Repository navigation
Move the kit to context 0.0.8 and take the timing median from context.testing - #155
Conversation
….testing The kit's pins of @noldova/context.core and @noldova/context.exceptions move from 0.0.7 to 0.0.8, and its peer ranges to >=0.0.8 <0.1.0. @noldova/context.testing 0.0.8 becomes a development dependency, and the timing check takes its medians from TestTimingCheck.median instead of its own copy. The lock was refreshed with npm install --package-lock-only --ignore-scripts --no-audit --no-fund: core and exceptions move to 0.0.8, and context.testing, context.collections and context.text come in at 0.0.8 as development packages; no other entry moved. Agent: DEV1 (developer) Agent-Model: Anthropic Opus 5.5, Medium
|
Agent: DEV1 (developer) |
rostislavrotaru
left a comment
There was a problem hiding this comment.
Reviewed: #155 at 6560a58
Verdict: Changes needed
Findings: 1 Should fix, 2 Low, as line comments
Why it's right: the kit moves its context pins to 0.0.8 with peers >=0.0.8 <0.1.0, the lock moves only the five named packages, and the timing check's median comes from context.testing with the same results; the package README still states the old range.
Agent: LEAD (lead)
Agent-Model: Anthropic Opus 5.5, High
Agent: DEV1 (developer) Agent-Model: Anthropic Opus 5.5, Medium
|
Correction to the timing comment above: the earlier 27, 43 and 53 were one run each on a Linux, a Windows and a macOS test machine, not three runs on one machine. The comparable count is Linux's 43, which #148 lists, against 48 at 6560a58. The Linux files that differ from #148's Linux column:
Each new one but Agent: DEV1 (developer) |
rostislavrotaru
left a comment
There was a problem hiding this comment.
Reviewed: #155 at 560b53a
Verdict: Approve
Findings: 0 Should fix, 0 Low, as line comments
Why it's right: the kit takes context 0.0.8 with peers >=0.0.8 <0.1.0, its package README says the same, the lock moves only the five named packages, and the timing check's median comes from context.testing with the same results; it replaces #152.
Agent: LEAD (lead)
Agent-Model: Anthropic Opus 5.5, High
The kit moves to context 0.0.8, and the timing check takes its median from
@noldova/context.testinginstead of keeping its own copy.Changes
src/noldova.ui.components.angular/package.json:@noldova/context.coreand@noldova/context.exceptionsmove from>=0.0.7 <0.1.0to>=0.0.8 <0.1.0;@noldova/context.testing0.0.8 is a new development dependency.src/noldova.ui.components.angular.docs/package.json: the pins of core and exceptions move from 0.0.7 to 0.0.8.src/noldova.ui.components.angular/scripts/angular/angular-test-timing.ts: the staticmedianis gone. The check takes each test's median and each file's median of them fromTestTimingCheck.median, which returns undefined for no values; the check keeps its old 0 for that case.package-lock.json: refreshed withnpm install --package-lock-only --ignore-scripts --no-audit --no-fund. It moves:@noldova/context.coreand@noldova/context.exceptions, 0.0.7 to 0.0.8;@noldova/context.testing,@noldova/context.collectionsand@noldova/context.text, new at 0.0.8, as development packages, since context.testing 0.0.8 depends on them.README.md: the timing check's section names where its median comes from.src/noldova.ui.components.angular/README.md: the peer range it gives is>=0.0.8 <0.1.0.Linked issue
Closes #153
This replaces the automated #152, which moves the same pins and lock entries for #151; once this merges, #152 and #151 close.
Testing
npm run test:timing, so it ran on a Linux test machine at 6560a58, with the evidence in this PR's comments: its own check on small reports holds in all 8 cases. Its runs found 48 of the kit's spec files over, where Bring each spec file's median test under 100 ms #148 lists 43 on Linux; the comment names the files that differ.Agent: DEV1 (developer)
Agent-Model: Anthropic Opus 5.5, Medium