7f5a0cb63c
Answer prose round 1: one name for a place holding most of a postal code's land, the decision leaves the seeding to the code, and the test says what it checks
lilleman2026-10-11 16:38:42 +02:00
9360447254
Answer stability round 2: the test names a missing population column
lilleman2026-10-11 16:38:55 +02:00
66b5d699fa
Answer stability round 1: each US place's draw is seeded by its GEOID, so a changed source moves only the places near the cut; item 78 leaves the plan and the multi-county fetch is filed
lilleman2026-10-11 16:33:23 +02:00
121f9549f2
Test the US places' reach by the locality table's population column, read by its header
lilleman2026-10-11 16:32:19 +02:00
63a7ac156e
Record the reading cost of the last chunk and refresh item 167
lilleman2026-10-11 16:20:39 +02:00
5db699469b
Record the shipped places and streets as a decision, and say so in the README and changelog
lilleman2026-10-11 16:20:23 +02:00
99ba0b30ee
Re-pin the seeded sv_SE.address example to the new streets
lilleman2026-10-11 16:19:05 +02:00
88cea644b6
Ship a tenth of each place's streets, at least ten, and a tenth of the US places owning a ZCTA, drawn by population
lilleman2026-10-11 16:18:32 +02:00
b64b73cb9a
Test that a city ships a tenth of its streets and the US places reach below a town
lilleman2026-10-11 16:07:35 +02:00
2f8fffd237
Record the maintainer's decision to ship item 182 over the ruling at a632a92 with Shape fallen, and file item 197 to replace the namespace mount
lilleman2026-10-11 15:59:27 +02:00
33ddb49127
Answer architecture round 12: the namespace mount serves what a walk asks of a tree, a folder's entries and a file's bytes, so the folders above it need no file of their own
lilleman2026-10-11 15:45:44 +02:00
5b1f80c350
Test that a mounted tree lists and reads its files under its mount
lilleman2026-10-11 15:45:31 +02:00
e8f5fd47d5
Answer architecture round 11: a source states its mount in a field its manifest, its labels and its file reads share, so the mount can come before or after the manifest; the README says a stand-in for the places is a record with a format
lilleman2026-10-11 15:13:39 +02:00
ad1dfac447
Test that a mounted source names and reads its manifest beside its files, before or after the mount
lilleman2026-10-11 15:13:07 +02:00
d157e4526a
Answer architecture round 10: the paired ruling at fb2c809 is better; a mounted module names its manifest from its files' root, and the mount wrapper returns nothing beside an error
lilleman2026-10-11 15:10:23 +02:00
f89540e973
Answer product-owner round 6: the options table spreads Modules(), and the Library section shows two locales and what a stand-in for the places provides
lilleman2026-10-11 15:10:23 +02:00
230faf58a8
Answer stability round 8: a mounted data path names a file where it is on disk
lilleman2026-10-11 15:10:23 +02:00
ec865cfe53
Test that a mounted data path names a broken file where it is on disk
lilleman2026-10-11 15:09:44 +02:00
d242466859
Answer architecture round 9: each data package exports Modules, a locale's importing its country's geo package as a bench chose, and a namespace mounts a tree through an fs.FS wrapper, so internal/datafiles walks file-system folders alone
lilleman2026-10-11 15:05:56 +02:00
d4dee46086
Test that each shipped package's Modules holds what its categories read and renders them, and that a mounted tree is a valid fs.FS
lilleman2026-10-11 15:05:56 +02:00
203486a9d3
Answer architecture round 8: a shipped package's folder spells its namespace (data/sv_SE, data/geo/SE) under its lowered Go name, as a second bench chose, a mounted root folder reads as its namespace, and item 196 cites goal 9.2
lilleman2026-10-11 14:11:46 +02:00
26f3c95031
Test that each shipped module sits in the folder its namespace spells, that a locale alone fails at its address naming the missing geo, and that a mounted root folder is named by its namespace
lilleman2026-10-11 14:11:46 +02:00
cc1c3bbc7d
Answer architecture round 7: reads and its note move to item 196, a mounted module's errors read sv_SE (address.json), internal/datafiles handles namespace folders in one form, and a folder outside the mount is a broken invariant
lilleman2026-10-11 13:23:30 +02:00
6a62840373
Test that a mounted module names a broken file beside its namespace, that a manifest refuses reads, and that each shipped module renders beside the modules its categories reach
lilleman2026-10-11 13:23:30 +02:00
89cafb1a21
Answer architecture round 6: a manifest's reads names namespaces, loadReached notes only those nothing loaded provides, a mounted module labels its files by its namespace, Source.Load checks its mount, and the README keeps namespace and reads in an edited copy
lilleman2026-10-11 12:41:11 +02:00
3750770735
Test that a typo under loaded modules names no module to add, that reads names namespaces, that a mounted module names its files by its namespace, and that a namespace reaches a subfolder
lilleman2026-10-11 12:41:10 +02:00
7b70530a8a
Answer architecture round 5: the shipped packages are enus, svse, geose, geous and misc, each holding its files at its root under its manifest's namespace, and the errors use the wording a bench of the Audience chose
lilleman2026-10-11 12:15:41 +02:00
29a76d5d9a
Test that a manifest's namespace mounts its source under a dot path, that each shipped module loads from data/<lowered namespace>, and that a load failure's note reaches Fake, FakeTemplate and FakeStruct once
lilleman2026-10-11 12:15:41 +02:00
165b67e843
Answer architecture round 4: New keeps every module the loaded manifests read, and the note on a failed load names them all from one function beside noDataNote, so no category carries its source's modules
lilleman2026-10-11 11:49:18 +02:00
6e2555aee6
Test that a load failing on a read nothing provides names every module the loaded manifests read
lilleman2026-10-11 11:49:18 +02:00
7f7d260f83
Answer prose round 1: errNoData, the reads paragraph and the changelog entry read in one pass, the Library section shows the import block for one locale, and the decisions state why New checks no reads and which goal the embed serves
lilleman2026-10-10 13:33:26 +02:00
4caea56663
Answer stability round 2: New names every nil module WithDataFS was given
lilleman2026-10-10 13:30:09 +02:00
e6e7b804e1
Answer product-owner round 1: the Versioning table names the data packages, the README pairs each locale with its geo module, WithDataFS refuses a nil module, errNoData says New was given no data, and three changelog lines the move made false are true again
lilleman2026-10-10 13:28:42 +02:00
854d4d69ae
Answer stability round 1: a manifest's reads refuses a repeated module, and the test re-checking what the manifest tests pin goes
lilleman2026-10-10 13:28:42 +02:00
97ff2dfaa1
Test that WithDataFS refuses a nil module and a manifest's reads refuses a repeated module
lilleman2026-10-10 13:28:42 +02:00
949d2709da
Answer architecture round 3: the paired ruling at 70e552c against bf1becb is same
lilleman2026-10-10 13:24:41 +02:00
3da487f8ea
Answer architecture round 2: each category site carries its source's default modules, set where the walk or the index loads it, and the note is written beside the resolve that finds the read; the decision states why New checks no reads
lilleman2026-10-10 13:20:26 +02:00
2ddba2ce96
Test a broken reads in a manifest without an index
lilleman2026-10-10 13:20:26 +02:00
50e1a9fb22
Answer architecture round 1: a module's reads sits on its index entry or, for a source without an index, in the load that walks it, the note names default modules, the shipped modules are listed by name, datafiles drops its unused base, and item 195 files the builtins' second parse
lilleman2026-10-10 13:16:59 +02:00
e8b7d95559
Test that each shipped module renders beside the modules it reads, that the test's module list matches data.Modules(), and that a data path's manifest and an indexed one each name what they read
lilleman2026-10-10 13:16:59 +02:00
d79be8b50a
Record the reading cost of the last chunk, and refresh item 167's figures
lilleman2026-10-10 13:08:19 +02:00
b8eba1eb35
Ship en_US, sv_SE, misc, geo/SE and geo/US as Go packages exporting FS, listed by data.Modules(), and fail a bare New naming WithDataFS
lilleman2026-10-10 13:08:01 +02:00
3368f085fd
Test that the shipped data loads from a Go package per module, that a bare New fails naming WithDataFS, and that a read nothing provides names the modules its reader's manifest reads
lilleman2026-10-10 13:08:01 +02:00
bf1becbb59
Rescore: the paired ruling at b77ec94 against 2652779 is better
lilleman2026-10-10 11:21:22 +02:00
37b1e370b9
Answer product-owner round 2: the parent error says the source loads whole at start, and the exit codes name a missing -d folder
lilleman2026-10-10 11:18:44 +02:00
fcd7ebf736
Answer stability round 1: a manifest refuses an entry inside another category and a path an entry repeats
lilleman2026-10-10 11:16:52 +02:00
ce2da330f7
Test that a manifest refuses an entry inside another category and a repeated path
lilleman2026-10-10 11:16:07 +02:00
13e3de39fe
Answer product-owner round 1: the parent error and the README name the way to link under another source's table, a missing indexed file names the manifest, the README warns a copied data/ carries its index, and the docs say what a load failure and a replacement do
lilleman2026-10-10 11:14:19 +02:00
15477e989f
Test that an indexed category with no file names the manifest, and that an entry omitting a parent says it names none
lilleman2026-10-10 11:13:18 +02:00
a6d9623383
Record the paired ruling at d34cc42, move jsonKind's comment onto Kind, name the manifest in Layout, and note the normalised manifest in item 185
lilleman2026-10-10 11:10:13 +02:00
55d503317d
Answer architecture round 3: New loads only sources without an index and what they read, an entry drops reads, ErrLoad sits beside New, the manifest decoder moves to manifest.go and names each mistake once as jsonvalue.Kind words it
lilleman2026-10-10 11:07:16 +02:00
2cb124b8a7
Test that a replacement breaking a shipped reader fails that reader's first reach, that New's load errors match ErrLoad, and that a manifest without entry reads names every mistake once
lilleman2026-10-10 11:04:40 +02:00
aafcbd9025
Record the paired ruling at 0bdb100, and mark item 194 as the decision log cites it
lilleman2026-10-09 19:39:45 +02:00
498f81075e
Answer architecture round 2: a first reach checks only a table's parent, loadReached splits into a batch that loads, runs and puts back, the manifest decodes strictly and reports every mistake, and ErrLoad marks a load failing in New too
lilleman2026-10-09 19:36:27 +02:00
3c1af91433
Test that a manifest refuses every key and shape it does not define, that a table under a shipped table loads from a data path, that an omitted parent fails the first reach, and pin New's allocations
lilleman2026-10-09 19:36:27 +02:00
c783d4af23
Answer architecture round 1: one owner keeps a folder's loaded and unloaded names apart, each loaded category is checked against its index entry, addSource returns what it replaced, the manifest cache goes, ErrLoad exits 1, and the docs say what New and Fake now refuse
lilleman2026-10-09 19:26:54 +02:00
642fe3a506
Test that a stale index entry fails its first reach naming the manifest, that such a failure wraps ErrLoad and exits 1, and that a data path naming a file says it is no directory
lilleman2026-10-09 19:26:54 +02:00
2e56af5d63
Record the reading cost of the last chunk
lilleman2026-10-09 19:11:55 +02:00
c320344fac
Load each category of a source whose manifest carries an index on first reach, and move the shipped index into data/.fejkdata.json
lilleman2026-10-09 19:11:37 +02:00
7d5b32fe3f
Test that a source whose manifest carries an index loads each category on first reach, and that a data path loads only what it reads and what reaches what it replaces
lilleman2026-10-09 19:04:46 +02:00
8d309cd827
Rescore: the paired ruling at 8b5a972 against 2652779 is the same
lilleman2026-10-09 18:54:01 +02:00
f663e1af1e
Record the maintainer's answers: a manifest names no fejkdata version and item 183 goes, and the shipped data is held under 100 MiB by item 192, past which item 193 carves out Go modules
lilleman2026-10-09 18:52:50 +02:00
3744b60f7f
Answer goals audit round 1: item 182 names the import guard and the embedded manifest, and item 183 says Go module
lilleman2026-10-09 18:51:02 +02:00
687a3d37bf
Answer prose round 1: module means fejkdata's unit and Go module the Go one, only anyone else's manifest names a version, goal 2.3 says a program links what it imports, and a carved-out module requires the core release that drops its package
lilleman2026-10-09 18:49:59 +02:00
22ca7caa40
Record the reading cost of the last chunk
lilleman2026-10-09 18:47:24 +02:00
3e12849f65
Ship the data as Go packages in the core's module: goal 2.3 and the shipped-data decision say packages, item 179 goes, items 180 to 186 follow, and item 191 files the release tag-guard bug
lilleman2026-10-09 18:47:23 +02:00
dc32e24188
Plan item 190: decide whether fejkdata pseudonymises real datasets
lilleman2026-10-09 15:47:34 +02:00
b08b3e94d5
Answer prose round 1: the exact-match reason reads one idea a sentence, and item 180 says its goal-13 effect once
lilleman2026-10-09 14:57:58 +02:00
c597d0a2e0
Record the maintainer's answers: an exact version match, goal 6.3 for modules that load on first reach, and items 79 and 54 folded into items 180 and 179
lilleman2026-10-09 14:55:37 +02:00
24a30161f7
Answer prose round 1: the decision keeps only what the goals leave open, and the new items' taglines read once
lilleman2026-10-09 14:42:27 +02:00
65ccc132ae
Record the reading cost of the last chunk
lilleman2026-10-09 14:40:17 +02:00
9558472c96
Split item 77 into items 179 to 189, and record its module decision in docs/decisions.md
lilleman2026-10-09 14:40:13 +02:00
fa734fda3c
Answer prose round 1: AGENTS.md names the paired ruling as what still blocks a 0.1.0 merge, item 107 promises what item 178 plans, and the floor entry says one thing a sentence
lilleman2026-10-09 14:14:29 +02:00
30a4445a47
Ship 0.1.0 under the comprehension floor, drop item 176, and plan item 107 for 0.2.0 behind an architect's plan for grouping and untangling the engine
lilleman2026-10-09 14:12:37 +02:00
265277953c
Rescore: the paired ruling at 344023d against 1c342bc is better
lilleman2026-10-09 13:09:34 +02:00
edec28f58f
Answer prose round 1: the Data paragraph names the set first, the choice comment says what shared counts, and the plan's new items read once
lilleman2026-10-09 13:07:59 +02:00
0243e8d42d
Answer product-owner round 1: the Data section names the geo trees, and Layout says what the docs records hold
lilleman2026-10-09 13:06:05 +02:00
18a264db59
Answer stability round 1: the choice comment defines shared, and names none of its readers
lilleman2026-10-09 13:06:05 +02:00
fd6d80e2c2
Wrap the README's Data paragraph
lilleman2026-10-09 13:04:43 +02:00
f30bdc1380
Record the reading cost of the chunk before item 153
lilleman2026-10-09 13:04:18 +02:00
5d85944359
Record the 2026-10-09 panel's scores, file what it named as items 170 to 177, and correct four claims it found false
lilleman2026-10-09 13:03:51 +02:00
b9d1af6565
Answer prose round 1: loadCallerPath's name and docstring say it loads, and that the caller walks next
lilleman2026-10-09 12:39:56 +02:00
51c0bd518e
Record the reading cost of the chunk before item 151
lilleman2026-10-09 12:37:43 +02:00
00f31e6f59
Load a caller's path in the one method that splits it, so Fake and FakeRecord walk only what has loaded
lilleman2026-10-09 12:37:31 +02:00
1c342bc2af
Rescore: the paired ruling at 8433689 against d74c0ed is better
lilleman2026-10-09 12:29:19 +02:00
a318dc7f42
Answer prose round 1: fence's docstring splits rule, shapes and result, each drop message names what was missing, the US geo row says a place needs a street, and file item 169
lilleman2026-10-09 12:27:58 +02:00
32e01565b4
Answer architecture round 3: fence's docstring says a locality needs entries in both child maps, and zcta_localities says what it maps to
lilleman2026-10-09 12:26:00 +02:00
c032cb779c
Answer architecture round 2: street.tsv is written from the kept localities, fence checks a child holds rows, and its names say what they hold
lilleman2026-10-09 12:21:14 +02:00
b49e0cac6e
Rescore: the paired ruling at c27e8b7 against d74c0ed is better
lilleman2026-10-09 12:16:38 +02:00
71b17d104b
Answer architecture round 1: fence keeps a locality holding both a street and a postal code, returns what it dropped so the SE import still exits on a lost kommun and the US one counts its counties, and file item 168 for data-import tests
lilleman2026-10-09 12:16:33 +02:00
4f369b961c
Record the reading cost, and refresh item 167's figures
lilleman2026-10-09 12:09:00 +02:00
6d285249e4
Meet the parent-has-a-child fence in one shared geo.fence step, which states the locality row each country hands it
lilleman2026-10-09 12:08:54 +02:00
825b01408b
Answer prose round 1: the skips comment holds for routeSteps, the finder is named for what it finds, and the failure names which list to extend
lilleman2026-10-09 11:57:18 +02:00
5945a32bc3
Answer stability round 1: a method is keyed by its receiver, and panicsOnOtherNodeKinds names what it checks
lilleman2026-10-09 11:56:02 +02:00
ae5a4e5a6c
Record the reading cost, and refresh item 167's files
lilleman2026-10-09 11:53:42 +02:00
8338f0a5e3
Find every node switch with a panicking default from the source, and test routeSteps beside the seven listed by hand
lilleman2026-10-09 11:53:27 +02:00
588cb328bf
Rename item 167 to the title larv-review now files
lilleman2026-10-09 11:48:16 +02:00
lilleman