incubator-general mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Sam Corbett <>
Subject Re: [QUESTION] Rewriting history
Date Fri, 20 Nov 2015 16:50:57 GMT

To give some context on the objects in question..

There are 14 files in Brooklyn's history larger than 1Mb. Four of these are
larger than 20Mb. The largest is 57Mb(!). These files make for a
significant penalty when cloning the repository.

Of the fourteen large files, five were present in the Brooklyn repo when
the project joined the incubator (1 May 2015). All of them were deleted
before its first incubator release (0.7.0-M2 (incubating)) on 23 December
2014. The files are also not to be found in Brooklyn's last pre-ASF release
(they are all either examples or in a 'sandbox' module).

The other nine files were all deleted before Brooklyn even had its first
public GA release (0.4.0) on January 16 2013. The worst offending file is
part of this group.



On 20 November 2015 at 16:42, Ted Dunning <> wrote:

> On Sat, Nov 21, 2015 at 12:02 AM, Hadrian Zbarcea <>
> wrote:
> > The question was asked on general@ [1] many moons ago, but received no
> > answer (the original thread [1] indicates the offenders). We'd like to
> use
> > this occasion to split the git repo into multiple smaller once (more
> > naturally tailored to the project), so one idea would be, I think, to
> leave
> > incubator-brooklyn.git as is and just extract what we need (everything
> > except the large binaries) in new repos. Note: afaiui, those artifacts
> > we're never part of an ASF release.
> >
> > I am not sure what would be acceptable per ASF (intensely debated)
> > policies.
> >
> Surely copying and leaving some stuff behind is a fine thing to do.

  • Unnamed multipart/alternative (inline, None, 0 bytes)
View raw message