| commit | cf5f603718484931f8b1c59208066052bb203c5e | |
| author | Shawn O. Pearce <spearce@spearce.org> | |
| Tue, 23 Dec 2008 18:03:43 +0000 (23 10:03 -0800) | ||
| committer | Robin Rosenberg <robin.rosenberg@dewire.com> | |
| Sun, 4 Jan 2009 00:01:51 +0000 (4 01:01 +0100) | ||
| tree | 4d51adc413a92e1a5ec0a4c1444861bb7ab0657c | treesnapshot (tar.gz zip) |
| parent | ab27f138a4db9a9cd66a01f722d3755aa7106b4e | commitdiff |
Fix "fetch pulled too many objects" when auto-following tags
If we don't take into consideration the objects obtained during
the first connection when we open a second to auto-follow tags
we will download a large chunk of the repository a second time.
This is very wasteful of network bandwidth, and is an abuse of
the server.
Because we delay all ref updates until the very end of the fetch
process we need to hold onto the set of objects we requested in
the first connection, and pass that set into the subsequent one
so it can be considered reachable.
Issue: http://code.google.com/p/egit/issues/detail?id=22
Signed-off-by: Shawn O. Pearce <spearce@spearce.org>
Signed-off-by: Robin Rosenberg <robin.rosenberg@dewire.com>
If we don't take into consideration the objects obtained during
the first connection when we open a second to auto-follow tags
we will download a large chunk of the repository a second time.
This is very wasteful of network bandwidth, and is an abuse of
the server.
Because we delay all ref updates until the very end of the fetch
process we need to hold onto the set of objects we requested in
the first connection, and pass that set into the subsequent one
so it can be considered reachable.
Issue: http://code.google.com/p/egit/issues/detail?id=22
Signed-off-by: Shawn O. Pearce <spearce@spearce.org>
Signed-off-by: Robin Rosenberg <robin.rosenberg@dewire.com>