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

Re: [PATCH] imap-send: replace auto-probe libcurl with hard dependency

From
Ævar Arnfjörð Bjarmason <avarab@gmail.com>
Date
Feb 1, 2023, 23:56 UTC
Message-ID
<230202.86zg9xormj.gmgdl@evledraar.gmail.com>
In-Reply-To
<xmqqlelhx973.fsf@gitster.g>
On Wed, Feb 01 2023, Junio C Hamano wrote:
Show 9 quoted lines
> Ævar Arnfjörð Bjarmason  <avarab@gmail.com> writes:
>
>> Change the "imap-send" command to have a hard dependency on libcurl,
>> before this it had an optional dependency on both libcurl and OpenSSL,
>> now only the OpenSSL dependency is optional.
>>
>> This simplifies our dependency matrix my getting rid of yet another
>
> "my" -> "by", I think.
Thanks, I'll fix that in a re-roll, pending the below...
Show 23 quoted lines
>> special-case. Given the prevalence of libcurl and portability of
>> libcurl it seems reasonable to say that "git imap-send" cannot be used
>> without libcurl, almost everyone building git needs to be able to push
>> or pull over http(s), so they'll be building with libcurl already.
>
> OK.
>
>> So let's remove the previous "USE_CURL_FOR_IMAP_SEND" knob. Whether we
>> build git-imap-send or not is now controlled by the "NO_CURL"
>> knob.
>
> OK.
>
>> Let's also hide the old --curl and --no-curl options, and die if
>> "--no-curl" is provided.
>
> In other words, if we are building imap-send, we sure know cURL is
> there, and there is no need to tell a running imap-send not to use
> cURL to talk to the IMAP service?  I am not sure the linkage of this
> change with the rest of the patch.  Isn't that a totally orthogonal
> issue?  Your imap-send might be cURL enabled, but unless we stop to
> ship with our own IMAP routines compiled into imap-send, --no-curl
> does have a purpose.

The equivalent of USE_CURL_FOR_IMAP_SEND is now always true, and that's what "--curl" would enable.

The "--no-curl" option would then have us use the OpenSSL codepath, but that'll no longer be supported, we'll always use curl.

The link to the rest of the patch is then that "USE_CURL_FOR_IMAP_SEND" and "curl_check" etc. was needed to check if we had curl with imap-send, now we declare that we'll always need it.

And the link to the thread-at-large is that Jiang Xin's upthread version moves those checks from the Makefile into the code itself, I agree that wolud be an improvement, but if we're happy to just make it a hard dependency we won't need it there either...

> Or did you just forget to document that we stop to ship with our own
> IMAP routines in the above?  If so, as long as it is made a bit more
> prominent in the proposed log message in a reroll, I would be happy
> with such a change rolled into the same patch.

I'm not sure what you mean here, we still ship with the same routines, we just always take the "curl" codepath for the non-tunnel codepath now.

Is this perhaps confusion because while we do make curl mandatory, we're not dropping the OpenSSL code? That's because we're dropping its use for the non-tunnel codepath, but unfortunately for the tunnel case we'll still need it.

So we only make curl mandatory, but OpenSSL remains optional.
Previous: Ævar Arnfjörð BjarmasonNext: Junio C Hamano
Message 7 of 37 in “Makefile: not use mismatched curl_config to check version”
  1. 1/2 Makefile: not use mismatched curl_config to check versionJiang Xin, Feb 1, 2023
  2. 2/2 imap-send: not define USE_CURL_FOR_IMAP_SEND in MakefileJiang Xin, Feb 1, 2023
  3. imap-send: replace auto-probe libcurl with hard dependencyÆvar Arnfjörð Bjarmason, Feb 1, 2023
  4. Junio C HamanoFeb 1, 2023
  5. Jeff KingFeb 1, 2023
  6. Ævar Arnfjörð BjarmasonFeb 2, 2023
  7. Ævar Arnfjörð BjarmasonFeb 1, 2023
  8. Junio C HamanoFeb 2, 2023
  9. 0/6 imap-send: replace auto-probe libcurl with hard dependencyÆvar Arnfjörð Bjarmason, Feb 2, 2023
  10. 1/6 imap-send: note "auth_method", not "host" on auth method failureÆvar Arnfjörð Bjarmason, Feb 2, 2023
  11. Junio C HamanoFeb 2, 2023
  12. 2/6 imap-send doc: the imap.sslVerify is used with imap.tunnelÆvar Arnfjörð Bjarmason, Feb 2, 2023
  13. 3/6 imap-send: replace auto-probe libcurl with hard dependencyÆvar Arnfjörð Bjarmason, Feb 2, 2023
  14. Junio C HamanoFeb 2, 2023
  15. 4/6 imap-send: make --curl no-optionalÆvar Arnfjörð Bjarmason, Feb 2, 2023
  16. Junio C HamanoFeb 2, 2023
  17. Ævar Arnfjörð BjarmasonFeb 3, 2023
  18. Junio C HamanoFeb 4, 2023
  19. 6/6 imap-send: correctly report "host" when using "tunnel"Ævar Arnfjörð Bjarmason, Feb 2, 2023
  20. Junio C HamanoFeb 2, 2023
  21. Jeff KingFeb 3, 2023
  22. Ævar Arnfjörð BjarmasonFeb 3, 2023
  23. Jeff KingFeb 4, 2023
  24. Ævar Arnfjörð BjarmasonFeb 5, 2023
  25. Jeff KingFeb 7, 2023
  26. Ævar Arnfjörð BjarmasonFeb 7, 2023
  27. Junio C HamanoFeb 7, 2023
  28. Ævar Arnfjörð BjarmasonFeb 7, 2023
  29. Jeff KingFeb 7, 2023
  30. Jeff KingFeb 7, 2023
  31. Junio C HamanoFeb 7, 2023
  32. Ævar Arnfjörð BjarmasonFeb 8, 2023
  33. Jeff KingFeb 17, 2023
  34. Junio C HamanoFeb 6, 2023
  35. 5/6 imap-send: remove old --no-curl codepathÆvar Arnfjörð Bjarmason, Feb 2, 2023
  36. Junio C HamanoFeb 2, 2023
  37. Junio C HamanoFeb 1, 2023

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.