[PATCH v2 0/1] update MAINTAINERS file with git trailers

[PATCH v2 0/1] update MAINTAINERS file with git trailers

[Bruno Larsen]

Some private chats highlighted that the currently in use git trailers
aren't explained well enough. This patch aims to fix that by adding the
information in a more verbose way to guarantee that everyone is on the
same page about them and refine anything that might still need work.

The first version of this update didn't get much traction, so I decided
to add every active global maintainer as CC, since everyone is affected
by this and I want some consensus before comitting.

Right now there is one big unanswered question: Should we have a
specific tag to explicitly signal when a patch has been partially
approved? Eli asked for it to avoid people mechanically reading tags
from thinking that a patch has been fully approved when it was only
partial. If we agree that there should be one, what would it look like?
I suggested using Acked-By, but Simon uses it differently, and I opted
to keep his usage of it.

There is a secondary question for me, which is to add a hard requirement
for approval tags. If we add it, the sourceware repository could have a
pre-receive hook (or update hook) that automatically reject pushes if
they dont have an approval tag. This would be especially useful to
avoid accidentally pushing the wrong commit, but is more annoying for
global maintainers (need to approve their own patches) and we'd need to
see what to do about the obvious fix rule. My first answer for that is
adding a special "Approved-By: Obvious fix" or similar.

Thoughts? Comments? Suggestions?

Version 1: https://inbox.sourceware.org/gdb-patches/20230516143826.3431583-1-blarsen@redhat.com/

Bruno Larsen (1):
  [gdb]: add git trailer information on gdb/MAINTAINERS

 gdb/MAINTAINERS | 56 ++++++++++++++++++++++++++++++++++++++++++-------
 1 file changed, 48 insertions(+), 8 deletions(-)

-- 
2.40.1

[PATCH v2 1/1] [gdb]: add git trailer information on gdb/MAINTAINERS

[Bruno Larsen]

The project has been using Tested-By (tb), Reviewed-By (rb) and
Approved-By (ab) for some time, but there has been no information to be
found in the actual repository. This commit changes that by adding
information about all git trailers to the MAINTAINERS file, so that it
can be easily double-checked.

The upstream discussion also brought up the use of Acked-by, which is
better defined in this commit.
---
 gdb/MAINTAINERS | 56 ++++++++++++++++++++++++++++++++++++++++++-------
 1 file changed, 48 insertions(+), 8 deletions(-)

diff --git a/gdb/MAINTAINERS b/gdb/MAINTAINERS
index 175595d5f17..22f7ecbd471 100644
--- a/gdb/MAINTAINERS
+++ b/gdb/MAINTAINERS
@@ -43,14 +43,9 @@ patch without review from another maintainer.  This especially includes
 patches which change internal interfaces (e.g. global functions, data
 structures) or external interfaces (e.g. user, remote, MI, et cetera).
 
-The term "review" is used in this file to describe several kinds of feedback
-from a maintainer: approval, rejection, and requests for changes or
-clarification with the intention of approving a revised version.  Review is
-a privilege and/or responsibility of various positions among the GDB
-Maintainers.  Of course, anyone - whether they hold a position but not the
-relevant one for a particular patch, or are just following along on the
-mailing lists for fun, or anything in between - may suggest changes or
-ask questions about a patch!
+The word "contributor" is used in this document to refer to any GDB
+developer listed above as well as folks who may have suggested some patches
+but aren't part of one of those categories for any reason.
 
 There's also a couple of other people who play special roles in the GDB
 community, separately from the patch process:
@@ -78,6 +73,51 @@ consensus among the global maintainers and any other involved parties.
 In cases where consensus can not be reached, the global maintainers may
 ask the official FSF-appointed GDB maintainers for a final decision.
 
+The term "review" is used in this file to describe several kinds of feedback
+from a maintainer: approval, rejection, and requests for changes or
+clarification with the intention of approving a revised version.  Approval is
+a privilege and/or responsibility of various positions among the GDB
+Maintainers.  Of course, anyone - whether they hold a position but not the
+relevant one for a particular patch, or are just following along on the
+mailing lists for fun, or anything in between - may suggest changes,
+ask questions about a patch or say if they believe a patch is fit for
+upstreaming!
+
+To ensure that patches are only pushed when approved, and to thank the
+contributors who take the time to review incoming patches, the GDB project
+uses trailers to identify who contributed and how.  All patches pushed
+upstream should have at least one Approved-By patches (with the exception of
+obvious patches, see below).  The trailers (or tags) currently in use are:
+
+ - Acked-By:
+
+   Used when a contributor has taken a quick glance at a patch and agrees
+   with the direction outlined in the commit message, but hasn't evaluated
+   the code for correctness or regressions.
+
+ - Tested-by:
+
+   Used when a contributor does not want to comment on the quality
+   of the code in the patch, but has tested and sees no regressions on their
+   hardware.
+
+ - Reviewed-by:
+
+   Used when a contributors have looked at code and agrees with the changes,
+   but either don't have the authority or don't feel comfortable approving
+   the patch (usually due to unfamiliarity with the relevant part of the code).
+
+ - Approved-by:
+
+   Used by responsible maintainers or global maintainers when
+   a patch is ready to be upstreamed.  Some patches may touch multiple areas
+   and require multiple approvals before landing (such as a maintainer only
+   approving documentation), it is up to the maintainer giving the approval tag
+   to make it clear when that a tag is not sufficient.
+   Responsible, Global and Official FSF-appointed maintainers may approve their
+   own patches, but it is recommended that they seek external approval before
+   doing so.
+
 
 			The Obvious Fix Rule
 			--------------------
-- 
2.40.1

Re: [PATCH v2 1/1] [gdb]: add git trailer information on gdb/MAINTAINERS

[Kevin Buettner]

Hi,

Thanks for taking this on.  See my thoughts inline below...

On Mon, 12 Jun 2023 12:01:40 +0200
Bruno Larsen <blarsen@redhat.com> wrote:

> The project has been using Tested-By (tb), Reviewed-By (rb) and
> Approved-By (ab) for some time, but there has been no information to be
> found in the actual repository. This commit changes that by adding
> information about all git trailers to the MAINTAINERS file, so that it

If you wish to document *all* git trailers, then 'Bug:' and
'Co-Authored-By:' should be documented too.

> can be easily double-checked.
> 
> The upstream discussion also brought up the use of Acked-by, which is
> better defined in this commit.
> ---
>  gdb/MAINTAINERS | 56 ++++++++++++++++++++++++++++++++++++++++++-------
>  1 file changed, 48 insertions(+), 8 deletions(-)
> 
> diff --git a/gdb/MAINTAINERS b/gdb/MAINTAINERS
> index 175595d5f17..22f7ecbd471 100644
> --- a/gdb/MAINTAINERS
> +++ b/gdb/MAINTAINERS
> @@ -43,14 +43,9 @@ patch without review from another maintainer.  This especially includes
>  patches which change internal interfaces (e.g. global functions, data
>  structures) or external interfaces (e.g. user, remote, MI, et cetera).
>  
> -The term "review" is used in this file to describe several kinds of feedback
> -from a maintainer: approval, rejection, and requests for changes or
> -clarification with the intention of approving a revised version.  Review is
> -a privilege and/or responsibility of various positions among the GDB
> -Maintainers.  Of course, anyone - whether they hold a position but not the
> -relevant one for a particular patch, or are just following along on the
> -mailing lists for fun, or anything in between - may suggest changes or
> -ask questions about a patch!
> +The word "contributor" is used in this document to refer to any GDB
> +developer listed above as well as folks who may have suggested some patches
> +but aren't part of one of those categories for any reason.
>  
>  There's also a couple of other people who play special roles in the GDB
>  community, separately from the patch process:
> @@ -78,6 +73,51 @@ consensus among the global maintainers and any other involved parties.
>  In cases where consensus can not be reached, the global maintainers may
>  ask the official FSF-appointed GDB maintainers for a final decision.
>  
> +The term "review" is used in this file to describe several kinds of feedback
> +from a maintainer: approval, rejection, and requests for changes or
> +clarification with the intention of approving a revised version.  Approval is

In the line above, you've replaced the word "Review" with "Approval". 
I think I like "Review" better.  (But there may be an even better word
or phrase to use here.)

> +a privilege and/or responsibility of various positions among the GDB
> +Maintainers.  Of course, anyone - whether they hold a position but not the
> +relevant one for a particular patch, or are just following along on the
> +mailing lists for fun, or anything in between - may suggest changes,
> +ask questions about a patch or say if they believe a patch is fit for
> +upstreaming!
> +
> +To ensure that patches are only pushed when approved, and to thank the
> +contributors who take the time to review incoming patches, the GDB project
> +uses trailers to identify who contributed and how.  All patches pushed
> +upstream should have at least one Approved-By patches (with the exception of
> +obvious patches, see below).  The trailers (or tags) currently in use are:
> +
> + - Acked-By:
> +
> +   Used when a contributor has taken a quick glance at a patch and agrees
> +   with the direction outlined in the commit message, but hasn't evaluated
> +   the code for correctness or regressions.
> +
> + - Tested-by:
> +
> +   Used when a contributor does not want to comment on the quality
> +   of the code in the patch, but has tested and sees no regressions on their
> +   hardware.

For "Tested-by", I suggest this instead:

    Used when a contributor has tested the patch and finds that it
    fixes the claimed problem.  It may also be used to indicate that
    the contributor has performed regression testing.  By itself, this
    tag says nothing about the quality of the fix implemented by the
    patch.

> +
> + - Reviewed-by:
> +
> +   Used when a contributors have looked at code and agrees with the changes,

s/contributors have/contributor has/
s/code/the code/

> +   but either don't have the authority or don't feel comfortable approving

s/don't/doesn't/ ??

> +   the patch (usually due to unfamiliarity with the relevant part of the code).

I think I'd remove the parenthetical remark.  There could be a variety of
reasons for feeling uncomfortable with outright approving the change. 
One of the reasons that I sometimes don't approve a patch with which I
agree is that I want to give other maintainers a chance to look at it
first, especially when I know that they were involved with review of
earlier iterations of the patch.

I sometimes ask that a contributor wait a few days or a week for
others to comment prior to pushing it.  I'm not sure how that would work
when using these tags.  Could I give conditional 'Approved-by:' tag?  Maybe
something like this?:

    Approved-by: (name & email here)
    (But please wait a week for others to comment first.  If there are any
    objections or concerns, please address those first.)

> +
> + - Approved-by:
> +
> +   Used by responsible maintainers or global maintainers when
> +   a patch is ready to be upstreamed.  Some patches may touch multiple areas
> +   and require multiple approvals before landing (such as a maintainer only
> +   approving documentation), it is up to the maintainer giving the approval tag
> +   to make it clear when that a tag is not sufficient.
> +   Responsible, Global and Official FSF-appointed maintainers may approve their
> +   own patches, but it is recommended that they seek external approval before
> +   doing so.
> +

In the above paragraph (and perhaps in a few other placess too), it
seems that the formatting needs some adjustment.  (Some lines are
quite short while others seem a bit long.)

>  
>  			The Obvious Fix Rule
>  			--------------------
> -- 
> 2.40.1
> 

Re: [PATCH v2 1/1] [gdb]: add git trailer information on gdb/MAINTAINERS

[Bruno Larsen]

On 13/06/2023 01:59, Kevin Buettner wrote:
> Hi,
>
> Thanks for taking this on.  See my thoughts inline below...
>
> On Mon, 12 Jun 2023 12:01:40 +0200
> Bruno Larsen <blarsen@redhat.com> wrote:
>
>> The project has been using Tested-By (tb), Reviewed-By (rb) and
>> Approved-By (ab) for some time, but there has been no information to be
>> found in the actual repository. This commit changes that by adding
>> information about all git trailers to the MAINTAINERS file, so that it
> If you wish to document *all* git trailers, then 'Bug:' and
> 'Co-Authored-By:' should be documented too.

Good point, I will add them to v3.

I will just wait a bit for other people's input before sending a v3.

>> can be easily double-checked.
>>
>> The upstream discussion also brought up the use of Acked-by, which is
>> better defined in this commit.
>> ---
>>   gdb/MAINTAINERS | 56 ++++++++++++++++++++++++++++++++++++++++++-------
>>   1 file changed, 48 insertions(+), 8 deletions(-)
>>
>> diff --git a/gdb/MAINTAINERS b/gdb/MAINTAINERS
>> index 175595d5f17..22f7ecbd471 100644
>> --- a/gdb/MAINTAINERS
>> +++ b/gdb/MAINTAINERS
>> @@ -43,14 +43,9 @@ patch without review from another maintainer.  This especially includes
>>   patches which change internal interfaces (e.g. global functions, data
>>   structures) or external interfaces (e.g. user, remote, MI, et cetera).
>>   
>> -The term "review" is used in this file to describe several kinds of feedback
>> -from a maintainer: approval, rejection, and requests for changes or
>> -clarification with the intention of approving a revised version.  Review is
>> -a privilege and/or responsibility of various positions among the GDB
>> -Maintainers.  Of course, anyone - whether they hold a position but not the
>> -relevant one for a particular patch, or are just following along on the
>> -mailing lists for fun, or anything in between - may suggest changes or
>> -ask questions about a patch!
>> +The word "contributor" is used in this document to refer to any GDB
>> +developer listed above as well as folks who may have suggested some patches
>> +but aren't part of one of those categories for any reason.
>>   
>>   There's also a couple of other people who play special roles in the GDB
>>   community, separately from the patch process:
>> @@ -78,6 +73,51 @@ consensus among the global maintainers and any other involved parties.
>>   In cases where consensus can not be reached, the global maintainers may
>>   ask the official FSF-appointed GDB maintainers for a final decision.
>>   
>> +The term "review" is used in this file to describe several kinds of feedback
>> +from a maintainer: approval, rejection, and requests for changes or
>> +clarification with the intention of approving a revised version.  Approval is
> In the line above, you've replaced the word "Review" with "Approval".
> I think I like "Review" better.  (But there may be an even better word
> or phrase to use here.)

I replaced it on purpose, because the full original wording is "Review 
is a privilege and/or responsibility (...) [of] GDB maintainers." This 
seems to imply that a new contributor may not review a patch if they 
aren't a maintainer yet, when the feeling I had of the patch and 
community as a whole was that anyone can review, what is a privilege 
that needs to be earned is approving patches.  If you think my reasoning 
is wrong, do let me know!

>
>> +a privilege and/or responsibility of various positions among the GDB
>> +Maintainers.  Of course, anyone - whether they hold a position but not the
>> +relevant one for a particular patch, or are just following along on the
>> +mailing lists for fun, or anything in between - may suggest changes,
>> +ask questions about a patch or say if they believe a patch is fit for
>> +upstreaming!
>> +
>> +To ensure that patches are only pushed when approved, and to thank the
>> +contributors who take the time to review incoming patches, the GDB project
>> +uses trailers to identify who contributed and how.  All patches pushed
>> +upstream should have at least one Approved-By patches (with the exception of
>> +obvious patches, see below).  The trailers (or tags) currently in use are:
>> +
>> + - Acked-By:
>> +
>> +   Used when a contributor has taken a quick glance at a patch and agrees
>> +   with the direction outlined in the commit message, but hasn't evaluated
>> +   the code for correctness or regressions.
>> +
>> + - Tested-by:
>> +
>> +   Used when a contributor does not want to comment on the quality
>> +   of the code in the patch, but has tested and sees no regressions on their
>> +   hardware.
> For "Tested-by", I suggest this instead:
>
>      Used when a contributor has tested the patch and finds that it
>      fixes the claimed problem.  It may also be used to indicate that
>      the contributor has performed regression testing.  By itself, this
>      tag says nothing about the quality of the fix implemented by the
>      patch.
I like this better! I will use this on v3.
>
>> +
>> + - Reviewed-by:
>> +
>> +   Used when a contributors have looked at code and agrees with the changes,
> s/contributors have/contributor has/
> s/code/the code/
oops, sorry, will change.
>
>> +   but either don't have the authority or don't feel comfortable approving
> s/don't/doesn't/ ??
ah, yes, leftover from a previous draft.
>
>> +   the patch (usually due to unfamiliarity with the relevant part of the code).
> I think I'd remove the parenthetical remark.  There could be a variety of
> reasons for feeling uncomfortable with outright approving the change.
> One of the reasons that I sometimes don't approve a patch with which I
> agree is that I want to give other maintainers a chance to look at it
> first, especially when I know that they were involved with review of
> earlier iterations of the patch.
That's fair. I guess I don't need to give hints to the maintainers 
themselves here.
>
> I sometimes ask that a contributor wait a few days or a week for
> others to comment prior to pushing it.  I'm not sure how that would work
> when using these tags.  Could I give conditional 'Approved-by:' tag?  Maybe
> something like this?:
>
>      Approved-by: (name & email here)
>      (But please wait a week for others to comment first.  If there are any
>      objections or concerns, please address those first.)
This is how I would recommend you go about things, yes. With the rb tag, 
I would have to wait an unknown amount of time before pushing, until 
some other maintainer gives me the ab tag or until you decide to look at 
a patch you've already reviewed, which is extra work for you. 
Conditional approvals, on the other hand, are already how we deal with 
partial approval.
>> +
>> + - Approved-by:
>> +
>> +   Used by responsible maintainers or global maintainers when
>> +   a patch is ready to be upstreamed.  Some patches may touch multiple areas
>> +   and require multiple approvals before landing (such as a maintainer only
>> +   approving documentation), it is up to the maintainer giving the approval tag
>> +   to make it clear when that a tag is not sufficient.
>> +   Responsible, Global and Official FSF-appointed maintainers may approve their
>> +   own patches, but it is recommended that they seek external approval before
>> +   doing so.
>> +
> In the above paragraph (and perhaps in a few other placess too), it
> seems that the formatting needs some adjustment.  (Some lines are
> quite short while others seem a bit long.)
Ah... the wonders of multiple drafts in an editor that doesn't format 
automatically, hahaha. Thanks for pointing this out!

-- 
Cheers,
Bruno

>
>>   
>>   			The Obvious Fix Rule
>>   			--------------------
>> -- 
>> 2.40.1
>>