discuss: stop adding 'via' from CHANGES.txt entries (take two)

classic Classic list List threaded Threaded
12 messages Options
Reply | Threaded
Open this post in threaded view
|

discuss: stop adding 'via' from CHANGES.txt entries (take two)

Robert Muir
I raised this issue a few years ago, and no consensus was reached [1]

I'm asking if we can take the time to revisit the issue. Back then it
was subversion days, and you had "patch-uploaders" and "contributors".
With git now, I believe the situation is even a bit more extreme,
because the committer is the contributor and the lucene "committer"
was really the "pusher".

On the other hand, there were some reasons against removing this
before. In particular some mentioned that it conveyed meaning about
who might be the best person to ping about a particular area of the
code. If this is still the case, I'd ask that we discuss alternative
ways that it could be accomplished (such as wiki page perhaps
linked-to HowToContribute that ppl can edit).

I wrote a new summary/argument inline, but see the linked thread for
the previous discussion:


In the past CHANGES.txt entries from a contributor have also had the
name of the committer with a 'via' entry.

e.g.:

LUCENE-1234: optimized FooBar. (Jane Doe via Joe Schmoe).

I propose we stop adding the committer name (via Joe Schmoe). It seems
to diminish the value of the contribution. It reminds me of a
professor adding a second author by default or something like that. If
someone really wants to know who committed the change, I think its
fair that they look at version control history?

1. http://mail-archives.apache.org/mod_mbox/lucene-dev/201206.mbox/%3CCAOdYfZW65MXrzyRPsvBD0C6c4X%2BLuQX4oVec%3DyR_PCPgTQrnhQ%40mail.gmail.com%3E

---------------------------------------------------------------------
To unsubscribe, e-mail: [hidden email]
For additional commands, e-mail: [hidden email]

Reply | Threaded
Open this post in threaded view
|

Re: discuss: stop adding 'via' from CHANGES.txt entries (take two)

Erick Erickson
Robert:

I don't have strong feelings either way. Personally I use "via" to
indicate that I didn't have much to do with the hard parts, I just was
the "committer fingers". If I've been more involved I just add my name
as a co-contributor (last). Basically it's a question of "how much
credit do I think I deserve?". If very little I use "via". If I'm more
involved, I just add a comma and my name.

But that's a nuance that I suppose varies by person, so I'm happy
either way. Your point that the tighter integration with Git is well
taken, we can trace things back to whoever committed things pretty
easily.

I'm -1 to having to remember to go to another place like a Wiki page,
too easy to forget. And I don't think we really need it.

Erick


On Mon, Jun 4, 2018 at 5:47 PM, Robert Muir <[hidden email]> wrote:

> I raised this issue a few years ago, and no consensus was reached [1]
>
> I'm asking if we can take the time to revisit the issue. Back then it
> was subversion days, and you had "patch-uploaders" and "contributors".
> With git now, I believe the situation is even a bit more extreme,
> because the committer is the contributor and the lucene "committer"
> was really the "pusher".
>
> On the other hand, there were some reasons against removing this
> before. In particular some mentioned that it conveyed meaning about
> who might be the best person to ping about a particular area of the
> code. If this is still the case, I'd ask that we discuss alternative
> ways that it could be accomplished (such as wiki page perhaps
> linked-to HowToContribute that ppl can edit).
>
> I wrote a new summary/argument inline, but see the linked thread for
> the previous discussion:
>
>
> In the past CHANGES.txt entries from a contributor have also had the
> name of the committer with a 'via' entry.
>
> e.g.:
>
> LUCENE-1234: optimized FooBar. (Jane Doe via Joe Schmoe).
>
> I propose we stop adding the committer name (via Joe Schmoe). It seems
> to diminish the value of the contribution. It reminds me of a
> professor adding a second author by default or something like that. If
> someone really wants to know who committed the change, I think its
> fair that they look at version control history?
>
> 1. http://mail-archives.apache.org/mod_mbox/lucene-dev/201206.mbox/%3CCAOdYfZW65MXrzyRPsvBD0C6c4X%2BLuQX4oVec%3DyR_PCPgTQrnhQ%40mail.gmail.com%3E
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: [hidden email]
> For additional commands, e-mail: [hidden email]
>

---------------------------------------------------------------------
To unsubscribe, e-mail: [hidden email]
For additional commands, e-mail: [hidden email]

Reply | Threaded
Open this post in threaded view
|

Re: discuss: stop adding 'via' from CHANGES.txt entries (take two)

Adrien Grand
+1 to remove the name of the pusher from the changelog

Le mar. 5 juin 2018 à 06:29, Erick Erickson <[hidden email]> a écrit :
Robert:

I don't have strong feelings either way. Personally I use "via" to
indicate that I didn't have much to do with the hard parts, I just was
the "committer fingers". If I've been more involved I just add my name
as a co-contributor (last). Basically it's a question of "how much
credit do I think I deserve?". If very little I use "via". If I'm more
involved, I just add a comma and my name.

But that's a nuance that I suppose varies by person, so I'm happy
either way. Your point that the tighter integration with Git is well
taken, we can trace things back to whoever committed things pretty
easily.

I'm -1 to having to remember to go to another place like a Wiki page,
too easy to forget. And I don't think we really need it.

Erick


On Mon, Jun 4, 2018 at 5:47 PM, Robert Muir <[hidden email]> wrote:
> I raised this issue a few years ago, and no consensus was reached [1]
>
> I'm asking if we can take the time to revisit the issue. Back then it
> was subversion days, and you had "patch-uploaders" and "contributors".
> With git now, I believe the situation is even a bit more extreme,
> because the committer is the contributor and the lucene "committer"
> was really the "pusher".
>
> On the other hand, there were some reasons against removing this
> before. In particular some mentioned that it conveyed meaning about
> who might be the best person to ping about a particular area of the
> code. If this is still the case, I'd ask that we discuss alternative
> ways that it could be accomplished (such as wiki page perhaps
> linked-to HowToContribute that ppl can edit).
>
> I wrote a new summary/argument inline, but see the linked thread for
> the previous discussion:
>
>
> In the past CHANGES.txt entries from a contributor have also had the
> name of the committer with a 'via' entry.
>
> e.g.:
>
> LUCENE-1234: optimized FooBar. (Jane Doe via Joe Schmoe).
>
> I propose we stop adding the committer name (via Joe Schmoe). It seems
> to diminish the value of the contribution. It reminds me of a
> professor adding a second author by default or something like that. If
> someone really wants to know who committed the change, I think its
> fair that they look at version control history?
>
> 1. http://mail-archives.apache.org/mod_mbox/lucene-dev/201206.mbox/%3CCAOdYfZW65MXrzyRPsvBD0C6c4X%2BLuQX4oVec%3DyR_PCPgTQrnhQ%40mail.gmail.com%3E
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: [hidden email]
> For additional commands, e-mail: [hidden email]
>

---------------------------------------------------------------------
To unsubscribe, e-mail: [hidden email]
For additional commands, e-mail: [hidden email]

Reply | Threaded
Open this post in threaded view
|

Re: discuss: stop adding 'via' from CHANGES.txt entries (take two)

Alan Woodward-3
+1 to Robert’s suggestion

On 5 Jun 2018, at 09:03, Adrien Grand <[hidden email]> wrote:

+1 to remove the name of the pusher from the changelog

Le mar. 5 juin 2018 à 06:29, Erick Erickson <[hidden email]> a écrit :
Robert:

I don't have strong feelings either way. Personally I use "via" to
indicate that I didn't have much to do with the hard parts, I just was
the "committer fingers". If I've been more involved I just add my name
as a co-contributor (last). Basically it's a question of "how much
credit do I think I deserve?". If very little I use "via". If I'm more
involved, I just add a comma and my name.

But that's a nuance that I suppose varies by person, so I'm happy
either way. Your point that the tighter integration with Git is well
taken, we can trace things back to whoever committed things pretty
easily.

I'm -1 to having to remember to go to another place like a Wiki page,
too easy to forget. And I don't think we really need it.

Erick


On Mon, Jun 4, 2018 at 5:47 PM, Robert Muir <[hidden email]> wrote:
> I raised this issue a few years ago, and no consensus was reached [1]
>
> I'm asking if we can take the time to revisit the issue. Back then it
> was subversion days, and you had "patch-uploaders" and "contributors".
> With git now, I believe the situation is even a bit more extreme,
> because the committer is the contributor and the lucene "committer"
> was really the "pusher".
>
> On the other hand, there were some reasons against removing this
> before. In particular some mentioned that it conveyed meaning about
> who might be the best person to ping about a particular area of the
> code. If this is still the case, I'd ask that we discuss alternative
> ways that it could be accomplished (such as wiki page perhaps
> linked-to HowToContribute that ppl can edit).
>
> I wrote a new summary/argument inline, but see the linked thread for
> the previous discussion:
>
>
> In the past CHANGES.txt entries from a contributor have also had the
> name of the committer with a 'via' entry.
>
> e.g.:
>
> LUCENE-1234: optimized FooBar. (Jane Doe via Joe Schmoe).
>
> I propose we stop adding the committer name (via Joe Schmoe). It seems
> to diminish the value of the contribution. It reminds me of a
> professor adding a second author by default or something like that. If
> someone really wants to know who committed the change, I think its
> fair that they look at version control history?
>
> 1. http://mail-archives.apache.org/mod_mbox/lucene-dev/201206.mbox/%3CCAOdYfZW65MXrzyRPsvBD0C6c4X%2BLuQX4oVec%3DyR_PCPgTQrnhQ%40mail.gmail.com%3E
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: [hidden email]
> For additional commands, e-mail: [hidden email]
>

---------------------------------------------------------------------
To unsubscribe, e-mail: [hidden email]
For additional commands, e-mail: [hidden email]


Reply | Threaded
Open this post in threaded view
|

Re: discuss: stop adding 'via' from CHANGES.txt entries (take two)

Yonik Seeley
In reply to this post by Robert Muir
I don't have much of an opinion about "via" one way or the other,
however I think we should avoid using the mental model of authorship
for CHANGES.txt.
We've generally been listing people who made meaningful contributions
to the patch, including sometimes the person who opened the issue for
example.

-Yonik

---------------------------------------------------------------------
To unsubscribe, e-mail: [hidden email]
For additional commands, e-mail: [hidden email]

Reply | Threaded
Open this post in threaded view
|

Re: discuss: stop adding 'via' from CHANGES.txt entries (take two)

Mark Miller-3
I  have the same opinion as last time. Taking ownership of actually committing something to the code base is an important attribution and that is why it has been included in CHANGES. I don't agree that it takes away credit at all - via means the commit went through you, which is an accurate reflection of things. Committing others work is a major contribution and should be called out, for the positives that it creates as well as the responsibility for that change you have undertaken by being a very key part of the via route.

- Mark

On Wed, Jun 6, 2018 at 8:10 AM Yonik Seeley <[hidden email]> wrote:
I don't have much of an opinion about "via" one way or the other,
however I think we should avoid using the mental model of authorship
for CHANGES.txt.
We've generally been listing people who made meaningful contributions
to the patch, including sometimes the person who opened the issue for
example.

-Yonik

---------------------------------------------------------------------
To unsubscribe, e-mail: [hidden email]
For additional commands, e-mail: [hidden email]

--
Reply | Threaded
Open this post in threaded view
|

Re: discuss: stop adding 'via' from CHANGES.txt entries (take two)

Joel Bernstein
I agree with Marks position on this, the information of who committed has significant value.


On Wed, Jun 6, 2018 at 10:14 PM, Mark Miller <[hidden email]> wrote:
I  have the same opinion as last time. Taking ownership of actually committing something to the code base is an important attribution and that is why it has been included in CHANGES. I don't agree that it takes away credit at all - via means the commit went through you, which is an accurate reflection of things. Committing others work is a major contribution and should be called out, for the positives that it creates as well as the responsibility for that change you have undertaken by being a very key part of the via route.

- Mark

On Wed, Jun 6, 2018 at 8:10 AM Yonik Seeley <[hidden email]> wrote:
I don't have much of an opinion about "via" one way or the other,
however I think we should avoid using the mental model of authorship
for CHANGES.txt.
We've generally been listing people who made meaningful contributions
to the patch, including sometimes the person who opened the issue for
example.

-Yonik

---------------------------------------------------------------------
To unsubscribe, e-mail: [hidden email]
For additional commands, e-mail: [hidden email]

--

Reply | Threaded
Open this post in threaded view
|

Re: discuss: stop adding 'via' from CHANGES.txt entries (take two)

Jan Høydahl / Cominvent
I agree with Mark that it is a huge and important part of keeping Lucene/Solr a welcoming
community, that the existing committers take time to guide contributors. Keeping the "via"
part of the changelog also makes it very easy to spot potential candiates for committership,
and avoid the "I thought he/she was already a committer" type of comments we've seen :)

--
Jan Høydahl, search solution architect
Cominvent AS - www.cominvent.com

7. jun. 2018 kl. 14:59 skrev Joel Bernstein <[hidden email]>:

I agree with Marks position on this, the information of who committed has significant value.


On Wed, Jun 6, 2018 at 10:14 PM, Mark Miller <[hidden email]> wrote:
I  have the same opinion as last time. Taking ownership of actually committing something to the code base is an important attribution and that is why it has been included in CHANGES. I don't agree that it takes away credit at all - via means the commit went through you, which is an accurate reflection of things. Committing others work is a major contribution and should be called out, for the positives that it creates as well as the responsibility for that change you have undertaken by being a very key part of the via route.

- Mark

On Wed, Jun 6, 2018 at 8:10 AM Yonik Seeley <[hidden email]> wrote:
I don't have much of an opinion about "via" one way or the other,
however I think we should avoid using the mental model of authorship
for CHANGES.txt.
We've generally been listing people who made meaningful contributions
to the patch, including sometimes the person who opened the issue for
example.

-Yonik

---------------------------------------------------------------------
To unsubscribe, e-mail: [hidden email]
For additional commands, e-mail: [hidden email]

--


Reply | Threaded
Open this post in threaded view
|

Re: discuss: stop adding 'via' from CHANGES.txt entries (take two)

Michael McCandless-2
In reply to this post by Robert Muir
+1 to remove via; who actually pushed the change is (or should be) unimportant, and is easily derived from git history if we really do ever need it.

Your ideas will go further if you don't insist on going with them ;)


On Mon, Jun 4, 2018 at 8:47 PM, Robert Muir <[hidden email]> wrote:
I raised this issue a few years ago, and no consensus was reached [1]

I'm asking if we can take the time to revisit the issue. Back then it
was subversion days, and you had "patch-uploaders" and "contributors".
With git now, I believe the situation is even a bit more extreme,
because the committer is the contributor and the lucene "committer"
was really the "pusher".

On the other hand, there were some reasons against removing this
before. In particular some mentioned that it conveyed meaning about
who might be the best person to ping about a particular area of the
code. If this is still the case, I'd ask that we discuss alternative
ways that it could be accomplished (such as wiki page perhaps
linked-to HowToContribute that ppl can edit).

I wrote a new summary/argument inline, but see the linked thread for
the previous discussion:


In the past CHANGES.txt entries from a contributor have also had the
name of the committer with a 'via' entry.

e.g.:

LUCENE-1234: optimized FooBar. (Jane Doe via Joe Schmoe).

I propose we stop adding the committer name (via Joe Schmoe). It seems
to diminish the value of the contribution. It reminds me of a
professor adding a second author by default or something like that. If
someone really wants to know who committed the change, I think its
fair that they look at version control history?

1. http://mail-archives.apache.org/mod_mbox/lucene-dev/201206.mbox/%3CCAOdYfZW65MXrzyRPsvBD0C6c4X%2BLuQX4oVec%3DyR_PCPgTQrnhQ%40mail.gmail.com%3E

---------------------------------------------------------------------
To unsubscribe, e-mail: [hidden email]
For additional commands, e-mail: [hidden email]


Reply | Threaded
Open this post in threaded view
|

Re: discuss: stop adding 'via' from CHANGES.txt entries (take two)

Gus Heck
FWIW As a non-committer contributor I don't mind the "via". Seems legit to give the committer some credit for their review and merging work. 

On Tue, Jun 12, 2018 at 2:53 PM, Michael McCandless <[hidden email]> wrote:
+1 to remove via; who actually pushed the change is (or should be) unimportant, and is easily derived from git history if we really do ever need it.

Your ideas will go further if you don't insist on going with them ;)


On Mon, Jun 4, 2018 at 8:47 PM, Robert Muir <[hidden email]> wrote:
I raised this issue a few years ago, and no consensus was reached [1]

I'm asking if we can take the time to revisit the issue. Back then it
was subversion days, and you had "patch-uploaders" and "contributors".
With git now, I believe the situation is even a bit more extreme,
because the committer is the contributor and the lucene "committer"
was really the "pusher".

On the other hand, there were some reasons against removing this
before. In particular some mentioned that it conveyed meaning about
who might be the best person to ping about a particular area of the
code. If this is still the case, I'd ask that we discuss alternative
ways that it could be accomplished (such as wiki page perhaps
linked-to HowToContribute that ppl can edit).

I wrote a new summary/argument inline, but see the linked thread for
the previous discussion:


In the past CHANGES.txt entries from a contributor have also had the
name of the committer with a 'via' entry.

e.g.:

LUCENE-1234: optimized FooBar. (Jane Doe via Joe Schmoe).

I propose we stop adding the committer name (via Joe Schmoe). It seems
to diminish the value of the contribution. It reminds me of a
professor adding a second author by default or something like that. If
someone really wants to know who committed the change, I think its
fair that they look at version control history?

1. http://mail-archives.apache.org/mod_mbox/lucene-dev/201206.mbox/%3CCAOdYfZW65MXrzyRPsvBD0C6c4X%2BLuQX4oVec%3DyR_PCPgTQrnhQ%40mail.gmail.com%3E

---------------------------------------------------------------------
To unsubscribe, e-mail: [hidden email]
For additional commands, e-mail: [hidden email]





--
Reply | Threaded
Open this post in threaded view
|

Re: discuss: stop adding 'via' from CHANGES.txt entries (take two)

Uwe Schindler
+1 to remove via if it's a stupid merge.

But there are cases where the patch is taken and improved by the committer. In that case the committer should be added with a comma only.

Uwe

Am June 12, 2018 7:05:57 PM UTC schrieb Gus Heck <[hidden email]>:
FWIW As a non-committer contributor I don't mind the "via". Seems legit to give the committer some credit for their review and merging work. 

On Tue, Jun 12, 2018 at 2:53 PM, Michael McCandless <[hidden email]> wrote:
+1 to remove via; who actually pushed the change is (or should be) unimportant, and is easily derived from git history if we really do ever need it.

Your ideas will go further if you don't insist on going with them ;)


On Mon, Jun 4, 2018 at 8:47 PM, Robert Muir <[hidden email]> wrote:
I raised this issue a few years ago, and no consensus was reached [1]

I'm asking if we can take the time to revisit the issue. Back then it
was subversion days, and you had "patch-uploaders" and "contributors".
With git now, I believe the situation is even a bit more extreme,
because the committer is the contributor and the lucene "committer"
was really the "pusher".

On the other hand, there were some reasons against removing this
before. In particular some mentioned that it conveyed meaning about
who might be the best person to ping about a particular area of the
code. If this is still the case, I'd ask that we discuss alternative
ways that it could be accomplished (such as wiki page perhaps
linked-to HowToContribute that ppl can edit).

I wrote a new summary/argument inline, but see the linked thread for
the previous discussion:


In the past CHANGES.txt entries from a contributor have also had the
name of the committer with a 'via' entry.

e.g.:

LUCENE-1234: optimized FooBar. (Jane Doe via Joe Schmoe).

I propose we stop adding the committer name (via Joe Schmoe). It seems
to diminish the value of the contribution. It reminds me of a
professor adding a second author by default or something like that. If
someone really wants to know who committed the change, I think its
fair that they look at version control history?

1. http://mail-archives.apache.org/mod_mbox/lucene-dev/201206.mbox/%3CCAOdYfZW65MXrzyRPsvBD0C6c4X%2BLuQX4oVec%3DyR_PCPgTQrnhQ%40mail.gmail.com%3E

---------------------------------------------------------------------
To unsubscribe, e-mail: [hidden email]
For additional commands, e-mail: [hidden email]





--

--
Uwe Schindler
Achterdiek 19, 28357 Bremen
https://www.thetaphi.de
Reply | Threaded
Open this post in threaded view
|

Re: discuss: stop adding 'via' from CHANGES.txt entries (take two)

Dawid Weiss-2
In reply to this post by Michael McCandless-2
> +1 to remove via; who actually pushed the change is (or should be) unimportant,

I don't care much for credit and am pretty much indifferent to any of
these options. The way I see it is that 'via'  is not much
of a credit but a *responsibility* -- the committer should have
reviewed the change she or he is contributing on behalf of the
non-committer. Sure, git history is fine too but note that it's not
ideal -- it may not mention the actual contributor of a patch
if it's a manual application of a diff file and in cases like this
it'd require more complex processing than just scanning through
CHANGES.txt.... so I'd opt for simplicity of keeping it together in
CHANGES.txt (easier to scan through, doesn't take any
credit away from the author), but I'll live with any decision.

Dawid

---------------------------------------------------------------------
To unsubscribe, e-mail: [hidden email]
For additional commands, e-mail: [hidden email]