On Tue, Feb 21, 2023 at 11:10 AM Taylor Blau [off-list ref] wrote:

On Fri, Feb 17, 2023 at 01:12:23PM -0800, Emily Shaffer wrote:

quoted

This turned out pretty long-winded, so a quick summary before I dive in:

- We want to compile parts of Git as independent libraries
- We want to do it by making incremental code quality improvements to Git
- Let's avoid promising stability of the interfaces of those libraries
- We think it'll let Git do cool stuff like unit tests and allowing
  purpose-built plugins
- Hopefully by example we can convince the rest of the project to join
  in the effort

Like others, I am less interested in the VFS-specific components you
mention here, but I suspect that is just one particular instance of
something that would be benefited by making git internals exposed via a
linkable library.

I don't have objections to things like reducing our usage of globals,
making fewer internal functions die() when they encounter an error, and
so on. But like Junio, I suspect that this is definitely an instance of
a "devil's in the details" kind of problem.

That's definitely my main concern: that this turns out to be much more
complicated than imagined and that we leave the codebase in a worse
state without much to show.

Yeah, I'm really hoping we don't end up with ugly half-changes too.
Some examples of "partial credit" that I'd be happy with:

- Fewer internal libraries relying on globals like
the_repository/the_index/etc (we've already started this effort,
libification or no)
- An "ugly" library interface becoming clearer and easier to use (and
internal callers updated)
- Figuring out an "error reporting type" that works well for us

There are some things that *are* ugly, for example, calling a library
via a vtable. But I do feel comfortable waiting to introduce that kind
of thing until we really need it, at which point I suspect we'll have
already made some successful strides with libification in general.

It's not so great to just trust me to say "I promise not to make ugly
changes" - I'd appreciate the community's help pushing back if we
propose doing something in an untidy way without clear justification.

A lesser version of that outcome would be
that we cause a lot of churn in the tree with not much to show either.

I'm actually not so concerned about this! The "churn", as I see it,
comes in the form of code cleanup that already makes Git more
understandable for Git hackers. We do spend some time on that now, as
a project, but I wouldn't be unhappy if we spent even more :)

So I think we'd want to see some more concrete examples with clear
benefits to gauge whether this is a worthwhile direction. I think that
strbuf.h is too trivial an example to demonstrate anything useful. Being
able to extract config.h into its own library so that another non-Git
program could link against it and implement 'git config'-like
functionality would be much more interesting.

Sure - I'm also looking forward to seeing it.

Thanks for your thoughtful reply.
 - Emily