Files
multica/server
Bohan Jiang 8db619c1cd fix(email): wire SMTP_EHLO_NAME through self-host config + docs [MUL-2984] (#3749)
* fix(email): wire SMTP_EHLO_NAME through self-host config + docs

Follow-up to #3679, which added SMTP_EHLO_NAME in code but never exposed
it to operators.

- docker-compose.selfhost.yml: pass SMTP_EHLO_NAME through to the backend
  container. The compose env block is an explicit allowlist, so without
  this the override set in .env was silently dropped and never reached
  the process — making the escape hatch unusable on the docker path.
- Document the var alongside its SMTP_* siblings: .env.example,
  SELF_HOSTING_ADVANCED.md, environment-variables.mdx, auth-setup.mdx,
  and self-host-quickstart.mdx (the last two with a strict-relay example).
- email.go: log when os.Hostname() fails instead of silently falling back
  to net/smtp's lazy "localhost" — the exact greeting strict relays reject.
- Add TestNewEmailService_EHLOName covering the env override, trimming,
  and the hostname fallback.

MUL-2984

Co-authored-by: multica-agent <github@multica.ai>

* fix(email): gate EHLO resolution to SMTP mode + sync docs to zh/ja/ko

Addresses review nits on this PR:

- email.go: resolve smtpEHLOName only when SMTP_HOST is set, so the
  Resend / DEV-stdout paths never call os.Hostname() or emit its
  failure log. The EHLO name is only ever used on the SMTP send path.
- docs: add SMTP_EHLO_NAME to the zh/ja/ko variants of
  environment-variables, self-host-quickstart, and auth-setup, in sync
  with the English docs updated earlier in this PR.

Note: the ja/ko self-host-quickstart and auth-setup pages were already
missing the port-465 implicit-TLS example (pre-existing i18n drift from
an earlier SMTP_TLS change, unrelated to this PR); the new EHLO block is
inserted at the correct logical anchor regardless. A full ja/ko re-sync
is left as a separate follow-up.

MUL-2984

Co-authored-by: multica-agent <github@multica.ai>

---------

Co-authored-by: J <j@multica.ai>
Co-authored-by: multica-agent <github@multica.ai>
2026-06-04 14:44:55 +08:00
..
2026-06-03 14:32:44 +08:00