Re: [Patch] Prevent cloning over http from spewing

From: Jeff King <peff@peff.net>
Date: Jun 3, 2009, 19:10 UTC
Message-ID: <20090603191050.GB29564@coredump.intra.peff.net>
In-Reply-To: <7vmy8p8947.fsf@alter.siamese.dyndns.org>

On Wed, Jun 03, 2009 at 11:28:40AM -0700, Junio C Hamano wrote:

Show 9 quoted lines

> You _could_ do something like this patch.  Instead of showing "walk %s"
> and "got %s" lines, with occasional "Getting pack %s\n which contains %s",
> it shows and recycles a single line that shows the number of packs, walk
> actions and got actions.
> 
> This is a toy patch; it hiccups for too long while getting each pack, and
> it does not cleanly restore the display after it finishes, but I'll leave
> it to interested readers as an exercise to properly do this using the
> progress API.

This is much better, and I had roughly the same thought when I saw the original "stop http from spewing" patch. The output is confusing, but we need _some_ progress indicator.

Unfortunately, I don't think it's possible to give a "percent complete" because we are fetching as we walk, meaning we don't know where the end is until we get there (but I might be wrong, as I have never looked too closely at the http fetch code).

So I think the best we could do is probably:

  1. summarize what we have fetched (N packs, N loose objects)
  2. show what we are currently fetching (object or pack)
  3. show the number of bytes retrieved for the current item
  4. if the server provides content-length, show the percentage
     completed for this object
  5. show the current throughput

That should give the user some eye candy to show something is happening, and for a mostly packed repo, should come close to being an overall progress indicator (since most of the time will be spent on one big pack, that will be most of what the user sees, and its progress indicator will dominate the overall time).

I suspect items 3-5 could mostly re-use the existing progress code, too.

-Peff