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

Re: What's cooking in git.git (Jun 2016, #01; Thu, 2)

From
Torsten Bögershausen <tboegi@web.de>
Date
Jun 4, 2016, 14:51 UTC
Message-ID
<1e3801ae-742f-61c4-2b72-350ada0b2220@web.de>
In-Reply-To
<20160604051458.GA3149@glandium.org>
On 2016-06-04 07.14, Mike Hommey wrote:
Show 16 quoted lines
> On Fri, Jun 03, 2016 at 04:47:33PM -0700, Junio C Hamano wrote:
>> Mike Hommey <mh@glandium.org> writes:
>>
>>> In fact, the parser doesn't even reject the one that is considered
>>> invalid (the first).
>>
>> My question was what the desired behaviour is, and if your "fix"
>> gives us that desired outcome.
> 
> From my POV, the desired outcome from this patch series is that there is
> no change of behavior, and Torsten's fix makes
> git://[example.com:123]:/path/to/repo urls handled the same before and
> after the patch series.
> 
> Whether that's the desired behavior is another topic, that we can,
> IMHO, leave for later.
This is the old behavior:
 GIT_TRACE=2  git clone -v git://[github.com:9418]:/tboegi/emacs.d.git $$
16:32:28.692918 git.c:350               trace: built-in: git 'clone' '-v'
'git://[github.com:9418]:/tboegi/emacs.d.git' '95214'
Cloning into '95214'...
Looking up github.com:9418 ... fatal: Unable to look up github.com:9418 (port
9418) (nodename nor servname provided, or not known)
# Everything inside [] goes into the host part (good),
# At the same time :9418 is printed as the port (weird)

This is the new behavior: GIT_TRACE=2 git clone -v git://[github.com:9418]:/tboegi/emacs.d.git $$ Connecting to github.com (port 9418) ... 192.30.252.122 done. or GIT_TRACE=2 git clone -v git://[github.com:9418]/tboegi/emacs.d.git $$ Connecting to github.com (port 9418) ... 192.30.252.122 done.

In any case, everyting inside the [] should go into the host part, for all protocols that have a scheme. But. There is a little but. the ssh protocol did, in early days, not parse ssh://user@[::1]/path/to/repo correctly :-( The only workaround to use a user name with hostname in [] was to use ssh://[user@::1]/path/to/repo

The thing I'm asking is, if we should handle

git://[github.com:9418]/tboegi/emacs.d.git or git://[github.com:9418]:/tboegi/emacs.d.git the same as git://github.com:9418/tboegi/emacs.d.git

Previous: Mike HommeyNext: Mike Hommey
Message 8 of 13 in “What's cooking in git.git (Jun 2016, #01; Thu, 2)”
  1. Junio C HamanoJun 2, 2016
  2. Mike HommeyJun 2, 2016
  3. Torsten BögershausenJun 3, 2016
  4. Junio C HamanoJun 3, 2016
  5. Mike HommeyJun 3, 2016
  6. Junio C HamanoJun 3, 2016
  7. Mike HommeyJun 4, 2016
  8. Torsten BögershausenJun 4, 2016
  9. Mike HommeyJun 4, 2016
  10. Junio C HamanoJun 4, 2016
  11. Torsten BögershausenJun 5, 2016
  12. 1/1 connect: [host:port] is lagacy for sshtboegi@web.de, Jun 5, 2016
  13. Eric SunshineJun 5, 2016

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.