Attribution for patches

[ro...@us.ibm.com]

Someone sent in a small fix for the URL class the other day. I made that
change and checked it in. He then sent a message indicating that other
Apache projects give attributions (in the change log?) for patches, and why
didn't we do it. I'm saying 'he' here because it was sent to a mail account
on one of the apache servers, and I don't have any normal mail account
there. It kind of showed up partly when I logged onto the server via secure
CRT, so I could only see some of it.

Anyway, I guess we can discuss it. Personally, having written almost all of
it and not having given myself any attributions per se, I didn't really
think about it. And, personally, I'm not sure I want to add any more gorp
to the change logs. But, if everyone else besides me thinks this is
important to get people to contribute, then certainly I will agree.

We do have a 'contributors' file, but I assume its reserved for more
significant (sizewise I mean, any bug fix is significant in terms of
usefulness :-) contributions.

Anyway.. anyone want to share their feelings on this subject? Do we need to
put in attributions for provided patches, no matter how big or how small
(Wasn't that a Dr. Seus story?)

----------------------------------------
Dean Roddey
Software Weenie
IBM Center for Java Technology - Silicon Valley
roddey@us.ibm.com

Re: Attribution for patches

[Stefano Mazzocchi <st...@apache.org>]

Dirk-Willem van Gulik wrote:
> 
> roddey@us.ibm.com wrote:
> 
> > We do have a 'contributors' file, but I assume its reserved for more
> > significant (sizewise I mean, any bug fix is significant in terms of
> > usefulness :-) contributions.
> 
> The 'Changes' file in the httpd project has short and to the point
> entries like:
> 
>   *) Add an explicit charset=iso-8859-1 to pages generated by
>      ap_send_error_response(), such as the default 404 page.
>      [Marc Slemko]
> 
>   *) Added protocol(%m)/method(%H) logging to the log format.
>      Suggested by Peter W <pe...@usa.net> [dirkx]
> 
>   *) Fix intermittent SEGV in ap_proxy_cache_error() in
>      src/modules/proxy_util.c where a NULL filepointer and
>      temporary filename were closed and unlinked.
>      [Graham Leggett <mi...@sharp.fm>,
>      Tim Costello <tj...@socs.uts.edu.au>] PR#3178
> 
> or whatever. I.e. who suggested it where appropriate, and
> who actually commited it (and should get part of the blame)
> or either of the two. Or a ptr to a bug #.
> 
> This is a good compromize I guess.

You might want to consider the Changes DTD we use in Cocoon to keep
track of contributions: here is an example of usage of our DTD.

<changes title="History of Changes">
 <devs>
  <person name="Bill Gates" email="$$$@microsoft.com" id="BG"/>
  ...
 </devs>

 <release version="52.3.4" date="to be determined">
  <action dev="BG" type="add" due-to="John Smith"
due-to-email="john@smith.com">
   Added something but I don't know what it does.
  </action>
 </release>

 ...

</changes>

then this is transformed into HTML and placed on the web site for
reference.

We should have something like this for every xml.apache project.

Stefano.

Re: Attribution for patches

[Dirk-Willem van Gulik <di...@webweaving.org>]

roddey@us.ibm.com wrote:

> We do have a 'contributors' file, but I assume its reserved for more
> significant (sizewise I mean, any bug fix is significant in terms of
> usefulness :-) contributions.

The 'Changes' file in the httpd project has short and to the point
entries like:

  *) Add an explicit charset=iso-8859-1 to pages generated by
     ap_send_error_response(), such as the default 404 page.
     [Marc Slemko]

  *) Added protocol(%m)/method(%H) logging to the log format.
     Suggested by Peter W <pe...@usa.net> [dirkx]

  *) Fix intermittent SEGV in ap_proxy_cache_error() in
     src/modules/proxy_util.c where a NULL filepointer and
     temporary filename were closed and unlinked.
     [Graham Leggett <mi...@sharp.fm>,
     Tim Costello <tj...@socs.uts.edu.au>] PR#3178

or whatever. I.e. who suggested it where appropriate, and
who actually commited it (and should get part of the blame)
or either of the two. Or a ptr to a bug #.

This is a good compromize I guess.

Dw.

Re: Attribution for patches

[Stefano Mazzocchi <st...@apache.org>]

Rahul Jain wrote:
> 
> Not everyone likes their email address to be mentioned. I have to
> now go any remove this person's email address for the patch that I
> checked into the repository.
> 
> We really need some kind of policy here. I don't know if there is
> one already. In case its not there, How about...
> 
> Unless the patch submitter specifically asks not to:
>   the person's name and email address will be provided
>   in the commit message.
> 
> Request to omit the email address may be entertained if
> mentioned at the time the patch was submitted. Once committed,
> its tedious to make edits on all the affected files.
> 
> Request to omit name's will not be entertained, i.e. no anonymous
> submissions.

I like this policy, but I would love more to have a more strict policy
that includes the usage of a common format for changes. If we use a DTD,
we can place the name/email pairs in the DTD and specify this guy
doesn't want the mail address to be specified.

Then the XSLT stylesheet will format the output depending on such a
flag, but we keep track of all people and their mail addresses.

What do you think?

Stefano.

Re: Attribution for patches

[Rahul Jain <ra...@apache.org>]

Not everyone likes their email address to be mentioned. I have to
now go any remove this person's email address for the patch that I
checked into the repository.

We really need some kind of policy here. I don't know if there is
one already. In case its not there, How about...


Unless the patch submitter specifically asks not to:
  the person's name and email address will be provided 
  in the commit message.

Request to omit the email address may be entertained if
mentioned at the time the patch was submitted. Once committed,
its tedious to make edits on all the affected files.

Request to omit name's will not be entertained, i.e. no anonymous
submissions.


rahul

Re: Attribution for patches

[Pierpaolo Fumagalli <pi...@apache.org>]

Stefano Mazzocchi wrote:
> 
> If felt so good, now I want my name everywhere :) ...yeah, I know...
> that's probably too much :) but you get my point.

... You already have your name EVERYWHERE :) :) :)

	Pier

-- 
--------------------------------------------------------------------
-          P              I              E              R          -
stable structure erected over water to allow the docking of seacraft
<ma...@betaversion.org>    <http://www.betaversion.org/~pier/>
--------------------------------------------------------------------
- ApacheCON Y2K: Come to the official Apache developers conference -
-------------------- <http://www.apachecon.com> --------------------

Re: Attribution for patches

[Stefano Mazzocchi <st...@apache.org>]

twleung@sauria.com wrote:
> 
> Patches should always be attributed.  The project convention is to
> attribute patches.  I vote -1 on any proposal to stop attributing patches.

I'd quit this project if this ever happens. Serious.

Credits are the only thing that drives open source and the only way an
open source project can pay back its contributors. During JServ
development there was a big discussion about "what" should be credited.
Most of us agreed that anything that influences the state of the program
should be given credits: even a very good bug report with no patch.

Of course, people understand themselves when it's honest to ask for such
credits to be included, but normally people are afraid to talk... look
at cocoon changes.... lots of them, lots of people listed...some have
big names, some don't (probably, yet!).

My first patch on an apache project happened in nov 97: jserv logs were
using Stack instead of Vector and you were getting the logs backwards
causing pain for servlet debugging. The patch was included and I loved
to see my name on that list.

If felt so good, now I want my name everywhere :) ...yeah, I know...
that's probably too much :) but you get my point.

Stefano.

Re: Attribution for patches

[tw...@sauria.com]

Patches should always be attributed.  The project convention is to
attribute patches.  I vote -1 on any proposal to stop attributing patches.

Ted
----- Original Message ----- 
From: <ro...@us.ibm.com>
To: <xe...@xml.apache.org>
Sent: Wednesday, February 02, 2000 3:41 PM
Subject: Attribution for patches


> 
> 
> 
> Someone sent in a small fix for the URL class the other day. I made that
> change and checked it in. He then sent a message indicating that other
> Apache projects give attributions (in the change log?) for patches, and why
> didn't we do it. I'm saying 'he' here because it was sent to a mail account
> on one of the apache servers, and I don't have any normal mail account
> there. It kind of showed up partly when I logged onto the server via secure
> CRT, so I could only see some of it.
> 
> Anyway, I guess we can discuss it. Personally, having written almost all of
> it and not having given myself any attributions per se, I didn't really
> think about it. And, personally, I'm not sure I want to add any more gorp
> to the change logs. But, if everyone else besides me thinks this is
> important to get people to contribute, then certainly I will agree.
> 
> We do have a 'contributors' file, but I assume its reserved for more
> significant (sizewise I mean, any bug fix is significant in terms of
> usefulness :-) contributions.
> 
> Anyway.. anyone want to share their feelings on this subject? Do we need to
> put in attributions for provided patches, no matter how big or how small
> (Wasn't that a Dr. Seus story?)
> 
> ----------------------------------------
> Dean Roddey
> Software Weenie
> IBM Center for Java Technology - Silicon Valley
> roddey@us.ibm.com
> 
> 
>