[dependabot]

Bumps the github-actions-major group with 1 update in the / directory: astral-sh/setup-uv.

Updates astral-sh/setup-uv from 7.6.0 to 10.3.3

Release notes

Sourced from astral-sh/setup-uv's releases.

v10.3.3 🌈 Use Astral mirror for version manifests with GitHub fallback

Changes

This release contains something that should have been included a very long time ago. Sorry for sleeping on that.

This action now retrieves the release manifest that contains the available versions, their download links and checksums from the Astral mirror, rather than trying GitHub first. This should present much more stable. (Known checksums are still checked, so the security impact does not change)

Also a big thank you to @​soujasK for a set of bugfixes 👏

🐛 Bug fixes

• Use Astral mirror for version manifests with GitHub fallback @​eifinger (#1119)
• fix: allow whitespace in uv requirement specifiers @​soujasK (#1112)
• fix: resolve ~= specifiers with PEP 440 semantics @​soujasK (#1111)
• fix: only report a saved cache when saving succeeded @​soujasK (#1110)

🧰 Maintenance

• chore: update known checksums for 0.13.6 @github-actions[bot] (#1117)

📚 Documentation

• docs: use JSDoc for compatible release specifier comment @​eifinger (#1113)
• docs: update version references to v10.3.2 @github-actions[bot] (#1104)

v10.3.2 🌈 update known checksums for 0.13.5

Changes

🧰 Maintenance

• chore: update known checksums for 0.13.5 @github-actions[bot] (#1103)

📚 Documentation

• docs: update version references to v10.3.1 @github-actions[bot] (#1102)

v10.3.1 🌈 preserve cache error class names in save-cache bundle

Changes

This release includes a fix so that multiple concurrent workflows trying to upload the same cache no longer warn

##[warning]Failed to save: Unable to reserve cache with key ..., another job may be creating this cache.

🐛 Bug fixes

• fix: preserve cache error class names in save-cache bundle @​dannyward630 (#1093)

🧰 Maintenance

... (truncated)

Commits

• 68559e9 Use Astral mirror for version manifests with GitHub fallback (#1119)
• 0192b69 chore: update known checksums for 0.13.6 (#1117)
• 2f29784 docs: use JSDoc for compatible release specifier comment (#1113)
• 9c73321 fix: allow whitespace in uv requirement specifiers (#1112)
• 80fbe01 fix: resolve ~= specifiers with PEP 440 semantics (#1111)
• 36b7b57 fix: only report a saved cache when saving succeeded (#1110)
• 56618ca docs: update version references to v10.3.2 (#1104)
• 07a047e chore: update known checksums for 0.13.5 (#1103)
• e1b8536 docs: update version references to v10.3.1 (#1102)
• 18d369b chore(deps): roll up Dependabot updates (#1101)
• Additional commits viewable in compare view

[github-actions]

DNS Resolver Benchmark

this PR @ c409bb6 against dev @ ab2568b, base → head, median of 3 interleaved rounds. Bold is significant across all 56 comparisons (Benjamini-Hochberg at 1%); the rest is noise.

regime	qps	unanswered	p99	attempts/query
ceiling/1-resolver	24,938 → 24,845 (-0.4%)	0.00% → 0.00%	0.800s → 0.803s (+0.4%)	1.00 → 1.00
pool/32-resolvers	24,621 → 24,583 (-0.2%)	0.00% → 0.00%	0.810s → 0.811s (+0.1%)	1.00 → 1.00
pool/32-persistent	23,899 → 24,803 (+3.8%)	0.00% → 0.00%	0.835s → 0.804s (-3.7%)	1.00 → 1.00
impaired/10%-loss	200 → 200 (+0.0%)	0.00% → 0.00%	8.134s → 8.134s (+0.0%)	1.11 → 1.11
impaired/20%-REFUSED	3,337 → 3,345 (+0.2%)	0.00% → 0.00%	0.599s → 0.597s (-0.2%)	1.24 → 1.24
impaired/50ms-latency	621 → 621 (+0.1%)	0.00% → 0.00%	3.220s → 3.218s (-0.1%)	1.00 → 1.00
impaired/20%-then-2.5%-dead-names	237 → 257 (+8.5%)	5.43% → 5.43%	10.660s → 10.250s (-3.8%)	1.05 → 1.05

All metrics

regime	p50	p100	sockets	conntrack
ceiling/1-resolver	0.440s → 0.434s (-1.5%)	0.802s → 0.805s (+0.4%)	42 → 42 (+0.0%)	17,038 → 12,580 (-26%)
pool/32-resolvers	0.441s → 0.442s (+0.1%)	0.812s → 0.813s (+0.2%)	73 → 73 (+0.0%)	19,292 → 19,072 (-1.1%)
pool/32-persistent	0.460s → 0.452s (-1.7%)	0.837s → 0.806s (-3.6%)	42 → 42 (+0.0%)	29 → 31 (+6.9%)
impaired/10%-loss	4.278s → 4.276s (-0.0%)	9.989s → 9.987s (-0.0%)	46 → 46 (+0.0%)	2,153 → 2,034 (-5.5%)
impaired/20%-REFUSED	0.413s → 0.414s (+0.1%)	0.599s → 0.598s (-0.2%)	46 → 46 (+0.0%)	2,316 → 2,313 (-0.1%)
impaired/50ms-latency	1.837s → 1.836s (-0.0%)	3.220s → 3.218s (-0.1%)	46 → 46 (+0.0%)	507 → 1,803 (+256%)
impaired/20%-then-2.5%-dead-names	3.771s → 4.160s (+10%)	12.676s → 11.678s (-7.9%)	39 → 39 (+0.0%)	2,929 → 2,919 (-0.3%)

[singlerider]

🟢 setup-uv 7 to 10 is used only in benchmark.yml (pull_request trigger, no custom manifest, no cache inputs). The breaking changes are: v8 drops floating major/minor tags (we pin a SHA), v9 stops pruning the cache by default (slightly more cache storage), and v10 disables the cache on pull_request_target (not used here). The benchmark ran green on this PR.

[pjhartlieb]

@dependabot rebase