"BreamoreBoy" is back on tracker touching hundreds of issues without adding any new information. This is certainly not the first time. Can we ban him?
On 05.10.14 21:15, Benjamin Peterson wrote:
"BreamoreBoy" is back on tracker touching hundreds of issues without adding any new information. This is certainly not the first time. Can we ban him?
I think it would be enough just to say him stop. He is not malicious.
Actually there was few cases when his reminders was helpful.
On Sun, Oct 5, 2014, at 14:42, Serhiy Storchaka wrote:
On 05.10.14 21:15, Benjamin Peterson wrote:
"BreamoreBoy" is back on tracker touching hundreds of issues without adding any new information. This is certainly not the first time. Can we ban him?
I think it would be enough just to say him stop. He is not malicious.
I did. http://bugs.python.org/issue18372#msg228613
Actually there was few cases when his reminders was helpful.
That's true, though, I'm not sure noise is worth the few issues that are closed.
On 10/5/2014 2:15 PM, Benjamin Peterson wrote:
"BreamoreBoy" is back on tracker touching hundreds of issues without adding any new information. This is certainly not the first time. Can we ban him?
Can we? Of course, but that is a rare last resort. Should we? No.
I have closed several issues and responded to other because of his pings.
On core-mentorship list, people are told that after a few months with no response, particularly after posting a patch, it is ok to ping. I suspect Mark thinks he is acting in accordance with this.
I have not seen Mark do now what lead to suspension of tracker admin privileges a few years ago -- which was to demand that someone respond or see the issue closed, which would then do. It was Mark's choice then to leave the tracker completely (contrary to Guido's wishes). If he has done the above on issues not involving me, then he should be told to stop.
It appears that someone somewhat recently gave Mark at least partial admin privileges (to change Versions). If so, that would have been a message that he was doing ok after his return. If he is not doing ok, let us give him more guidance.
If you are annoyed, lets discuss specifics and see if we can agree on refined guidelines to give him. It could even be "leave Benjamin's issues alone".
tjr
On Sun, 05 Oct 2014 14:47:06 -0400, Benjamin Peterson <benjamin@python.org> wrote:
On Sun, Oct 5, 2014, at 14:42, Serhiy Storchaka wrote:
On 05.10.14 21:15, Benjamin Peterson wrote:
"BreamoreBoy" is back on tracker touching hundreds of issues without adding any new information. This is certainly not the first time. Can we ban him?
I think it would be enough just to say him stop. He is not malicious.
I did. http://bugs.python.org/issue18372#msg228613
Actually there was few cases when his reminders was helpful.
That's true, though, I'm not sure noise is worth the few issues that are closed.
It is certainly true that I for one ignore anything with his name on it, because most of the time it is noise and it isn't worth the effort to figure out which ones aren't noise.
--David
On Oct 5, 2014, at 12:24 , R. David Murray <rdmurray@bitdance.com> wrote:
It is certainly true that I for one ignore anything with his name on it, because most of the time it is noise and it isn't worth the effort to figure out which ones aren't noise.
To me, the main issue is that the noise is not just directed at python committers but also to the python users who have submitted those issues or otherwise following them (via nosy or otherwise). I think the risk is that his noise sends a wrong message to those users: i.e. that python-dev has suddenly taken an interest in this issue and that, by taking the time to create a patch, the issue will somehow get magically resolved. That won't happen, of course, unless a core developer chooses to get involved.
The point of having the issue tracker is to solve problems, not to have a kind of contest about how many issues can be closed. Yes, all things being equal, it is better to have fewer open issues but that's not the primary goal. And I am uncomfortable with the risk of users potentially inferring that he is somehow a de-facto "project leader" of Python maintenance.
-- Ned Deily nad@acm.org -- []
On 10/5/2014 2:47 PM, Benjamin Peterson wrote:
On Sun, Oct 5, 2014, at 14:42, Serhiy Storchaka wrote:
On 05.10.14 21:15, Benjamin Peterson wrote:
"BreamoreBoy" is back on tracker touching hundreds of issues without adding any new information. This is certainly not the first time. Can we ban him?
I think it would be enough just to say him stop. He is not malicious.
On this one, he was a bit 'bossy'. Not quite as extreme as a few years ago, but definitely in the wrong direction. I responded on the issue, and will say more privately.
Actually there was few cases when his reminders was helpful.
That's true, though, I'm not sure noise is worth the few issues that are closed.
I guess I have been luckier ;-)
tjr
On Oct 5, 2014, at 12:24 , R. David Murray <rdmurray@bitdance.com> wrote:
It is certainly true that I for one ignore anything with his name on it, because most of the time it is noise and it isn't worth the effort to figure out which ones aren't noise.
To me, the main issue is that the noise is not just directed at python committers but also to the python users who have submitted those issues or otherwise following them (via nosy or otherwise). I think the risk is that his noise sends a wrong message to those users: i.e. that python-dev has suddenly taken an interest in this issue and that, by taking the time to create a patch, the issue will somehow get magically resolved. That won't happen, of course, unless a core developer chooses to get involved.
The point of having the issue tracker is to solve problems, not to have a kind of contest about how many issues can be closed. Yes, all
Le 05/10/2014 21:36, Ned Deily a écrit : things being equal, it is better to have fewer open issues but that's not the primary goal.
I agree with Ned. Closing idle issues is nice, but it's hardly a benefit to Python's quality.
Regards
Antoine.
On 10/05/2014 09:36 PM, Ned Deily wrote:
On Oct 5, 2014, at 12:24 , R. David Murray <rdmurray@bitdance.com> wrote:
It is certainly true that I for one ignore anything with his name on it, because most of the time it is noise and it isn't worth the effort to figure out which ones aren't noise.
To me, the main issue is that the noise is not just directed at python committers but also to the python users who have submitted those issues or otherwise following them (via nosy or otherwise). I think the risk is that his noise sends a wrong message to those users: i.e. that python-dev has suddenly taken an interest in this issue and that, by taking the time to create a patch, the issue will somehow get magically resolved. That won't happen, of course, unless a core developer chooses to get involved.
Most of the messages like "can someone look at this" don't seem to send any wrong messages. However, I agree that some of them are a bit pompous, like this one: http://bugs.python.org/issue1284316#msg228480 Responses to hints tend to sound offended: http://bugs.python.org/issue1284316#msg228483
The point of having the issue tracker is to solve problems, not to have a kind of contest about how many issues can be closed. Yes, all things being equal, it is better to have fewer open issues but that's not the primary goal.
The tracker does profit from having less inactive issues that are ready to be closed after a trivial commit, or being out of date or missing requested feedback. I've closed a few such in the last few days because of Mark's pings.
Remember that one thing we'd like users to see before reporting is to search the tracker for similar issues: the less noise they find there the better.
And I am uncomfortable with the risk of users potentially inferring that he is somehow a de-facto "project leader" of Python maintenance.
Well, those users can easily be informed about the circumstances should a question arise.
In total, I think there's no grounds for a ban (yet), but his tone has to be watched. If hints from our side are con ignored or receive ad-hominem responses, that'll change the situation in my opinion.
cheers, Georg
Le 05/10/2014 21:02, Terry Reedy a écrit :
If you are annoyed, lets discuss specifics and see if we can agree on refined guidelines to give him. It could even be "leave Benjamin's issues alone".
I'd like him to stop blindly pinging issues without trying to investigate them. I routinely get several e-mail notifications per day just because of his messages. It's highly annoying.
Regards
Antoine.
In article <m0sah7$4ci$1@ger.gmane.org>, Georg Brandl <g.brandl@gmx.net> wrote:
On 10/05/2014 09:36 PM, Ned Deily wrote:
To me, the main issue is that the noise is not just directed at python committers but also to the python users who have submitted those issues or otherwise following them (via nosy or otherwise). I think the risk is that his noise sends a wrong message to those users: i.e. that python-dev has suddenly taken an interest in this issue and that, by taking the time to create a patch, the issue will somehow get magically resolved. That won't happen, of course, unless a core developer chooses to get involved. Most of the messages like "can someone look at this" don't seem to send any wrong messages.
I was thinking more of the messages to non-python-dev users along the lines of "Can you supply a patch?" with an implied promise that this will cause the issue to be resolved, often without any particular insight into whether such a patch should be written.
The tracker does profit from having less inactive issues that are ready to be closed after a trivial commit, or being out of date or missing requested feedback. I've closed a few such in the last few days because of Mark's pings.
Remember that one thing we'd like users to see before reporting is to search the tracker for similar issues: the less noise they find there the better.
I don't disagree with that.
And I am uncomfortable with the risk of users potentially inferring that he is somehow a de-facto "project leader" of Python maintenance. Well, those users can easily be informed about the circumstances should a question arise.
How would we know? They are likely unfamiliar with the python-dev project and they receive these emails from an unknown person, sometimes even offering apologies on behalf of an indefinite "we".
In total, I think there's no grounds for a ban (yet), but his tone has to be watched. If hints from our side are con ignored or receive ad-hominem responses, that'll change the situation in my opinion.
I dunno. We've been down this road more than once over the years, always ending in some dust-up. I really don't think it's healthy for python-dev or our users to keep repeating that.
-- Ned Deily, nad@acm.org
On 10/5/2014 3:39 PM, Terry Reedy wrote:
I responded on the issue, and will say more privately.
I wrote him and tried to communicate three ideas.
- Closing issues and responding to new messages, while important, are subsidiary to the primary goal of improving Python.
- We would prefer more quality and less quantity.
- We each set our priorites; he should not try to do so for us.
Terry
On 10/05/2014 11:01 PM, Ned Deily wrote:
In article <m0sah7$4ci$1@ger.gmane.org>, Georg Brandl <g.brandl@gmx.net> wrote:
On 10/05/2014 09:36 PM, Ned Deily wrote:
To me, the main issue is that the noise is not just directed at python committers but also to the python users who have submitted those issues or otherwise following them (via nosy or otherwise). I think the risk is that his noise sends a wrong message to those users: i.e. that python-dev has suddenly taken an interest in this issue and that, by taking the time to create a patch, the issue will somehow get magically resolved. That won't happen, of course, unless a core developer chooses to get involved. Most of the messages like "can someone look at this" don't seem to send any wrong messages.
I was thinking more of the messages to non-python-dev users along the lines of "Can you supply a patch?" with an implied promise that this will cause the issue to be resolved, often without any particular insight into whether such a patch should be written.
Submitting patches is almost never a bad idea. But I agree the wholesale nature of the commenting without insight into the issue is a bit worrying.
Just like the initial submission, the submission of a patch generates an event with a certain probability of being noticed by "the right person" who'll take it further. I assume there are (established or aspiring) core developers searching explicitly for issues with patch when looking for potential work.
And I am uncomfortable with the risk of users potentially inferring that he is somehow a de-facto "project leader" of Python maintenance. Well, those users can easily be informed about the circumstances should a question arise.
How would we know? They are likely unfamiliar with the python-dev project and they receive these emails from an unknown person, sometimes even offering apologies on behalf of an indefinite "we".
You're right.
In total, I think there's no grounds for a ban (yet), but his tone has to be watched. If hints from our side are con ignored or receive ad-hominem responses, that'll change the situation in my opinion.
I dunno. We've been down this road more than once over the years, always ending in some dust-up. I really don't think it's healthy for python-dev or our users to keep repeating that.
Time for the COC overlords to chime in, I suppose.
cheers, Georg
Le 05/10/2014 23:22, Terry Reedy a écrit :
On 10/5/2014 3:39 PM, Terry Reedy wrote:
I responded on the issue, and will say more privately.
I wrote him and tried to communicate three ideas.
- Closing issues and responding to new messages, while important, are subsidiary to the primary goal of improving Python.
- We would prefer more quality and less quantity.
- We each set our priorites; he should not try to do so for us.
Thanks! I hope your mediation will change things for the better.
Regards
Antoine.
On Oct 5, 2014, at 1:44 PM, Antoine Pitrou <antoine@python.org> wrote:
Le 05/10/2014 21:02, Terry Reedy a écrit :
If you are annoyed, lets discuss specifics and see if we can agree on refined guidelines to give him. It could even be "leave Benjamin's issues alone".
I'd like him to stop blindly pinging issues without trying to investigate them. I routinely get several e-mail notifications per day just because of his messages. It's highly annoying.
FWIW, I'm also finding the behavior to be disruptive. It is interfering with using the tracker to know which issues are currently active.
Once in a while, it has a mildly beneficial effect of reminding us about something that has lain dormant for too long, but most of time the posts make the signal to noise ratio worse (mostly because Mark's posts usually have almost zero information content).
I'm not at all into banning, but it would be nice if Mark's interest could be channelled into something useful.
Raymond
On Sun, Oct 5, 2014 at 6:34 PM, Raymond Hettinger < raymond.hettinger@gmail.com> wrote:
I'm not at all into banning, but it would be nice if Mark's interest could be channelled into something useful.
Yes, given that he is reasonable on emails. So requesting him to not just prod/churn issues, but pick up issues to work or contribute via email discussions would be better.
Thanks, Senthil
On 6 October 2014 07:22, Terry Reedy <tjreedy@udel.edu> wrote:
On 10/5/2014 3:39 PM, Terry Reedy wrote:
I responded on the issue, and will say more privately.
I wrote him and tried to communicate three ideas.
- Closing issues and responding to new messages, while important, are subsidiary to the primary goal of improving Python.
- We would prefer more quality and less quantity.
- We each set our priorites; he should not try to do so for us.
Thanks Terry.
I know Mark is particularly concerned about the "open issues" tally and constantly seeks ways to bring that down. Scattergun pinging of open issues is one of the easiest, but also one of the ones with a high hidden cost in annoying people - for developers, it may mean getting dozens of email notifications, for issue submitters, it may lead to anticipation of action, only to be disappointed when it's just a ping asking for a status update.
Regards, Nick.
-- Nick Coghlan | ncoghlan@gmail.com | Brisbane, Australia
On 10/6/2014 1:02 AM, Nick Coghlan wrote:
On 6 October 2014 07:22, Terry Reedy <tjreedy@udel.edu> wrote:
On 10/5/2014 3:39 PM, Terry Reedy wrote:
I responded on the issue, and will say more privately.
I wrote him and tried to communicate three ideas.
- Closing issues and responding to new messages, while important, are subsidiary to the primary goal of improving Python.
- We would prefer more quality and less quantity.
- We each set our priorites; he should not try to do so for us.
Thanks Terry.
I know Mark is particularly concerned about the "open issues" tally and constantly seeks ways to bring that down.
I think I can suggest some better ways.
Scattergun pinging of open issues is one of the easiest, but also one of the ones with a high hidden cost in annoying people - for developers, it may mean getting dozens of email notifications, for issue submitters, it may lead to anticipation of action, only to be disappointed when it's just a ping asking for a status update.
Since all aspects of this annoyance were not completely obvious to me, I am sure they are not to him. With the discussion here, I understand much better and will try to explain to him.
Terry
On Oct 6, 2014 9:16 AM, "Terry Reedy" <tjreedy@udel.edu> wrote:
On 10/6/2014 1:02 AM, Nick Coghlan wrote:
On 6 October 2014 07:22, Terry Reedy <tjreedy@udel.edu> wrote:
On 10/5/2014 3:39 PM, Terry Reedy wrote:
I responded on the issue, and will say more privately.
I wrote him and tried to communicate three ideas.
- Closing issues and responding to new messages, while important, are subsidiary to the primary goal of improving Python.
- We would prefer more quality and less quantity.
- We each set our priorites; he should not try to do so for us.
Thanks Terry.
I know Mark is particularly concerned about the "open issues" tally and constantly seeks ways to bring that down.
I think I can suggest some better ways.
A while ago I talked with him about this issue, and suggested him to write down all the "interesting" issues on a list and mail it to python-dev once a week, instead of pinging the issues individually.
He followed my suggestion and sent the list, but iirc no one replied, so he probably deemed the approach ineffective and went back to pinging the issues.
I also suggested him to work on the bug tracker code or on similar projects -- not sure if he gave that a try yet.
Best Regards, Ezio Melotti
Scattergun pinging of open issues is one of the easiest, but also one of the ones with a high hidden cost in annoying people - for developers, it may mean getting dozens of email notifications, for issue submitters, it may lead to anticipation of action, only to be disappointed when it's just a ping asking for a status update.
Since all aspects of this annoyance were not completely obvious to me, I
am sure they are not to him. With the discussion here, I understand much better and will try to explain to him.
Terry
participants (11)
-
Antoine Pitrou -
Benjamin Peterson -
Ezio Melotti -
Georg Brandl -
Ned Deily -
Nick Coghlan -
R. David Murray -
Raymond Hettinger -
Senthil Kumaran -
Serhiy Storchaka -
Terry Reedy