Re: Bibisecting: Over 1000 bibisects served!

On 10/19/2015 05:10 PM, Norbert Thiebaud wrote:
> On Mon, Oct 19, 2015 at 10:02 AM, Stephan Bergmann <sbergman@redhat.com> wrote:
> > On 10/19/2015 04:53 PM, Norbert Thiebaud wrote:
> > > On Mon, Oct 19, 2015 at 3:23 AM, Stephan Bergmann <sbergman@redhat.com>
> > > wrote:
> > > > A reminder when updating a bug with information about a successful
> > > > bibisecting:  Do not only paste the commit message of the first bad
> > > > bibisect
> > > > repo commit, but:  "please [also] provide either the commit message of
> > > > the
> > > > (last good) commit in the bibisect repo directly preceding this [first
> > > > bad]
> > > > commit, or at least the URL of the bibisect repo (so I can determine that
> > > > preceding commit myself).  (Individual commits in bibisect repos often
> > > > correspond to a range of source commits, so it is always important to
> > > > know
> > > > that full range.)"
> > > > (<https://bugs.documentfoundation.org/show_bug.cgi?id=91895#c5>)
> > >
> > >
> > > on Mac and Windows bibisect, the bibisect commit list all the source
> > > commit sha since the previous bibisect commit,
> > > hence solving that problem.
> >
> >
> > No, that does not solve the problem if all you know is the commit message of
> > a bibisect commit mentioning a single source commit.  It could be a bibisect
> > commit from such a Mac or Windows bibisect repo that corresponds to a single
> > source commit, or it could be a bibsect commit from another bibisect repo
> > that may corresponds to a larger range of source commits.
>
> maybe an example will be clearer:
>
> commit c49af8fc6406f9e7e8e0b1dcebee6df87bdeb9aa
> Author: Norbert Thiebaud <nthiebaud@gmail.com>
> Date:   Thu May 21 01:05:45 2015 -0500
>
>      source sha:6586da0631ddcfd704538b0e1cf96d2ea0be7cd9
>
>      source sha:6586da0631ddcfd704538b0e1cf96d2ea0be7cd9
>      source sha:55828bef6926a0eec446ca0b8503f656cc5c4a7f
>      source sha:9a721f1fe6c30ec231744a730f9046aa64bfae0f
>      source sha:5591fdada7cf4f81f173f7954e2b4831c6dc89c9
>      source sha:b253d2bb93d170e1ba6e10fb9d74388b7c25e588
>      source sha:e32da3af7695617ca1fc4c8a5c1689952269ade5
>      source sha:b03f2cb6417c1f7a6892db23bcd3e807b2806e89
>      source sha:05c85d7a36f3ea69bba0fa69720bff61f2e0f573
>      source sha:8baaf76ae373a7c945e1cc2fa5cd57b59749d414
>      source sha:f387b04dc89d2bab96d77a6f1445d272e2b1d827
>      source sha:3dd259ffa1607fa91d598954b07f9ac8e5c0b08a
>      source sha:65f9d502d88e832e8b8af71b7be8f706519f7162
>      source sha:e964042c16efafd5f6656dbb7cea14c852a39822
>      source sha:bc56da8784a76e92878193baf4764c50333eb21f
>      source sha:4b57be7234ea61aad0a472a56a4f073270933c8e
>      source sha:23b439a664549e1b43347bc625c267fa3c4f4d3b
>      source sha:49e13aa41895e8562d83cff189ab6a78d5d902aa
>      source sha:456c379ffd8683cd8c22969268900787cf1ac28e
>      source sha:3cbdf64ad5240e6d9a73d4f7e005f7110d5e4002
>      source sha:662700703bebad38ca7ad74ca4eb040fe8b5b676
>      source sha:dcefc97c8bd5be9ba229098c32d2a5c73d084163
>      source sha:953f327818f565969b8de5d9b956bd6b9a7c64b2
>      source sha:25a47c5cd54c73e754de988bde8ec8a202d27717
>      source sha:7a0af37989d1f1b508a61f28e785c5b1f27d58af
>
>
> If a bibisect terminate on the bibisect commit above, then the listed
> source sha: are all the source commit that are in _this_ bibisect
> commit that are not in the bibisect commit just before that.

But that example is irrelevant to what I wrote above, as the commit 
message mentions a range of more than one source commits.