Re: [PATCH v3 00/32] .mailmap: Add mailmap file to have a cleaner git history

["Philippe Mathieu-Daudé" <philmd@redhat.com>]

On 1/6/20 11:41 AM, Laszlo Ersek wrote:
> On 01/03/20 14:56, Philippe Mathieu-Daudé wrote:
>> On 1/3/20 2:48 PM, Laszlo Ersek wrote:
> 
>>> Apparently, there is a patch that
>>> prevents the series from being merged? Which one?
>>
>> This comment was about patch 13/32, but Jiewen Yao already reviewed
>> it,
> 
> OK, thanks.
> 
>> so this series is fully reviewed by each developer.
> 
> I've now tried applying this series locally (for a github.com pull
> request). But git-am fails, on patch#6 ("corrupted patch").
> 
> I've checked the patch email (as saved from Thunderbird) and I can see
> it uses quoted-printable transfer encoding. That has basically never
> worked with edk2. Can you please send a v4, with 8-bit encoding?
> 
> Please see:
> 
>    [edk2-devel] [PATCH]
>    BaseTools/Scripts: Add sendemail.transferEncoding to SetupGit.py
> 
>    https://edk2.groups.io/g/devel/message/52323
>    http://mid.mail-archive.com/02A34F284D1DA44BB705E61F7180EF0AB5C64C6F@ORSMSX114.amr.corp.intel.com

Ah yes, I enabled this setting globally, but not in my edk2 identity 
(which is the one I use to avoid utf8 in my last name, which is the 
single community with UTF-8 issues I've to deal with).

    --identity=<identity>

      A configuration identity. When given, causes values
      in the sendemail.<identity> subsection to take
      precedence over values in the sendemail section. The
      default identity is the value of sendemail.identity.

I now fixed my config with:

$ git config --global sendemail.edk2.transferEncoding 8-bit

(btw I'm unable to apply the previously referred patch).

>>> Can you please monitor the remaining feedback for this series? If
>>> everyone is happy with the v3 updates, and you determine that the
>>> patch set is fully reviewed, I can help pick up the new tags, and
>>> merge the series.
>>
>> Since the change is a one-line, I'd prefer avoid respining the whole
>> series.
> 
> I think you mean the change is a one-liner *for a commit message*
> (namely, Jiewen's R-b).
> 
> I would normally agree (I'd just pick up that R-b from the list), but
> again, I can't apply the v3 patches in the first place.

Fair enough.

>> IMHO the series is ready for step #4 "stewards go through the backlog,
>> and decide one by one whether, for person X, the proposed mappings are
>> cross-domain or not, and ACK or NACK" which I don't really get.
>> I suppose this will be resolved in the open on the mailing list, so
>> individuals will get notified their patch got refused.
> 
> Let me quote the original suggestion in-line, from here:
> 
>    https://edk2.groups.io/g/devel/message/51906
>    http://mid.mail-archive.com/9e212952-a114-57d0-9880-6686edc4d8ae@redhat.com
> 
> On 12/06/19 13:23, Laszlo Ersek wrote:
>> - Collect feedback (from individual reviewers / contributors) for one
>> or two weeks
>>
>> - Accept cross-domain mappings in any patch where the person subject
>> to the patch is ACKing the patch from the mapped-to (i.e. "current")
>> email address. This is now a *lot* easier to do, because the patch
>> series is structured by person.
>>
>> - After one or two weeks, push all the patches that have been approved
>> like explained above.
> 
> This is where we are now. We're about to push the patches with the
> direct contributor approvals.
> 
> This covers 32 patches (v3) out of your original 105 (v2).
> 
>> - Repost the rest of the series, CC'd solely to the stewards. We will
>> then have to go through the backlog, and decide one by one whether,
>> for person X, the proposed mappings are cross-domain or not, and ACK
>> or NACK accordingly.
> 
> This is "step 4", and for that, you should please post the *rest* of the
> patches: approximately 105-32 = 73 patches.
> 
> That is the "backlog". The idea is that the stewards should review
> those, because we should be permitted to add email mappings even for
> such contributors that do not personally comment on the proposed
> mappings, as long as those mappings are not cross-domain.

OK, understood now :)

Another way can be to not fix broken/obsolete emails, but fix author names.

>>
>> - Push the second round, drop the rest for good.
> 
> And here the "rest" (to be dropped for good) stands for such patches
> that even the stewards don't approve (because the mappings appear
> cross-domain, which cannot be approved without direct contributor
> feedback).
> 
> Thanks!
> Laszlo
>