Commit Graph

Select branches

Hide Pull Requests

master

Mono Color

• d9dd1435a6 What was run, and the one thing a green build does not mean Joseph Ferano 2026-09-20 19:56:24 +07:00
• a983a46ea5 The review's follow-ups: a file decides the shadow, and a set answers the membership Joseph Ferano 2026-09-20 19:51:52 +07:00
• 5cfefa586b The rebase's four repairs, and two pins that were not pinning anything Joseph Ferano 2026-09-20 19:54:53 +07:00
• 77d6e43b09 The agent binds before main, so a dev program need not start it at all Joseph Ferano 2026-09-20 19:52:34 +07:00
• 0d34831199 A trial puts the whole context back, and the compiler now insists on it Joseph Ferano 2026-09-20 19:51:51 +07:00
• 63fb5c629c The other backend runs the class thunk too, and now something says so Joseph Ferano 2026-09-20 18:45:35 +07:00
• 7f92136401 Old instances of a redefined class now follow the class Joseph Ferano 2026-09-20 18:41:14 +07:00
• e03028819c A defn named after a builtin now wins, and says so once Joseph Ferano 2026-09-20 19:28:00 +07:00
• e9c0e96a9b (agent/start) takes no argument, and takes the socket away after itself Joseph Ferano 2026-09-20 19:44:34 +07:00
• 114ea391aa One more dev-daemon row that raced, seen once in five Joseph Ferano 2026-09-20 19:38:37 +07:00
• 8ea1ff4481 The build artefact, out of the tree for good this time Joseph Ferano 2026-09-20 19:36:06 +07:00
• eb0d883e84 The suite is green on its own merits now, and the sweep says so against the new tip Joseph Ferano 2026-09-20 19:33:32 +07:00
• b4dbbb67ab The review's three findings, and the lanes this one landed on top of Joseph Ferano 2026-09-20 19:25:17 +07:00
• 657f640ec7 A reconsidered operand must leave nothing behind, and a literal is never reconsidered Joseph Ferano 2026-09-20 19:21:16 +07:00
• 3efa261539 A build artefact does not belong in the tree Joseph Ferano 2026-09-20 18:45:39 +07:00
• 13b391b7bf The u64 pin is spelled the way the program that motivated it is Joseph Ferano 2026-09-20 18:45:34 +07:00
• b34a7bbf11 The raylib half of the sweep, and the let the note named Joseph Ferano 2026-09-20 18:44:37 +07:00
• 3b91afd47b What the widening lane changed, kept, and measured Joseph Ferano 2026-09-20 18:41:28 +07:00
• f9ae500949 The lattice pinned at its edges, and the two calls that could go the other way Joseph Ferano 2026-09-20 18:37:55 +07:00
• 3e4267f57c Every no-implicit-widening comment now says what is true instead Joseph Ferano 2026-09-20 18:33:15 +07:00
• d0e33331b5 An expectation outranks the join, because an expectation is information Joseph Ferano 2026-09-20 18:19:39 +07:00
• 0c50f34916 Widening happens at expect, and the wider operand decides a binary op Joseph Ferano 2026-09-20 18:15:42 +07:00
• 080d294bc9 The lattice of conversions that cannot change the number Joseph Ferano 2026-09-20 18:08:16 +07:00
• b87ae11fa8 The editor protocol never waits for the agent, and a parked delivery installs first Joseph Ferano 2026-09-20 19:27:12 +07:00
• bbacbcb666 dead-beef's byte arm learns arity's new ctx argument Joseph Ferano 2026-09-20 19:27:06 +07:00
• 0a4c52d5ee filled and dead-beef, the two byte fills Joseph Ferano 2026-09-20 19:25:08 +07:00
• 0cc229dc6c trap_park stops dying on the abort race, and Closed means what it means Joseph Ferano 2026-09-20 19:24:47 +07:00
• f6416858bb defmacro takes a real parameter list, and [args] means the first argument Joseph Ferano 2026-09-20 19:23:01 +07:00
• d225aa4172 int and float are the two spellings i32 and f32 answer to Joseph Ferano 2026-09-20 19:19:42 +07:00
• 26242388b9 int and float are the machine types, not a second name for them Joseph Ferano 2026-09-20 19:09:26 +07:00
• 2afa554cce The flake note said noise; the full-suite numbers say otherwise Joseph Ferano 2026-09-20 19:02:53 +07:00
• 2240100ee9 Review follow-ups: the default's payload, four wrong reasons, a doubled prefix Joseph Ferano 2026-09-20 18:58:32 +07:00
• 5ea6884d2c The diagnostics pass: every message shows, explains, and names the fix Joseph Ferano 2026-09-20 18:49:12 +07:00
• 3c39354833 Measure the Wire.Closed flake rather than assume it Joseph Ferano 2026-09-20 18:46:58 +07:00
• 78d9a0f051 The review's fixes: a suggestion that does not compile, and a confident wrong guess Joseph Ferano 2026-09-20 18:46:48 +07:00
• d4def945a9 An array is a value you can write, not a place you have to fill first Joseph Ferano 2026-09-20 18:30:25 +07:00
• e807986622 The dogfood batch: empty forms, comment, inc and dec, guards, limits, shorthand Joseph Ferano 2026-09-20 18:34:13 +07:00
• 7bd2c99353 sentinel-filled is now dead-beef, and takes the pattern Joseph Ferano 2026-09-20 18:33:05 +07:00
• 097161fd41 The park left for a re-run without draining its ring Joseph Ferano 2026-09-20 18:32:01 +07:00
• 9fd1eb113f No sentence for a & that cannot arrive Joseph Ferano 2026-09-20 18:27:02 +07:00
• 92bf091361 A struct writes itself where its type is already known, and positionally Joseph Ferano 2026-09-20 18:23:42 +07:00
• 865c7e0445 C-c C-c on a defclass went to the expression evaluator and was refused Joseph Ferano 2026-09-20 18:23:33 +07:00
• b2d1df300b macros.flan says which of the two spellings it is Joseph Ferano 2026-09-20 18:23:22 +07:00
• eeac54a8e3 The parameter list, its tests, and the note about what it breaks Joseph Ferano 2026-09-20 18:21:54 +07:00
• de3a3ed3c8 The writeup said eleven checker rows; there are fourteen Joseph Ferano 2026-09-20 18:19:45 +07:00
• 69646e534e A macro's parameter list, and one grammar for it Joseph Ferano 2026-09-20 18:18:24 +07:00
• 1702a62308 Pin the () body guards, the match-arm rule, and correct two comments Joseph Ferano 2026-09-20 18:18:20 +07:00
• f4fcbe7745 Say what C-c C-i actually takes: the innermost thing, not the enclosing form Joseph Ferano 2026-09-20 18:17:48 +07:00
• 99f519ba6f Two byte fills: (filled BYTE) and (sentinel-filled) Joseph Ferano 2026-09-20 18:15:19 +07:00
• 16f90a7ef5 The dot advice says assign where the form is an assignment Joseph Ferano 2026-09-20 18:14:58 +07:00
• ce900acfe4 Six editor complaints from the dogfooding notes Joseph Ferano 2026-09-20 18:14:25 +07:00
• c1b18cb181 Two struct spellings the parser was never able to decide Joseph Ferano 2026-09-20 18:11:03 +07:00
• 89481cf8ec The wait for the agent socket was in front of the accept loop Joseph Ferano 2026-09-20 18:07:33 +07:00
• fa2b56ba5a Empty fn bodies, the pins for all six items, and the FIX.org entry Joseph Ferano 2026-09-20 18:11:20 +07:00
• 859e3aa7f4 The shadowed builtin, and's misdirected caret, and the expansion count Joseph Ferano 2026-09-20 18:11:16 +07:00
• ea84394dd0 Three more: the return slot, the case payload, the missing collection Joseph Ferano 2026-09-20 18:06:13 +07:00
• 6c017c2cf8 Six dogfooding items: empty bodies, comment, inc/dec, () bodies, type limits, {.field} Joseph Ferano 2026-09-20 18:02:59 +07:00
• 8cba440aca Four more from the worst-20: the constant, the let annotation, the quote, the operand Joseph Ferano 2026-09-20 18:01:08 +07:00
• 5afa707d76 The notes live in DISCUSS.org; the empty stub goes Joseph Ferano 2026-09-20 17:59:21 +07:00
• 1526b6fe3f Typed structs do not version; an evolving shape is a defclass Joseph Ferano 2026-09-20 17:58:26 +07:00
• 8add0093ba The defvar follow-ups, the spellings other languages use, and two register warts Joseph Ferano 2026-09-20 17:55:59 +07:00
• db2f6c69b9 The call argument names the parameter, and the condition states the rule Joseph Ferano 2026-09-20 17:44:11 +07:00
• 2b5d793d7a The dot habit, and a did-you-mean over values Joseph Ferano 2026-09-20 17:40:45 +07:00
• 6d82221978 A dyn trap that says where: the nine trapping operators take a site Joseph Ferano 2026-09-20 17:36:10 +07:00
• ac2af7c537 The diagnostics worklist, ranked, so it stops floating Joseph Ferano 2026-09-20 17:24:35 +07:00
• 36624d6e33 An abort's reply races the process it ends Joseph Ferano 2026-09-20 17:04:54 +07:00
• dc39631db7 M2 item 6: defclass, generic functions, and multimethods on the dyn side Joseph Ferano 2026-09-20 15:49:15 +07:00
• 0b4bf850ec The random exit 1 is test_dev's trap rows, and it predates this lane Joseph Ferano 2026-09-20 15:47:06 +07:00
• bd29eee3e4 The rebase is what was run, so it is what the note says Joseph Ferano 2026-09-20 15:38:19 +07:00
• ac7608f94b A method's parameter names are bound in parallel, not in sequence Joseph Ferano 2026-09-20 15:36:02 +07:00
• 90110f525a Classes and generic functions, written down Joseph Ferano 2026-09-20 15:07:34 +07:00
• 0511cc1a01 A class's name and its keyword are one dispatch value Joseph Ferano 2026-09-20 15:06:36 +07:00
• 52a1ea183a A method added to a running program, proved end to end Joseph Ferano 2026-09-20 15:02:39 +07:00
• 5a62770e52 Classes and generic functions, milestone 2's last item Joseph Ferano 2026-09-20 14:45:38 +07:00
• c4e07256db A three-element defvar reads its third element: a type is zeroed static, anything else is dyn Joseph Ferano 2026-09-20 15:14:56 +07:00
• 9321547822 A defvar's third element decides: a type is the static, anything else is dyn Joseph Ferano 2026-09-20 14:57:39 +07:00
• b2558f04de A defconst is a compiler const, on both backends and dyn is not one Joseph Ferano 2026-09-20 14:50:01 +07:00
• 49ee8774d3 The case inside a struct literal is the same unwritable value as a bare one Joseph Ferano 2026-09-20 14:46:53 +07:00
• ce93ac7622 A defconst's value is what the linker writes, on both backends Joseph Ferano 2026-09-20 14:31:32 +07:00
• e19dd5be29 The follow-up batch: honest prose, one prefix, one shared label Joseph Ferano 2026-09-20 14:39:35 +07:00
• e6ac833626 The follow-ups the day's reviews left behind, each re-verified Joseph Ferano 2026-09-20 14:30:05 +07:00
• 6baed9a0ba The numeric casts open dyn boxes, coercing across kinds with one warning per site Joseph Ferano 2026-09-20 14:11:04 +07:00
• 34790e78af The cast that opens a box, written down Joseph Ferano 2026-09-20 13:58:15 +07:00
• a45e3806af The cast that opens a box, pinned on both backends Joseph Ferano 2026-09-20 13:57:43 +07:00
• 10610fb76b The corpus program for a cast that opens a box Joseph Ferano 2026-09-20 13:54:07 +07:00
• b6ea14f930 A numeric cast opens a dyn box itself Joseph Ferano 2026-09-20 13:52:27 +07:00
• 2127d4d070 The transfer out of an initialiser is a restart's, not an unhandled signal's Joseph Ferano 2026-09-20 13:21:11 +07:00
• b0ee9a3ce0 The runtime structs are one table, and the mangles one module Joseph Ferano 2026-09-20 13:20:47 +07:00
• 0e9707c9ef The review batch: honest headers, named temps, the seqlock's odd window Joseph Ferano 2026-09-20 13:11:36 +07:00
• 811c37dca4 Two sentences said more than was checked Joseph Ferano 2026-09-20 13:09:04 +07:00
• 5b39730f07 The init-once flag moves out of the namespace a program can write Joseph Ferano 2026-09-20 13:07:09 +07:00
• dd239d6cd5 Two review findings this lane could not fix, written down Joseph Ferano 2026-09-20 13:00:55 +07:00
• 2d29528cad Three escape tables became two, and the docs catch up Joseph Ferano 2026-09-20 13:00:55 +07:00
• 6461322aa9 The length write-back moves back inside the seqlock odd window Joseph Ferano 2026-09-20 13:00:55 +07:00
• 2daf7e9164 A short-circuit temp stops calling itself a destructuring one Joseph Ferano 2026-09-20 13:00:55 +07:00
• 9eed4cbd8c The dyn header claimed a guarantee the linker does not give Joseph Ferano 2026-09-20 13:00:55 +07:00
• 1829cd43b6 One field list per runtime struct, one spelling per symbol prefix Joseph Ferano 2026-09-20 13:00:52 +07:00
• f030c5f7f1 Memory diagnostics on demand: gc and native allocation sites, faintly Joseph Ferano 2026-09-20 12:57:30 +07:00
• cb58288f4d The test binaries share one support module instead of ten forks Joseph Ferano 2026-09-20 12:49:32 +07:00
• 26439dc22e The runtime's twin emitters share their bodies, and six dead symbols go Joseph Ferano 2026-09-20 12:06:48 +07:00