, gaia-gis.it among them, clone instead of failing with a size-limit error. The 0.3.6 Linux tarball is staged on the host and verifies; the full suite is green locally against zit 0.3.6. After the deploy, libspatialite gets re-added as a mirror. artifact: a0093e38fa user: dmestas | |
| 15:52 | • Edit [c52bd26dd7]: Mark "Closed". artifact: b7d455775d user: quarry | |
| 15:50 | • Edit [c52bd26dd7]: Add "ci-pass" with value "410e3a126588da9dbb5980cb7cdfc5d8". artifact: 38fd951fd7 user: quarry | |
| 15:49 | • Post: Pin zit 0.3.6. It reads the Content-Type media type rather than the whole header (zit#179), so servers that answer with a charset parameter, gaia-gis.it among them, clone instead of failing with a size-limit error. The 0.3.6 Linux tarball is staged on the host and verifies; the full suite is green locally against zit 0.3.6. After the deploy, libspatialite gets re-added as a mirror. artifact: 27fd67e616 user: zeitgeist | |
| 15:49 | Pin zit 0.3.6 for the Linux build (clones servers that answer with a charset on the media type, zit#179) closed check-in: c52bd26dd7 user: dmestas tags: zit-0.3.6 | |
| 15:48 | • Edit [8cf40c0aa9]: Add "ci-pass" with value "cb32dae8bfb3511a4e4d3280ef75a9ab". artifact: 2581b8ab9f user: quarry | |
| 15:47 | land ci-log-stream check-in: 8cf40c0aa9 user: dmestas tags: trunk | |
| 15:47 | • Edit [8cf40c0aa9]: Add "landing" with value "cbb8842df0035ebd31a6dc51865353a5f6307bf2425df31e6c7b595b27d8eeb3". artifact: b178f86ef8 user: quarry | |
| 15:47 | • Reply: landing: the decision card streams while CI runs, and reloads without script artifact: e229282e5b user: dmestas | |
| 15:47 | • Edit [1b4423c468]: Mark "Closed". artifact: 3ff9cb0db0 user: quarry | |
| 14:30 | • Edit [1b4423c468]: Add "ci-pass" with value "3c34332db43292b9c23e339184b90ad4". artifact: a4d1a73c58 user: quarry | |
| 14:29 | • Post: landing: the decision card streams while CI runs, and reloads without script artifact: cbb8842df0 user: dmestas | |
| 14:29 | landing: the decision card streams while CI runs, and reloads without script closed check-in: 1b4423c468 user: dmestas tags: ci-log-stream | |
| 14:12 | • Edit [677bed2a7a]: Add "ci-pass" with value "beedb61b306f691aa511e3ded78d8294". artifact: f76ca006d0 user: quarry | |
| 14:12 | Create new branch named "ci-log-stream" check-in: 677bed2a7a user: dmestas tags: ci-log-stream | |
| 14:10 | • Edit [8afe5afaf9]: Add "ci-pass" with value "cee8aeba647e13424cec015337664a5d". artifact: 0f515ff8bd user: quarry | |
| 14:10 | land fleet-edit-mirror check-in: 8afe5afaf9 user: dmestas tags: trunk | |
| 14:10 | • Edit [8afe5afaf9]: Add "landing" with value "1226a6dc082c1c9177e3f420672db9582e9adf79606b5bcffcfa1cf69ce07b09". artifact: 737eed77c5 user: quarry | |
| 14:10 | • Reply: fleet: edit mirror link selects the Mirrors tab, where the form lives artifact: cb589a0965 user: dmestas | |
| 14:10 | • Edit [ef78b6d72f]: Mark "Closed". artifact: bf4b0547e6 user: quarry | |
| 12:39 | • Edit [ef78b6d72f]: Add "ci-pass" with value "f67aab1080fbbdd7c0f408afa67fc7bb". artifact: 70ce10492d user: quarry | |
| 12:38 | landing: decision controls skip the override validation that silently blocked them closed check-in: ef78b6d72f user: dmestas tags: fleet-edit-mirror | |
| 12:30 | • Edit [ce6f05ea98]: Add "ci-fail" with value "396a13e999674abc633db7760c2b4dea". artifact: 78c4d33e0e user: quarry | |
| 12:29 | • Edit [52008b3274]: Add "ci-pass" with value "7740a6e4a6d3241336832a40393b6a82". artifact: e2445ef393 user: quarry | |
| 12:29 | land mirror-orphan check-in: 52008b3274 user: dmestas tags: trunk | |
| 12:29 | • Edit [52008b3274]: Add "landing" with value "542c226288b027bc8454f0cae08a922b6b61b900632b257e61ffd13f58a556b1". artifact: 2ec40f288b user: quarry | |
| 12:29 | • Reply: Two thousand exceptions on zeitforge.app came from one refused clone. Importing libspatialite recorded its mirror before the clone ran; the clone failed (zit refuses gaia-gis.it's reply size, filed as danmestas/zit#179), the import was marked failed and its directory removed, but the mirror row stayed. The scheduler then asked for a repository that did not exist on every one-minute tick, and because the lookup failed before the run was recorded the row never became due-at-interval: one journalled error per minute, forever. Now a failed import forgets its mirror, and PullMirror records a failed run even when the store is missing so any such row backs off to its interval. The orphan row on production was removed by hand and the errors have stopped. Test covers the run being recorded; the registrar path has no harness and is four lines. artifact: 899476655b user: dmestas | |
| 12:29 | • Edit [e097c7d028]: Mark "Closed". artifact: 91ca66ef67 user: quarry | |
| 12:27 | • Edit [e097c7d028]: Add "ci-pass" with value "244989e19fe9539322ac6f3b58be56d5". artifact: bd5d710712 user: quarry | |
| 12:27 | • Post: fleet: edit mirror link selects the Mirrors tab, where the form lives artifact: 1226a6dc08 user: dmestas | |
| 12:27 | • Edit [ce6f05ea98]: Edit check-in comment. artifact: 0e18b7b26e user: dmestas | |
| 12:26 | fleet: edit mirror link selects the Mirrors tab, where the form lives check-in: ce6f05ea98 user: dmestas tags: fleet-edit-mirror | |
| 12:24 | • Post: Two thousand exceptions on zeitforge.app came from one refused clone. Importing libspatialite recorded its mirror before the clone ran; the clone failed (zit refuses gaia-gis.it's reply size, filed as danmestas/zit#179), the import was marked failed and its directory removed, but the mirror row stayed. The scheduler then asked for a repository that did not exist on every one-minute tick, and because the lookup failed before the run was recorded the row never became due-at-interval: one journalled error per minute, forever. Now a failed import forgets its mirror, and PullMirror records a failed run even when the store is missing so any such row backs off to its interval. The orphan row on production was removed by hand and the errors have stopped. Test covers the run being recorded; the registrar path has no harness and is four lines. artifact: 542c226288 user: zeitgeist | |
| 12:24 | A failed import forgets its mirror, and a mirror with no store records its failed run: no more one error per tick for a clone that never happened closed check-in: e097c7d028 user: dmestas tags: mirror-orphan | |
| 12:16 | • Edit [2e17f29e28]: Add "ci-pass" with value "b5131bed51bb7e55a72461538a123014". artifact: da2ecae480 user: quarry | |
| 12:15 | land zit-0.3.5 check-in: 2e17f29e28 user: dmestas tags: trunk | |
| 12:15 | • Edit [2e17f29e28]: Add "landing" with value "b993e1dd0876464ddc730b3b6ed530970e3f3540f48c82c021f8988db4a84de0". artifact: a5419722d1 user: quarry | |
| 12:15 | • Reply: Pin zit 0.3.5 for the Linux build. The 0.3.5 tarball is staged on the host and verifies against SHA256SUMS; the full suite, including the zitcli contract canaries, is green locally against zit 0.3.5. The next deploy installs it beside quarry and the footer will say so. artifact: c31b093e10 user: dmestas | |
| 12:15 | • Edit [501b055504]: Mark "Closed". artifact: bc63ac60cd user: quarry | |
| 12:12 | • Edit [501b055504]: Add "ci-pass" with value "a632548aea14cd6aff60310a0113c861". artifact: 858985a9db user: quarry | |
| 12:12 | • Post: Pin zit 0.3.5 for the Linux build. The 0.3.5 tarball is staged on the host and verifies against SHA256SUMS; the full suite, including the zitcli contract canaries, is green locally against zit 0.3.5. The next deploy installs it beside quarry and the footer will say so. artifact: b993e1dd08 user: zeitgeist | |
| 12:12 | Pin zit 0.3.5 for the Linux build closed check-in: 501b055504 user: dmestas tags: zit-0.3.5 | |