threads / discuss / 59373

A nice, beauty progress metter for Git Clone + a feature request

Subject: A nice, beauty progress metter for Git Clone + a feature request

## tl;dr

5 messages between Mar 10, 2023 and Mar 28, 2023.

replies: 4people: 2as markdown or json

Sebastian Gniazdowski· Mar 10, 2023, 11:37 UTC · lore
Check out: https://asciinema.org/a/566216

Feature request is: add a pipe={cmd} option to git, that would just pass through the git clone --progress output into {cmd}. This is EXTREMELY EASY TO DO SO, ONLY

FILE *pipe=popen(cmd_option_str, "w") ;
is needed, and one can then write to `pipe` file handler.
-- 
Best regards,
Sebastian Gniazdowski
Jeff King· Mar 15, 2023, 18:02 UTC · re: Sebastian Gniazdowski · lore

Re: A nice, beauty progress metter for Git Clone + a feature request

On Fri, Mar 10, 2023 at 11:37:49AM +0000, Sebastian Gniazdowski wrote:
Show 9 quoted lines
> Check out: https://asciinema.org/a/566216
> 
> Feature request is: add a pipe={cmd} option to git, that would just
> pass through the git clone --progress output into {cmd}. This is
> EXTREMELY EASY TO DO SO, ONLY
> 
> FILE *pipe=popen(cmd_option_str, "w") ;
> 
> is needed, and one can then write to `pipe` file handler.

I think it would be neat to have options for prettier progress output. Unfortunately, it's not quite that easy.

Imagine we had a config option, progress.command or something, which specified a shell command to which we'd pipe machine-readable progress, and it would turn it into something pretty. There are two gotchas I can think of immediately:

  1. Progress is often created by multiple processes. For example, in a
     clone, you might see progress output from index-pack as it receives
     the pack and resolves the delta. But then you might also see
     progress during the checkout phase, as we write out the worktree
     files.
     One solution is to say that each progress command is independent,
     and we'd simply run it twice, once for each context. That's easy to
     do, but restricts some fancier options that a command could do.
     Another is for the parent clone command to kick off a progress
     process, and then tell child processes (perhaps through the
     environment) which descriptor they can use to talk to the progress
     command.
  2. Sometimes progress is shoved through a single descriptor along with
     other human-readable messages. I'm thinking specifically of the
     output you get from a clone/fetch _before_ we start receiving pack
     bytes from the server. There the server is giving progress on
     preparing the fetch. It comes over a sideband channel through Git's
     protocol, along with any regular stderr messages, and the client
     side just dumps it all to the local stderr.
     Obviously the remote side is not going to run our custom progress
     command. But it would be possible to add a protocol extension that
     says "hey, please send machine-readable progress data over sideband
     4" or something.

Now, none of that is _strictly_ necessary. We could just dump the same human-readable progress to the progress command that we dump to stderr now, and it could scrape it for things that look like progress, and pass everything else through. But then, you can already do that with:

  git clone --progress ... 2>&1 | my-progress-scraper

(and it looks like the asciinema you showed is probably just a syntactically nicer version of that with support from the shell?).

-Peff
Sebastian Gniazdowski· Mar 27, 2023, 10:59 UTC · re: Jeff King · lore

Re: A nice, beauty progress metter for Git Clone + a feature request

On Wed, 15 Mar 2023 at 18:02, Jeff King <peff@peff.net> wrote:
Show 11 quoted lines
> Now, none of that is _strictly_ necessary. We could just dump the same
> human-readable progress to the progress command that we dump to stderr
> now, and it could scrape it for things that look like progress, and pass
> everything else through. But then, you can already do that with:
>
>   git clone --progress ... 2>&1 | my-progress-scraper
>
> (and it looks like the asciinema you showed is probably just a
> syntactically nicer version of that with support from the shell?).
>
> -Peff

Yes, that's what the asciinema does – it pipes git clone --progress to a script, and that's pretty much it. That's why I thought about a core.pipe option, that would hold the my-progress-scrapper command. That's a very easy change, could it be added to git source code? The scrapper from the asciinema doesn't care if it's count, retrieve or resolve, etc. currently ongoing, it simply displays a gauge with the current percentage for the current stage. After a while one discovers that it's resolve that's the final stage, and the gauge is intuitive either before or after this revelation.

-- 
Best regards,
Sebastian Gniazdowski
Jeff King· Mar 28, 2023, 17:51 UTC · re: Sebastian Gniazdowski · lore

Re: A nice, beauty progress metter for Git Clone + a feature request

On Mon, Mar 27, 2023 at 10:59:56AM +0000, Sebastian Gniazdowski wrote:
Show 15 quoted lines
> On Wed, 15 Mar 2023 at 18:02, Jeff King <peff@peff.net> wrote:
> > Now, none of that is _strictly_ necessary. We could just dump the same
> > human-readable progress to the progress command that we dump to stderr
> > now, and it could scrape it for things that look like progress, and pass
> > everything else through. But then, you can already do that with:
> >
> >   git clone --progress ... 2>&1 | my-progress-scraper
> >
> > (and it looks like the asciinema you showed is probably just a
> > syntactically nicer version of that with support from the shell?).
> 
> Yes, that's what the asciinema does  – it pipes git clone --progress
> to a script, and that's pretty much it. That's why I thought about a
> core.pipe option, that would hold the my-progress-scrapper command.
> That's a very easy change, could it be added to git source code?

For the reasons I gave earlier, it has to scrape all of stderr, and not just progress output. I'm still skeptical that this buys a lot over just redirecting stderr entirely via the shell, but it's at least conceptually simple (it's basically the same as a pager for stderr).

The implementation might have some tricky bits to think about, though, since we have an extra process to handle. For real pagers, we take some care to terminate them on close. Maybe we wouldn't need the same here, though (since the filter isn't interactive, and we don't want Git to wait on exiting until it's done).

So I dunno. It might only be a few lines. It just feels like a weird feature to add.

-Peff
Sebastian Gniazdowski· Mar 28, 2023, 20:49 UTC · re: Jeff King · lore

Re: A nice, beauty progress metter for Git Clone + a feature request

On Tue, 28 Mar 2023 at 17:51, Jeff King <peff@peff.net> wrote:
> So I dunno. It might only be a few lines. It just feels like a weird
> feature to add.

I would perceive the feature as a nice, versatile trick instead of weird… Current git clone progress is quite unexpected – when I was starting to use Git I've had a sensation of disappointment each time I've run clone – was expecting some nice progress bar and was served with bunch of text lines with numbers that weren't any percentages… I'm not saying that the progress in the asciinema is some uber cool solution, but then, it may serve as an example to draw attention to beautifying the progress and the freedom of hooking any command via a core.progress_pipe might result in some other nice scripts/apps to bloom out there.

-- 
Best regards,
Sebastian Gniazdowski

← back to recent threads