git/list[1] front-page[2] threads[3] people[4] search[5] about
 

Re: Git dumb HTTP protocol should use WebDAV PROPFIND method

From
Milan Hauth <milahu@gmail.com>
Date
Sep 8, 2025, 19:09 UTC
Message-ID
<CAGiEHCuaqu0ia7H8FKAPLcL3Gi_b0gKYe6ugjYAbyVkbd266_w@mail.gmail.com>
In-Reply-To
<xmqqikhswzl7.fsf@gitster.g>
> assume that WebDAV exists on the server side
no

plan A: q: GET $url/info/refs r: HTTP 404

plan B: q: GET $url/.git/info/refs r: HTTP 404

plan C: q: OPTIONS r: Allow: GET,HEAD,PROPFIND,OPTIONS q: PROPFIND $url r: ...

plan D: q: PROPFIND $url/.git/ r: ...

(q = query, r = response)
so instead of throwing
> fatal: repository '$url' not found

git should try plan A, B, C, D, ... to fetch data from the repo

currently, git gives up after plan A
> the WHOLE point of supporting "dumb HTTP" transport
> is to allow a truly dumb HTTP serving infrastructure

there are at least 2 variants of "dumb HTTP servers": HTTP servers without WebDAV HTTP servers with WebDAV

> assume that WebDAV exists on the server side

for HTTP servers without WebDAV see my other thread

Git dumb HTTP protocol should work without update-server-info https://lore.kernel.org/git/CAGiEHCtP29bQRsEyLabNrLuiP96P-o7EEGi88B7pJbP0tfprxw@mail.gmail.com/

Previous: Junio C Hamano
Message 3 of 3 in “Git dumb HTTP protocol should use WebDAV PROPFIND method”
  1. Milan HauthSep 8, 2025
  2. Junio C HamanoSep 8, 2025
  3. Milan HauthSep 8, 2025

Read the whole thread, see it on lore, or plain text.

$ cat FOOTERMessages come from the public archive at lore.kernel.org/git, fetched every hour. The front page is chosen and written each morning by an AI editor and can be wrong; the threads themselves are the record. About and API. For agents: an MCP server at https://gitlist.dev/mcp, and any thread, story or person page as Markdown by adding .md to its URL (or sending Accept: text/markdown). Details in /llms.txt.