Repository navigation
docs: record libtailscale#61 and its tracking issue - #59
Merged
Merged
Conversation
Adds the MACOS_TARGET override fix to the Upstream sections of README.md and TAILSCALE.md, alongside its tracking issue tailscale#21044, following the same shape as #57/#58/#59/#60. The entry says plainly that this repo is NOT affected — build-tailscalekit.sh already passes MACOS_TARGET on the command line — because an upstream list that implies a carried burden where none exists is worse than no entry. What it records is why the bug is worth fixing for everyone else: the Go archive and the Swift target must move together, so lowering only one stamps the framework with a floor its objects do not support, otool cannot detect it (it reports the number the binary claims), and the only signal is a single `ld: warning` line. TAILSCALE.md's "Deployment floors" section already documented the trap as standing advice. It now points at the proposed fix and says the paragraph becomes history if #61 lands, so the advice does not outlive the bug. Two other states refreshed while in here, both verified against GitHub today rather than assumed: - the "states verified" date moves 2026-08-28 -> 2026-08-30 - #60 gains its runtime evidence: it was previously recorded as measured by building at each floor, which is compile-time only. On an iOS 17.5 simulator the app now demonstrably loads a framework built with that diff while the pre-gating build is refused by dyld, reproducible via tailscale/verify-floor-runtime.sh Co-Authored-By: Claude <noreply@anthropic.com>
prakashrj
added a commit
that referenced
this pull request
Aug 31, 2026
…ean (#62) libtailscale merged three on 2026-08-31: 8564835 #60 gate the listener API -> retires 0003 61e8513 #58 run tailscale_up off the actor -> retires 0002 59d4bb8 #61 MACOS_TARGET ?= 15.0 -> never carried here The important part is what has NOT changed. All three sit *ahead of* the commit `vendor/libtailscale` pins (5e89501), so this repo still applies 0002 and 0003 and will keep doing so until the submodule is bumped. Recording them as simply "merged" would imply the patches are gone, which is exactly the kind of claim this section already refuses to make for tailscale#20985 — merged there too, and carried by no release. So each entry now says three things rather than one: it merged, which commit, and that the pin predates it. The delete-do-not-rebase instruction is kept and made conditional on the bump, because that is the moment it becomes actionable. #57 remains open with changes requested and still gates v0.2. #59 remains open and is still not on the critical path. Also refreshes the "states verified" date and extends its caveat: whether a fix has *shipped* was already distinguished from whether it merged, and now so is whether it has reached this repo's pinned submodule. Co-authored-by: Claude <noreply@anthropic.com>
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Adds the
MACOS_TARGEToverride fix to the Upstream sections ofREADME.mdandTAILSCALE.md, alongside its tracking issue, following the same shape as #57/#58/#59/#60.The entry says this repo is not affected
build-tailscalekit.shalready passesMACOS_TARGETon the command line, so nothing is carried and nothing needs deleting when it lands. That's stated explicitly, because an upstream list implying a carried burden where none exists is worse than no entry at all.What it does record is why the bug is worth fixing for everyone else: the Go archive and the Swift target have to move together, so lowering only one stamps the framework with a floor its objects don't support.
otoolcan't detect that — it reports the number the binary claims — and the only signal is a singleld: warningline.Advice that expires
TAILSCALE.md's "Deployment floors" section already documented the trap as standing advice. It now points at the proposed fix and says the paragraph becomes history if #61 lands, so the advice doesn't outlive the bug it describes.Two states refreshed while in here
Both verified against GitHub today rather than assumed:
built for iOS-sim 18.1 which is newer than running OS). Reproducible withtailscale/verify-floor-runtime.sh.Docs only — no code changes.