Commit cca78fa
committed
fix(scripts): stop unconfigured siblings from falsely claiming the sync port base
Independent review caught a real logic defect: readSyncPort defaulted to
SYNC_PORT_BASE whenever a sibling worktree had no sync.port set at all,
so every worktree that hasn't run sync yet (the common case — sync: stays
absent until mongo.uri is present) looked like it was actively occupying
port 3010 to every other worktree's conflict check. That's the exact
opposite of the comment above findNextSyncPort, which assumes the base is
usually free since most worktrees have never had one assigned.
readSyncPort now returns null when sync.port isn't actually configured,
and readSiblingSyncPorts filters those out, so only a worktree that has
genuinely claimed a sync port counts as a claim.1 parent 8c3dc9c commit cca78fa
2 files changed
Lines changed: 17 additions & 10 deletions
| Original file line number | Diff line number | Diff line change | |
|---|---|---|---|
| |||
35 | 35 | | |
36 | 36 | | |
37 | 37 | | |
38 | | - | |
39 | | - | |
| 38 | + | |
| 39 | + | |
| 40 | + | |
| 41 | + | |
40 | 42 | | |
41 | 43 | | |
42 | | - | |
43 | | - | |
| 44 | + | |
| 45 | + | |
44 | 46 | | |
45 | 47 | | |
46 | 48 | | |
| |||
| Original file line number | Diff line number | Diff line change | |
|---|---|---|---|
| |||
31 | 31 | | |
32 | 32 | | |
33 | 33 | | |
34 | | - | |
| 34 | + | |
| 35 | + | |
| 36 | + | |
| 37 | + | |
| 38 | + | |
35 | 39 | | |
36 | 40 | | |
37 | | - | |
| 41 | + | |
| 42 | + | |
38 | 43 | | |
39 | | - | |
| 44 | + | |
40 | 45 | | |
41 | 46 | | |
42 | 47 | | |
| |||
93 | 98 | | |
94 | 99 | | |
95 | 100 | | |
96 | | - | |
97 | | - | |
98 | | - | |
| 101 | + | |
| 102 | + | |
| 103 | + | |
99 | 104 | | |
100 | 105 | | |
101 | 106 | | |
| |||
0 commit comments