Re: [ofbiz-framework] branch release17.12 updated: Fixed: add XML declaration in “web.xml” files (OFBIZ-6993)

[Jacques Le Roux]

Hi Mathieu, All,

Not a big deal (it's only in the commits log), I'm not sure we want stuff like:

Le 27/11/2019 à 17:48, [hidden email] a écrit :
> Author: Samuel Trégouët<[hidden email]>

Normally we rather use something like

    Thanks to Samuel Tregouet for this fix

I did not notice yet, we have already 10 of them in Git repo.

Not sure we want to avoid them, maybe if Samuel has an ICLA it's OK, unsure on my side...

I mean is that something Git forces us, or could we prevent it if needed?

Jacques

[Mathieu Lirzin]

Hello Jacques,

Jacques Le Roux <[hidden email]> writes:

> Not a big deal (it's only in the commits log), I'm not sure we want stuff like:
>
> Le 27/11/2019 à 17:48, [hidden email] a écrit :
>> Author: Samuel Trégouët<[hidden email]>
>
> Normally we rather use something like
>
>    Thanks to Samuel Tregouet for this fix
>
> I did not notice yet, we have already 10 of them in Git repo.
>
> Not sure we want to avoid them, maybe if Samuel has an ICLA it's OK, unsure on my side...
>
> I mean is that something Git forces us, or could we prevent it if needed?

The fact that Samuel was defined as the author of the commit was a
conscient action on my side (i.e. Git do not force us to do so).

Since it is a very common practice to commit somebody else work, in Git
there is a clear distinction between the author and the committer which
can is done with ‘git commit --author="..."’ when both are not the same
person. This provides more explicit credits to the actual author without
relying on ad-hoc conventions.

As I understand it, the practice of committing somebody else work under
the committer name (not the author name) and adding credits in the
commit log to the actual author was only the result of a limitation of
Subversion which does not handle the scenario of committing somebody
else work.

I didn't bother discuss the usage of this Git feature before hand
because I thought it would be obviously accepted by the community, but
since it seems not necessarily natural for you (an probably others) I am
open to the discussion now. :-)

Indeed I am aware that any author should sign an ICLA, I assumed that
Samuel have signed but did not actually check.

Samuel: have you signed the ICLA ?

--
Mathieu Lirzin
GPG: F2A3 8D7E EB2B 6640 5761  070D 0ADE E100 9460 4D37

[Jacques Le Roux]

Hi Mathieu, Samuel,

It was late, actually if I refer to http://www.apache.org/dev/committers.html#applying-patches you are actually right Mathieu.

Even an ICLA is not needed if Samuel openly contributed a patch[1]. Through Jira is preferred, I guess it was done that way.

Of course, if not done yet, Samuel signing would be good :)

I was confused with author in code (we don't want it), sorry for the noise.

[1] https://markmail.org/message/y3j3t25tjwsn2dhc (Yoav was one of our mentors when incubating)

Jacques


Le 27/11/2019 à 22:31, Mathieu Lirzin a écrit :

> Hello Jacques,
>
> Jacques Le Roux <[hidden email]> writes:
>
>> Not a big deal (it's only in the commits log), I'm not sure we want stuff like:
>>
>> Le 27/11/2019 à 17:48, [hidden email] a écrit :
>>> Author: Samuel Trégouët<[hidden email]>
>> Normally we rather use something like
>>
>>     Thanks to Samuel Tregouet for this fix
>>
>> I did not notice yet, we have already 10 of them in Git repo.
>>
>> Not sure we want to avoid them, maybe if Samuel has an ICLA it's OK, unsure on my side...
>>
>> I mean is that something Git forces us, or could we prevent it if needed?
> The fact that Samuel was defined as the author of the commit was a
> conscient action on my side (i.e. Git do not force us to do so).
>
> Since it is a very common practice to commit somebody else work, in Git
> there is a clear distinction between the author and the committer which
> can is done with ‘git commit --author="..."’ when both are not the same
> person. This provides more explicit credits to the actual author without
> relying on ad-hoc conventions.
>
> As I understand it, the practice of committing somebody else work under
> the committer name (not the author name) and adding credits in the
> commit log to the actual author was only the result of a limitation of
> Subversion which does not handle the scenario of committing somebody
> else work.
>
> I didn't bother discuss the usage of this Git feature before hand
> because I thought it would be obviously accepted by the community, but
> since it seems not necessarily natural for you (an probably others) I am
> open to the discussion now. :-)
>
> Indeed I am aware that any author should sign an ICLA, I assumed that
> Samuel have signed but did not actually check.
>
> Samuel: have you signed the ICLA ?
>