On Sun, Mar 06, 2005 at 01:10:27PM -0500, Clint Adams wrote:
> Assume the demarcated hypothetical scenario to be true for the questions
> which follow.
Okay. I'll break your allegory up by sentence.
> ===BEGIN ALLEGORY===
> I spend the next several hours sending Andrew Suffield chain letters
> that say "if you don't fix tla by Thursday, your liver will transform
> into mascarpone."
Not being a gourmet, I had not known what mascarpone is (it appears ispell
doesn't, either). Thanks for broadening my horizons. :)
> I send these emails without making use of Debian infrastructure, so as to
> not violate the DMUP.
Agreed. The DMUP's scope, as I understand it, is strictly limited to
project resources.
> Andrew buries his face in a pillow and cries until Saturday, at which
> point he realizes that his liver is normal.
Who among us has not shed a tear over our treatment at the hands of our
fellow developer? :)
--LINE OF DEMARCATION---
> Fuming with rage, he then declares to his favorite stuffed animal that,
> henceforth, he will ignore any bug I file.
This seems an overreaction, but even if necessary for the maintenance of
"Andrew's" sanity, he can establish workarounds for this.
* He could attempt to call you out (perhaps on -private) for your bad
behavior, shaming you before your peers. His ego may be sensitive,
however, and/or he may not wish to discuss this publicly.
or:
* He could make it clear to you that he will ignore any bug reports from
you that include irrelevancies such as mocking him for his mascarpone
gullibility; and deal only with bug reports from you that exhibit
courtesy and respect.
or:
* He could put out a call for a co-maintainer to help him out with the
package. This assistance may or may not be limited in scope to dealing
with bug reports from you. (In the ego-sensitive variant of this
scenario, this restriction would have to be communicated privately to the
co-maintainer, with or without justification depending on how amenable that
person is.) Since the Debian BTS is publicly accessible, this
co-maintainer need not even be affiliated with the Debian Project. I
could have an army of miserable gnome slaves in my house who help me
triage bug reports against Debian's XFree86 packages, for example[2].
> Furthermore, he will ignore any bug that anyone else files on tla, and
> will refuse to do any work on tla.
This is getting beyond the bounds of rational response, IMO.
> He does not announce this publicly, but his stuffed animal leaks this
> information to a select few other developers, and the rumor makes its way
> to the DPL.
Orphaning a package without making an announcement that one is doing so is
at the very least a discourtesy to the rest of the project. One can hardly
claim ignorance of the expected procedure in these cases, given that
they're documented in the Developers' Reference[1].
> When others offer to NMU tla, he informs them that it is his package and
> they must keep their grubby little paws away from it.
> ===END ALLEGORY===
Ultimately, the package "belongs" to the Debian Project collectively. We
grant package maintainers wide latitude to maintain a package as they see
fit, but they do actually have to maintain it.
From The Collaborative International Dictionary of English v.0.48 [gcide]:
1. To hold or keep in any particular state or condition; to
support; to sustain; to uphold; to keep up; not to suffer
to fail or decline; as, to maintain a certain degree of
heat in a furnace; to maintain a fence or a railroad; to
maintain the digestive process or powers of the stomach;
to maintain the fertility of soil; to maintain present
reputation.
If a package's bugs are not being dealt with -- if the package is not being
supported, sustained, or kept from failure or decline, then it is not being
maintained. The Maintainer: field of a package can only reflect reality;
it cannot *command* reality.
I daresay we're well within the realm of common sense here. That you raise
this allegory suggests that your hypothetical scenario is not sensible, or
in the event it is not actually hypothetical (except perhaps with respect to
the chain-letter/mascarpone/liver detail), then someone is not behaving
sensibly.
> Is any of this Debian's business?
Yes.
> If so, which parts?
The parts below the --LINE OF DEMARCATION-- I indicated above.
> What is the role of the DPL in such a situation?
As far as I know we don't really have a body tasked with dealing with this
sort of situation. It therefore falls to the DPL to hear the grievance,
determine whether action should be taken, and if so, execute that action.
> How would the answers to the previous three questions be different if
> either Andrew or I were a DPL delegate or an NM applicant?
In substance, they wouldn't, but:
* If either party were a DPL delegate and I were the DPL, I'd want to know
why my delegate was reflecting so poorly on the leadership (by knowingly
sending a deceptive and hurtful chain-letter to a fellow Debian
developer, or by refusing to formally orphan a package that one has
abandoned by all measurable standards, as applicable). I'd contact them
privately (at first) and ask them what was/is going on.
* If the package maintainer were an NM applicant and I were the DPL, I'd
contact his or her AM and see if I could get some more information, and
solicit the AM's opinion about the situation. Since NM's can't quite
completely maintain a package (given that they can't upload a signed
package that will be accepted into the archive), I'd also want to know if
the package's sponsor knew anything about the situation, and I would
contact them.
* If the mascarpone-mail-sender were an NM applicant and I had credible
evidence of the harassment, I would inform his or her AM, and invite the
AM to take this harassing behavior into account when weighing the
decision to recommend the applicant to the NM Committee.
> Should there be rules codified to address any of the actions in the
> aforementioned scenario? Should there be rules codified to address any
> interpersonal behavior not mentioned in said scenario?
My response above could easily be the germ of a general non-technical
grievance resolution process. As such, I welcome comment on it.
[1] If you can no longer maintain a package, you need to inform the others
about that, and see that the package is marked as orphaned. You should
set the package maintainer to Debian QA Group <packages@qa.debian.org>
and submit a bug report against the pseudo package wnpp. [...]
-- http://www.debian.org/doc/developers-reference/ch-pkgs.en.html#s-orphaning
[2] If I did, though, I'd have to conclude I'm not a very effective
drive of gnome slaves. In actual fact, I can use more volunteer help
handling xfree86 bug reports. :)
--
G. Branden Robinson | I just wanted to see what it looked
Debian GNU/Linux | like in a spotlight.
branden@debian.org | -- Jim Morrison
http://people.debian.org/~branden/ |
Attachment:
signature.asc
Description: Digital signature