The concrete incidence is https://bugs.debian.org/972695#28That's an automatically triggered message created by a gitlab hook,
Am I missing something sensible here, or do others also see a problem
in this setup?
El 31/5/25 a las 18:19, Jonas Smedegaard escribió:
The concrete incidence is https://bugs.debian.org/972695#28
Am I missing something sensible here, or do others also see a problemThat's an automatically triggered message created by a gitlab hook,
in this setup?
which parses the commit message for "closes: #nnnnn" and notifies the bug submitter when it's known that the bug is fixed in salsa.
That's an automatically triggered message created by a gitlab hook,
which parses the commit message for "closes: #nnnnn" and notifies the bug
submitter when it's known that the bug is fixed in salsa.
The problem is that the confident submitter is a bot.
In the concrete case, I replied to point out that the bug closure was a mistake. That reply bounced.
Is it wrong of me to cc the "person" interacting with a bugreport?
Is it wrong of me to expect being able to reach that "person"? Easily?
I don't see a big problem in those messages being "noreply",
considering that they are generated automatically,
Arguably, such setup is spam: I is a bot that messes with the bugs but
is not accountable for its actions, since it is only a one-way
communication. Sure, I can then investigate the email and figure out
which non-email side channel might reach the true originator of the
bot activity, but I have no interest in that added burden laid on me.
The concrete incidence is https://bugs.debian.org/972695#28
Am I missing something sensible here, or do others also see a problem
in this setup?
Your reply had:
To: 972695-quiet@bugs.debian.org,
972695-submitter@bugs.debian.org, Yadd <noreply@salsa.debian.org>
I am wondering did you add this 972695-quiet@bugs.debian.org manually
or was it in the Reply-To headers? In this case it would have been
best for all participants to get your update.
The problem is that the confident submitter is a bot.
In the concrete case, I replied to point out that the bug closure was a mistake. That reply bounced.
Is it wrong of me to cc the "person" interacting with a bugreport?
Is it wrong of me to expect being able to reach that "person"? Easily?
Hi Jonas,
Le 2025-05-31 21:41, Jonas Smedegaard a écrit :
The problem is that the confident submitter is a bot.
In the concrete case, I replied to point out that the bug closure was a mistake. That reply bounced.
Is it wrong of me to cc the "person" interacting with a bugreport?
Is it wrong of me to expect being able to reach that "person"? Easily?
I suppose that in this specific case you wanted to interact with the
author of the commit, not with the bot or the author of the bot or its admin.
It is not wrong, however with the current state of anti-spam measures
having bots that send mail with a "From: " address that could belong to
any foreign domain is not a good practice. So the sender address will probably have to remain as it is, as arguably a bounce is a better
feedback here than silently accepting a message that will be ignored.
But the "Reply-To: " field of the bot message could certainly be
populated with the names and addresse(s) of the author and/or committer
(and NNN@b.d.o). Would that work for your case?
Quoting Julien Plissonneau Duquène (2025-06-01 12:05:43)
Hi Jonas,
Le 2025-05-31 21:41, Jonas Smedegaard a écrit :
The problem is that the confident submitter is a bot.
In the concrete case, I replied to point out that the bug closure was a mistake. That reply bounced.
Is it wrong of me to cc the "person" interacting with a bugreport?
Is it wrong of me to expect being able to reach that "person"? Easily?
I suppose that in this specific case you wanted to interact with the author of the commit, not with the bot or the author of the bot or its admin.
It is not wrong, however with the current state of anti-spam measures having bots that send mail with a "From: " address that could belong to any foreign domain is not a good practice. So the sender address will probably have to remain as it is, as arguably a bounce is a better feedback here than silently accepting a message that will be ignored.
But the "Reply-To: " field of the bot message could certainly be
populated with the names and addresse(s) of the author and/or committer (and NNN@b.d.o). Would that work for your case?
Not entirely sure, but yes, that sounds like a sensible solution to me.
Sysop: | Keyop |
---|---|
Location: | Huddersfield, West Yorkshire, UK |
Users: | 486 |
Nodes: | 16 (2 / 14) |
Uptime: | 136:45:37 |
Calls: | 9,657 |
Calls today: | 5 |
Files: | 13,707 |
Messages: | 6,167,060 |