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

Re: On shipping more of our technical docs as manpages

From
Jeff King <peff@peff.net>
Date
Sep 27, 2018, 06:40 UTC
Message-ID
<20180927064055.GA2865@sigill.intra.peff.net>
In-Reply-To
<20180927063404.GB220288@aiede.svl.corp.google.com>
On Wed, Sep 26, 2018 at 11:34:04PM -0700, Jonathan Nieder wrote:
Show 18 quoted lines
> > We already have the /** convention I mentioned above. Could we have
> > another micro-format like:
> >
> >   /** API:strbuf - working with strings */
> >
> > in each header file? That would make generating such an index pretty
> > trivial.
> 
> Can you spell out the problem for me a little more?  E.g. if we had a
> convention that the first comment in strbuf.h should say
> 
> 	/* strbuf - Git's standard string type */
> 
> or even just
> 
> 	/* Git's standard string type */
> 
> would that do the trick?

I assumed that not all header files would want to advertise themselves as a public API, so we'd want to some way to differentiate these from first comments in "other" header files (and I assumed we did not want a separate list of "these are the API files to go into the index", at which point you might as well just write "standard string type" in that list).

But I'm happy with any convention that lets Ævar generate the index he wants.

Show 15 quoted lines
> > Interesting. I'm not totally opposed to putting the documentation
> > alongside the actual code. It does feel a bit cluttered to me, but I
> > think you're right that it keeps everything as close together as
> > possible.
> 
> I've experienced projects following both conventions.  One thing I like
> a lot about documentation in the header file is that it encourages
> people to make the API documentation self-contained.  That is, you
> describe the contract in a way that doesn't require reading the
> implementation for details.
> 
> It took me a while to get used to this kind of convention but now I
> really like it.  So I really do prefer to keep putting API
> documentation in the header files in Git.  (Implementation
> documentation in the source files is of course also very welcome.)
Yeah, I'd agree with all of that.
> I use kythe to get an outline view of the header files.
Looks neat. Doesn't seem to be package for Debian, though. ;)
> In Debian we just ship all the docs.  For something like
> technical/pack-heuristics, it's quite nice.  For the API docs, I think
> they belong in the headers.

Yes, I'd agree with that. About half of technical/* is design docs, etc, that might be of use to random folks. But the internal code APIs are really only useful if you have an actual clone of git.git.

-Peff
Previous: Jonathan NiederNext: Junio C Hamano
Message 24 of 51 in “[PATCHv3 0/8] fetch: make sure submodule oids are fetched”
  1. Stefan BellerSep 21, 2018
  2. 1/8 sha1-array: provide oid_array_filterStefan Beller, Sep 21, 2018
  3. Ævar Arnfjörð BjarmasonSep 22, 2018
  4. Stefan BellerSep 25, 2018
  5. Jeff KingSep 26, 2018
  6. Junio C HamanoSep 26, 2018
  7. Ævar Arnfjörð BjarmasonSep 26, 2018
  8. Junio C HamanoSep 26, 2018
  9. Jeff KingSep 26, 2018
  10. Ævar Arnfjörð BjarmasonSep 26, 2018
  11. Jeff KingSep 26, 2018
  12. Stefan BellerSep 26, 2018
  13. Junio C HamanoSep 26, 2018
  14. Stefan BellerSep 26, 2018
  15. Junio C HamanoSep 26, 2018
  16. Jeff KingSep 26, 2018
  17. Stefan BellerSep 26, 2018
  18. On shipping more of our technical docs as manpagesÆvar Arnfjörð Bjarmason, Sep 26, 2018
  19. Junio C HamanoSep 26, 2018
  20. Stefan BellerSep 26, 2018
  21. Ævar Arnfjörð BjarmasonSep 27, 2018
  22. Jeff KingSep 27, 2018
  23. Jonathan NiederSep 27, 2018
  24. Jeff KingSep 27, 2018
  25. Junio C HamanoSep 27, 2018
  26. Jeff KingSep 27, 2018
  27. Documentation/CodingGuidelines: How to document new APIsStefan Beller, Sep 27, 2018
  28. Junio C HamanoSep 27, 2018
  29. Ævar Arnfjörð BjarmasonSep 27, 2018
  30. Jonathan NiederSep 27, 2018
  31. Jeff KingSep 28, 2018
  32. Junio C HamanoSep 28, 2018
  33. strbuf.h: format according to coding guidelinesStefan Beller, Sep 28, 2018
  34. Junio C HamanoSep 28, 2018
  35. Junio C HamanoSep 28, 2018
  36. Junio C HamanoSep 28, 2018
  37. Jeff KingSep 29, 2018
  38. Junio C HamanoSep 28, 2018
  39. Stefan BellerSep 28, 2018
  40. Martin ÅgrenSep 28, 2018
  41. Jeff KingSep 29, 2018
  42. Stefan BellerSep 28, 2018
  43. Jeff KingSep 29, 2018
  44. Jeff KingSep 27, 2018
  45. 2/8 submodule.c: fix indentationStefan Beller, Sep 21, 2018
  46. 3/8 submodule.c: sort changed_submodule_names before searching itStefan Beller, Sep 21, 2018
  47. 4/8 submodule: move global changed_submodule_names into fetch submodule structStefan Beller, Sep 21, 2018
  48. 5/8 submodule.c: do not copy around submodule listStefan Beller, Sep 21, 2018
  49. 6/8 submodule: fetch in submodules git directory instead of in worktreeStefan Beller, Sep 21, 2018
  50. 7/8 fetch: retry fetching submodules if needed objects were not fetchedStefan Beller, Sep 21, 2018
  51. 8/8 builtin/fetch: check for submodule updates for non branch fetchesStefan Beller, Sep 21, 2018

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.