You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@commons.apache.org by Gilles <gi...@harfang.homelinux.org> on 2013/10/31 14:26:59 UTC
[Math] "due-to" attribute in "changes.xml"
Hello.
Are there criteria about filling the "due-to" attribute of an issue
record in the "changes.xml" file?
Current practice seems that reporting an issue does not by itself
warrants such an attribution.
Indeed, as I understand it, the attribute is a place-holder for when
an issue is fixed by a contributor who hasn't commit access. IMHO,
this implies that the reporter (or another contributor) provided a
patch or non-trivial insights that led to the fix.
IOW, when a developer with commit access fixes a bug or implements a
feature request mostly by himself, the name of the original reporter
should not appear in the release notes, as if he were the contributor.
Regards,
Gilles
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@commons.apache.org
For additional commands, e-mail: dev-help@commons.apache.org
Re: [Math] "due-to" attribute in "changes.xml"
Posted by Gilles <gi...@harfang.homelinux.org>.
On Mon, 4 Nov 2013 20:15:01 +0100, Jochen Wiedmann wrote:
> On Thu, Oct 31, 2013 at 2:26 PM, Gilles
> <gi...@harfang.homelinux.org>wrote:
>
>> Hello.
>>
>> Are there criteria about filling the "due-to" attribute of an issue
>> record in the "changes.xml" file?
>
>
>
> Be generous with using it: I've seen examples of people being quite
> stimulated for more work by them included in that list.
I don't have a problem with mentioning more people there. I have a
problem
with having no policy as to how to associate the cited work with the
people
who did it.
Beyond fairness, being specific about attribution also avoids the risk
of
bad feelings wondering why one name is mentioned while another isn't
(as is
the case _now_).
As far as I can deduce from past releases and the SVN log, there was an
assumed policy of mentioning people who produced patches.
IMO, this issue is indeed too important (for the reason you give above)
to
let each committer decide independently what is published in the
release
notes. If we agree on this, then I don't understand why people would
prefer
to sweep this under the rug (delivering warm-feeling generalities),
rather
than propose a solution (concrete policy).
Regards,
Gilles
> Jochen
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@commons.apache.org
For additional commands, e-mail: dev-help@commons.apache.org
Re: [Math] "due-to" attribute in "changes.xml"
Posted by Jochen Wiedmann <jo...@gmail.com>.
On Thu, Oct 31, 2013 at 2:26 PM, Gilles <gi...@harfang.homelinux.org>wrote:
> Hello.
>
> Are there criteria about filling the "due-to" attribute of an issue
> record in the "changes.xml" file?
Be generous with using it: I've seen examples of people being quite
stimulated for more work by them included in that list.
Jochen
Re: [Math] "due-to" attribute in "changes.xml"
Posted by Gilles <gi...@harfang.homelinux.org>.
On Thu, 31 Oct 2013 17:35:50 +0100, Thomas Neidhart wrote:
> On 10/31/2013 05:18 PM, Gilles wrote:
>> On Thu, 31 Oct 2013 15:53:29 +0000, sebb wrote:
>>> On 31 October 2013 14:24, Gilles <gi...@harfang.homelinux.org>
>>> wrote:
>>>> On Thu, 31 Oct 2013 13:37:02 +0000, sebb wrote:
>>>>>
>>>>> On 31 October 2013 13:26, Gilles <gi...@harfang.homelinux.org>
>>>>> wrote:
>>>>>>
>>>>>> Hello.
>>>>>>
>>>>>> Are there criteria about filling the "due-to" attribute of an
>>>>>> issue
>>>>>> record in the "changes.xml" file?
>>>>>>
>>>>>> Current practice seems that reporting an issue does not by
>>>>>> itself
>>>>>> warrants such an attribution.
>>>>>> Indeed, as I understand it, the attribute is a place-holder for
>>>>>> when
>>>>>> an issue is fixed by a contributor who hasn't commit access.
>>>>>> IMHO,
>>>>>> this implies that the reporter (or another contributor) provided
>>>>>> a
>>>>>> patch or non-trivial insights that led to the fix.
>>>>>>
>>>>>> IOW, when a developer with commit access fixes a bug or
>>>>>> implements a
>>>>>> feature request mostly by himself, the name of the original
>>>>>> reporter
>>>>>> should not appear in the release notes, as if he were the
>>>>>> contributor.
>>>>>
>>>>>
>>>>> The person who raised the bug still took the trouble to do so.
>>>>
>>>>
>>>> My question is still: is it sufficient?
>>>> Without filing a bug report, the reporter is harming himself.
>>>
>>> Not necessarily. I've certainly reported bugs that don't affect me.
>>>
>>>> Also, some reports are only feature requests. I deem it quite
>>>> unfair
>>>> that
>>>> the release notes would contain lines such as
>>>> * MATH-123456789: Algorithm Xxx implemented. Thanks to
>>>> <reporter>.
>>>
>>> So?
>>
>> It's just false.
>>
>>> They still made the effort.
>>> Maybe they did not provide a patch (yet) because it was not clear
>>> whether it would be accepted or not.
>>> And then the issue got fixed before they had a chance to provide
>>> the
>>> patch.
>>
>> Do we need to find corner cases just to not address the broader
>> issue?
>>
>>>
>>> The JIRA issue itself provides very little clue as to how much
>>> effort
>>> the person has expended.
>>
>> Really?
>> Attached patches are certainly more telling than a "Thanks to"
>> line...
>>
>>> But regardless, does it matter? It's probably quite a big step for
>>> some people to file the JIRA.
>>
>> Did I deny that?
>> How is it related to what I asked?
>>
>>>
>>> But one can of course add comments to the due-to text which can be
>>> used to clarify the perceived level of contribution.
>>
>> That's what I'm suggesting: a fair report.
>> So, is it possible to specify "Reported by" and "Fixed by"? That
>> would be
>> quite fine.
>>
>> I repeat: up until recently I never noticed undue (IMHO)
>> attributions.
>> That would mean that the practice was _not_ as you seem to describe.
>> Why would my question be met with dismissing comments?
>> There is a field in "changes.xml": how and when do we fill it?
>> I thought I knew (by looking at what others did) and now I see that
>> it
>> does not fit in some cases. Thus I ask for clarification.
>>
>> Why does this have to be controversial?
>
> I do not read any controversial comments here.
* Purporting allusions about me not being generous enough to thank
someone who has reported an issue. [This is false.]
* Defusing the original request by alleging that it does not matter.
[Even as I mentioned contradicting behaviours.]
Non-controversial would have been: I do <this thing> in <this
circumstance> and <that thing> in <that circumstance>.
> You started a thread with
> a specific question, and people answered with their opinion about
> this
> topic.
No, they inferred things about my supposedly evil reasons for raising
this question.
Gilles
> Here is mine:
>
> I did not make a difference between a simple reporter and a
> contributor
> that also provides a patch. In case the patch is provided by a
> different
> person than the original reporter, I added both.
>
> Actually I think that a person who raised the interest on a specific
> feature / topics, which then really gets included in the software is
> also worthwhile to mention as it really helps improve our software.
> E.g.
> most of the things I did were related to feature requests by other
> people and I am quite happy about that, because I know that somebody
> really wants to use this feature.
>
> If there should be two separate fields (reporter, patch-provider) or
> just one, hmmm it depends on personal preference imho. I am fine with
> just one field, but would not have a big problem if we have separate
> ones.
>
> Thomas
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@commons.apache.org
For additional commands, e-mail: dev-help@commons.apache.org
Re: [Math] "due-to" attribute in "changes.xml"
Posted by Thomas Neidhart <th...@gmail.com>.
On 10/31/2013 05:18 PM, Gilles wrote:
> On Thu, 31 Oct 2013 15:53:29 +0000, sebb wrote:
>> On 31 October 2013 14:24, Gilles <gi...@harfang.homelinux.org> wrote:
>>> On Thu, 31 Oct 2013 13:37:02 +0000, sebb wrote:
>>>>
>>>> On 31 October 2013 13:26, Gilles <gi...@harfang.homelinux.org> wrote:
>>>>>
>>>>> Hello.
>>>>>
>>>>> Are there criteria about filling the "due-to" attribute of an issue
>>>>> record in the "changes.xml" file?
>>>>>
>>>>> Current practice seems that reporting an issue does not by itself
>>>>> warrants such an attribution.
>>>>> Indeed, as I understand it, the attribute is a place-holder for when
>>>>> an issue is fixed by a contributor who hasn't commit access. IMHO,
>>>>> this implies that the reporter (or another contributor) provided a
>>>>> patch or non-trivial insights that led to the fix.
>>>>>
>>>>> IOW, when a developer with commit access fixes a bug or implements a
>>>>> feature request mostly by himself, the name of the original reporter
>>>>> should not appear in the release notes, as if he were the contributor.
>>>>
>>>>
>>>> The person who raised the bug still took the trouble to do so.
>>>
>>>
>>> My question is still: is it sufficient?
>>> Without filing a bug report, the reporter is harming himself.
>>
>> Not necessarily. I've certainly reported bugs that don't affect me.
>>
>>> Also, some reports are only feature requests. I deem it quite unfair
>>> that
>>> the release notes would contain lines such as
>>> * MATH-123456789: Algorithm Xxx implemented. Thanks to <reporter>.
>>
>> So?
>
> It's just false.
>
>> They still made the effort.
>> Maybe they did not provide a patch (yet) because it was not clear
>> whether it would be accepted or not.
>> And then the issue got fixed before they had a chance to provide the
>> patch.
>
> Do we need to find corner cases just to not address the broader issue?
>
>>
>> The JIRA issue itself provides very little clue as to how much effort
>> the person has expended.
>
> Really?
> Attached patches are certainly more telling than a "Thanks to" line...
>
>> But regardless, does it matter? It's probably quite a big step for
>> some people to file the JIRA.
>
> Did I deny that?
> How is it related to what I asked?
>
>>
>> But one can of course add comments to the due-to text which can be
>> used to clarify the perceived level of contribution.
>
> That's what I'm suggesting: a fair report.
> So, is it possible to specify "Reported by" and "Fixed by"? That would be
> quite fine.
>
> I repeat: up until recently I never noticed undue (IMHO) attributions.
> That would mean that the practice was _not_ as you seem to describe.
> Why would my question be met with dismissing comments?
> There is a field in "changes.xml": how and when do we fill it?
> I thought I knew (by looking at what others did) and now I see that it
> does not fit in some cases. Thus I ask for clarification.
>
> Why does this have to be controversial?
I do not read any controversial comments here. You started a thread with
a specific question, and people answered with their opinion about this
topic. Here is mine:
I did not make a difference between a simple reporter and a contributor
that also provides a patch. In case the patch is provided by a different
person than the original reporter, I added both.
Actually I think that a person who raised the interest on a specific
feature / topics, which then really gets included in the software is
also worthwhile to mention as it really helps improve our software. E.g.
most of the things I did were related to feature requests by other
people and I am quite happy about that, because I know that somebody
really wants to use this feature.
If there should be two separate fields (reporter, patch-provider) or
just one, hmmm it depends on personal preference imho. I am fine with
just one field, but would not have a big problem if we have separate ones.
Thomas
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@commons.apache.org
For additional commands, e-mail: dev-help@commons.apache.org
Re: [Math] "due-to" attribute in "changes.xml"
Posted by Gilles <gi...@harfang.homelinux.org>.
On Thu, 31 Oct 2013 15:53:29 +0000, sebb wrote:
> On 31 October 2013 14:24, Gilles <gi...@harfang.homelinux.org>
> wrote:
>> On Thu, 31 Oct 2013 13:37:02 +0000, sebb wrote:
>>>
>>> On 31 October 2013 13:26, Gilles <gi...@harfang.homelinux.org>
>>> wrote:
>>>>
>>>> Hello.
>>>>
>>>> Are there criteria about filling the "due-to" attribute of an
>>>> issue
>>>> record in the "changes.xml" file?
>>>>
>>>> Current practice seems that reporting an issue does not by itself
>>>> warrants such an attribution.
>>>> Indeed, as I understand it, the attribute is a place-holder for
>>>> when
>>>> an issue is fixed by a contributor who hasn't commit access. IMHO,
>>>> this implies that the reporter (or another contributor) provided a
>>>> patch or non-trivial insights that led to the fix.
>>>>
>>>> IOW, when a developer with commit access fixes a bug or implements
>>>> a
>>>> feature request mostly by himself, the name of the original
>>>> reporter
>>>> should not appear in the release notes, as if he were the
>>>> contributor.
>>>
>>>
>>> The person who raised the bug still took the trouble to do so.
>>
>>
>> My question is still: is it sufficient?
>> Without filing a bug report, the reporter is harming himself.
>
> Not necessarily. I've certainly reported bugs that don't affect me.
>
>> Also, some reports are only feature requests. I deem it quite unfair
>> that
>> the release notes would contain lines such as
>> * MATH-123456789: Algorithm Xxx implemented. Thanks to <reporter>.
>
> So?
It's just false.
> They still made the effort.
> Maybe they did not provide a patch (yet) because it was not clear
> whether it would be accepted or not.
> And then the issue got fixed before they had a chance to provide the
> patch.
Do we need to find corner cases just to not address the broader issue?
>
> The JIRA issue itself provides very little clue as to how much effort
> the person has expended.
Really?
Attached patches are certainly more telling than a "Thanks to" line...
> But regardless, does it matter? It's probably quite a big step for
> some people to file the JIRA.
Did I deny that?
How is it related to what I asked?
>
> But one can of course add comments to the due-to text which can be
> used to clarify the perceived level of contribution.
That's what I'm suggesting: a fair report.
So, is it possible to specify "Reported by" and "Fixed by"? That would
be
quite fine.
I repeat: up until recently I never noticed undue (IMHO) attributions.
That would mean that the practice was _not_ as you seem to describe.
Why would my question be met with dismissing comments?
There is a field in "changes.xml": how and when do we fill it?
I thought I knew (by looking at what others did) and now I see that it
does not fit in some cases. Thus I ask for clarification.
Why does this have to be controversial?
Gilles
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@commons.apache.org
For additional commands, e-mail: dev-help@commons.apache.org
Re: [Math] "due-to" attribute in "changes.xml"
Posted by sebb <se...@gmail.com>.
On 31 October 2013 14:24, Gilles <gi...@harfang.homelinux.org> wrote:
> On Thu, 31 Oct 2013 13:37:02 +0000, sebb wrote:
>>
>> On 31 October 2013 13:26, Gilles <gi...@harfang.homelinux.org> wrote:
>>>
>>> Hello.
>>>
>>> Are there criteria about filling the "due-to" attribute of an issue
>>> record in the "changes.xml" file?
>>>
>>> Current practice seems that reporting an issue does not by itself
>>> warrants such an attribution.
>>> Indeed, as I understand it, the attribute is a place-holder for when
>>> an issue is fixed by a contributor who hasn't commit access. IMHO,
>>> this implies that the reporter (or another contributor) provided a
>>> patch or non-trivial insights that led to the fix.
>>>
>>> IOW, when a developer with commit access fixes a bug or implements a
>>> feature request mostly by himself, the name of the original reporter
>>> should not appear in the release notes, as if he were the contributor.
>>
>>
>> The person who raised the bug still took the trouble to do so.
>
>
> My question is still: is it sufficient?
> Without filing a bug report, the reporter is harming himself.
Not necessarily. I've certainly reported bugs that don't affect me.
> Also, some reports are only feature requests. I deem it quite unfair that
> the release notes would contain lines such as
> * MATH-123456789: Algorithm Xxx implemented. Thanks to <reporter>.
So?
They still made the effort.
Maybe they did not provide a patch (yet) because it was not clear
whether it would be accepted or not.
And then the issue got fixed before they had a chance to provide the patch.
The JIRA issue itself provides very little clue as to how much effort
the person has expended.
But regardless, does it matter? It's probably quite a big step for
some people to file the JIRA.
But one can of course add comments to the due-to text which can be
used to clarify the perceived level of contribution.
>
>> Without the report, would the bug have been noticed and fixed as quickly?
>
>
> With the report, but without patch, would the bug have been fixed at all?
> [That's interesting: If the report is as important as the fix, then
> shouldn't
> all reports (even unfixed issues) be part of the release notes?]
>
>
>> IOW, the bug fix is still due-to the reporter, even if the
>> contribution is just the bug report.
>
>
> [Then, all these years, _many_ attributions were not acknowledged in
> this way.]
>
> I understand the argument that reporters are important in the development
> chain. But my point is that it is unfair to have people who notice a bug
> and people who analyze and fix that bug on the same footing.
> It makes sense to have a name in the release notes only if the person
> contributed to the implementation; unless I'm mistaken, it is the usual
> meaning of such references in such a file.
> Doing otherwise will just make this information meaningless and useless.
>
> Detailed role attribution can be retrieved from the bug-tracking system:
> when one reports a bug, he is referred to as the "reporter" there.
>
>
>
> Regards,
> Gilles
>
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: dev-unsubscribe@commons.apache.org
> For additional commands, e-mail: dev-help@commons.apache.org
>
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@commons.apache.org
For additional commands, e-mail: dev-help@commons.apache.org
Re: [Math] "due-to" attribute in "changes.xml"
Posted by Gilles <gi...@harfang.homelinux.org>.
On Thu, 31 Oct 2013 11:30:16 -0400, Ted Dunning wrote:
> On Thu, Oct 31, 2013 at 10:24 AM, Gilles
> <gi...@harfang.homelinux.org>wrote:
>
>>
>>> The person who raised the bug still took the trouble to do so.
>>>
>>
>> My question is still: is it sufficient?
>> Without filing a bug report, the reporter is harming himself.
>>
>> Also, some reports are only feature requests. I deem it quite unfair
>> that
>> the release notes would contain lines such as
>> * MATH-123456789: Algorithm Xxx implemented. Thanks to <reporter>.
>
>
> How is it controversial to say thank you for contributions? The
> report is
> a contribution and being nice could encourage more contributions.
How is it that you, more often than not, attempt to reverse the meaning
of
what is being said?
Could you please report instances where I would not have said "Thank
you"
to a contributor?
> Being all officious about what suffices to be worthy enough to make
> the oh
> so mighty gatekeepers be generous is a great way to turn people off.
Here no one should need to be generous or not. The release notes are an
unfair report of the contributions if some names appear there while
others
do not. It's a fact, no more no less.
Who gives and who receives?
If _you_ do not want to be fair, then just say it; do not try to imply
anything about me. Please.
Gilles
P.S. I've seen people here that produced reports from SVN, from JIRA,
etc.
If we want to reward every contribution, we can do so in reports
that
will properly attribute everything. The release notes are not the
place
for that.
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@commons.apache.org
For additional commands, e-mail: dev-help@commons.apache.org
Re: [Math] "due-to" attribute in "changes.xml"
Posted by Ted Dunning <te...@gmail.com>.
On Thu, Oct 31, 2013 at 10:24 AM, Gilles <gi...@harfang.homelinux.org>wrote:
>
>> The person who raised the bug still took the trouble to do so.
>>
>
> My question is still: is it sufficient?
> Without filing a bug report, the reporter is harming himself.
>
> Also, some reports are only feature requests. I deem it quite unfair that
> the release notes would contain lines such as
> * MATH-123456789: Algorithm Xxx implemented. Thanks to <reporter>.
How is it controversial to say thank you for contributions? The report is
a contribution and being nice could encourage more contributions.
Being all officious about what suffices to be worthy enough to make the oh
so mighty gatekeepers be generous is a great way to turn people off.
Re: [Math] "due-to" attribute in "changes.xml"
Posted by Gilles <gi...@harfang.homelinux.org>.
On Thu, 31 Oct 2013 13:37:02 +0000, sebb wrote:
> On 31 October 2013 13:26, Gilles <gi...@harfang.homelinux.org>
> wrote:
>> Hello.
>>
>> Are there criteria about filling the "due-to" attribute of an issue
>> record in the "changes.xml" file?
>>
>> Current practice seems that reporting an issue does not by itself
>> warrants such an attribution.
>> Indeed, as I understand it, the attribute is a place-holder for when
>> an issue is fixed by a contributor who hasn't commit access. IMHO,
>> this implies that the reporter (or another contributor) provided a
>> patch or non-trivial insights that led to the fix.
>>
>> IOW, when a developer with commit access fixes a bug or implements a
>> feature request mostly by himself, the name of the original reporter
>> should not appear in the release notes, as if he were the
>> contributor.
>
> The person who raised the bug still took the trouble to do so.
My question is still: is it sufficient?
Without filing a bug report, the reporter is harming himself.
Also, some reports are only feature requests. I deem it quite unfair
that
the release notes would contain lines such as
* MATH-123456789: Algorithm Xxx implemented. Thanks to <reporter>.
> Without the report, would the bug have been noticed and fixed as
> quickly?
With the report, but without patch, would the bug have been fixed at
all?
[That's interesting: If the report is as important as the fix, then
shouldn't
all reports (even unfixed issues) be part of the release notes?]
> IOW, the bug fix is still due-to the reporter, even if the
> contribution is just the bug report.
[Then, all these years, _many_ attributions were not acknowledged in
this way.]
I understand the argument that reporters are important in the
development
chain. But my point is that it is unfair to have people who notice a
bug
and people who analyze and fix that bug on the same footing.
It makes sense to have a name in the release notes only if the person
contributed to the implementation; unless I'm mistaken, it is the usual
meaning of such references in such a file.
Doing otherwise will just make this information meaningless and
useless.
Detailed role attribution can be retrieved from the bug-tracking
system:
when one reports a bug, he is referred to as the "reporter" there.
Regards,
Gilles
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@commons.apache.org
For additional commands, e-mail: dev-help@commons.apache.org
Re: [Math] "due-to" attribute in "changes.xml"
Posted by sebb <se...@gmail.com>.
On 31 October 2013 13:26, Gilles <gi...@harfang.homelinux.org> wrote:
> Hello.
>
> Are there criteria about filling the "due-to" attribute of an issue
> record in the "changes.xml" file?
>
> Current practice seems that reporting an issue does not by itself
> warrants such an attribution.
> Indeed, as I understand it, the attribute is a place-holder for when
> an issue is fixed by a contributor who hasn't commit access. IMHO,
> this implies that the reporter (or another contributor) provided a
> patch or non-trivial insights that led to the fix.
>
> IOW, when a developer with commit access fixes a bug or implements a
> feature request mostly by himself, the name of the original reporter
> should not appear in the release notes, as if he were the contributor.
The person who raised the bug still took the trouble to do so.
Without the report, would the bug have been noticed and fixed as quickly?
IOW, the bug fix is still due-to the reporter, even if the
contribution is just the bug report.
>
> Regards,
> Gilles
>
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: dev-unsubscribe@commons.apache.org
> For additional commands, e-mail: dev-help@commons.apache.org
>
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@commons.apache.org
For additional commands, e-mail: dev-help@commons.apache.org