mirror of
https://github.com/ancsemi/Haven.git
synced 2026-10-26 08:38:20 +00:00
ancsemi
31175ba712
The last line of .gitattributes was getting truncated mid-token to 'GUIDE.md text working-tree-enc' on save, dropping the 'oding=UTF-8 eol=lf' suffix and the trailing newline. Restored the full rule. Also normalized the file to LF line endings. The new '* text=auto eol=lf' rule a couple lines above is a one-shot fix for the LF-vs-CRLF churn that was generating phantom 'modified' status on ~120 files in working trees on Windows; with this rule in place, future checkouts normalize cleanly. No content-affecting change to the rest of the repo.
21 lines
1.1 KiB
Plaintext
21 lines
1.1 KiB
Plaintext
# Force Unix line endings for shell scripts so Docker builds
|
|
# work regardless of the user's core.autocrlf setting.
|
|
*.sh text eol=lf
|
|
docker-entrypoint.sh text eol=lf
|
|
Dockerfile text eol=lf
|
|
* text=auto eol=lf
|
|
|
|
# Force UTF-8 (with a working-tree BOM) on the public landing pages and
|
|
# anything else that holds emoji. PowerShell 5.1's Set-Content defaults to
|
|
# Windows-1252, which has silently corrupted these files multiple times
|
|
# (every emoji becomes "?" / every em-dash becomes 0x97). The BOM makes the
|
|
# encoding self-identifying so editors and PS5 won't round-trip them through
|
|
# cp1252. See AGENT_INSTRUCTIONS.md ("Emoji / UTF-8 safety") for the workflow.
|
|
docs/index.html text working-tree-encoding=UTF-8 eol=lf
|
|
website/index.html text working-tree-encoding=UTF-8 eol=lf
|
|
docs/support.html text working-tree-encoding=UTF-8 eol=lf
|
|
website/support.html text working-tree-encoding=UTF-8 eol=lf
|
|
CHANGELOG.md text working-tree-encoding=UTF-8 eol=lf
|
|
README.md text working-tree-encoding=UTF-8 eol=lf
|
|
GUIDE.md text working-tree-encoding=UTF-8 eol=lf
|