[Spice-devel] [PATCH spice-protocol] build-sys: simplify autogen

[Uri Lublin]

On 12/08/2014 02:53 PM, Marc-André Lureau wrote:
> On Mon, Dec 8, 2014 at 1:47 PM, Uri Lublin <uril at redhat.com> wrote:
>>> With current workflow, you have no guaratee that unreviewed patch go
>>> there by mistake or maliciously. We would need a tool for that.
>>> For me this is the job of maintainer to quickly review each commit
>>> before release.
>>
>> I disagree.
>> When doing a release, a maintainer should _not_ review all commits.
>> Those commits should have been reviewed before being committed.
>> The number of patches added since the previous release can be large,
>> and re-reviewing all of them can be too much work for little gain.
> This contradicts a bit the fact that you can do commit without review.
>   I said "quickly", doing thorough review of all commits before a
> release is not doable. But it is the role of the maintainer to check
> all the commits that go in a repository, because we don't have ways to
> enforce that all patches are the one that are ack on ML. (and I don't
> think we need that, to me the process works well so far)

OK.

I agree that a maintainer should "quick review" the patches.
At least to know what's new in the release and update some files (e.g. 
NEWS).

Hopefully the maintainer would catch "bad" commits, pushed by mistake.
(Let's (naively?) assume no malicious commits are pushed).
That does not happen often (or at least not enough data about 
bad-commits is available).

A maintainer can push a commit without review, but one should not for 
most patches.
And those trivial patches, by definition, do not need to get reviewed by 
the maintainer
doing the release.

Uri