1 |
(Warning: wall of text) |
2 |
|
3 |
On 11/07/2016 04:20 AM, Rich Freeman wrote: |
4 |
> On Mon, Nov 7, 2016 at 2:43 AM, Daniel Campbell <zlg@g.o> wrote: |
5 |
>> |
6 |
>> Gentoo is ultimately better off if it takes its stance of |
7 |
>> non-involvement more seriously and butts out of any disputes until |
8 |
>> they affect Gentoo itself, i.e. commit wars, spam or harassment in |
9 |
>> IRC/forums, etc. |
10 |
> |
11 |
> As far as I can tell that is basically the status quo. About the only |
12 |
> times I've seen Comrel doing anything is when it involves fairly |
13 |
> serious issues that fall into the sorts of categories you describe |
14 |
> here. |
15 |
> |
16 |
> Sure, the folks involved might go telling everybody they know that |
17 |
> they're being persecuted for whatever. Since Gentoo does not disclose |
18 |
> why people are asked to leave, there is really nothing to contradict |
19 |
> their side of the story. That is frustrating, but I don't think we |
20 |
> really have any reasonable alternatives. |
21 |
|
22 |
A group whose sole purpose is to mediate disputes and makes the business |
23 |
of others its business is, to me, the exact opposite of staying out of |
24 |
disputes and letting adults be adults. |
25 |
|
26 |
>> Why is comrel necessary? Do we not have policy documents outlining the |
27 |
>> consequences of violating the CoC? |
28 |
> |
29 |
> This is like saying why do we need a kernel project when we have |
30 |
> documents explaining how the kernel gets maintained. |
31 |
> |
32 |
> Comrel is the body that actually administers enforcement of the CoC. |
33 |
|
34 |
And how do they enforce that? Somebody has to have access to infra, |
35 |
forums, mailing lists, etc. Does comrel have its hands in that many |
36 |
cookie jars? If so, something is very wrong with that picture. The most |
37 |
powerful groups in Gentoo are given their power on the condition that they |
38 |
meet re-election and are accountable to what amounts to constituents |
39 |
(devs and/or foundation members). Trustees are voted on, Council is |
40 |
voted on. Ignoring the metastructure revisiting that we've seen in other |
41 |
threads, frankly comrel doesn't deserve that much power when there is |
42 |
already staff who can (and already does) do comrel's job. They interact |
43 |
directly with problem users/devs and "hold'er steady". Their job is |
44 |
mostly transparent and thankless, but they already handle it. |
45 |
|
46 |
Having people wearing both hats is also dangerous, as it creates a |
47 |
chilling effect on anyone on the business end of comrel. If 2 or 3 |
48 |
members of Comrel are also Councilors, one of two things happens: those |
49 |
people sit out of appeals (making the remaining council votes more |
50 |
powerful than usual) or they double down on their decision and create a |
51 |
stacked deck. Neither outcome is ideal or impartial. If you or someone |
52 |
else were on the business end of such decisions, would you take an |
53 |
honest shot at an appeal, knowing that a few of the same people who made |
54 |
the initial decision were overseeing its appeal? To me, that's a red |
55 |
flag for corruption, and some may consider that grounds to petition the |
56 |
Foundation for a recall. In my searching, I didn't see a time where that |
57 |
happened. I hope nothing comes to a head like that. |
58 |
|
59 |
>> Removing comrel can reduce legal liability in that there _is_ no |
60 |
>> investigative team involved, no slander or libel to worry about. |
61 |
> |
62 |
> When somebody goes doing "commit wars, spam or harassment in |
63 |
> IRC/forums, etc" how would you propose dealing with those problems |
64 |
> without having some sort of investigative team? |
65 |
|
66 |
That honestly depends on the incident. Disrupting discussion or support |
67 |
in an official Gentoo channel on IRC is often handled by ops with |
68 |
warnings, kicks, and if it's bad enough, temporary bans. Comrel's not |
69 |
even involved with that unless it chooses to be. |
70 |
|
71 |
Commit wars are trivial to mediate, and git does a great job of |
72 |
creating a "paper trail". We have policies to point to. We can enact |
73 |
similar consequences, or require peer review for X amount of time. No |
74 |
investigation is needed because it's public. |
75 |
|
76 |
The forums are basically identical to IRC; someone screws up, the |
77 |
software has means to deal with things, and admins have access to PMs, |
78 |
logs, etc without any need to formally investigate and make a scene. |
79 |
|
80 |
>> Then |
81 |
>> there's the added bonus of those devs gaining free time to do things |
82 |
>> that they enjoy rather than listen to greivances and being expected to |
83 |
>> please two or more parties. |
84 |
> |
85 |
> They're volunteers. They're not forced to do anything. Volunteer |
86 |
> resources generally aren't fungible. If the Council said that MIPS is |
87 |
> a waste of time and the MIPS team is disbanded it isn't like they'd |
88 |
> just start spending the time they spend on MIPS on other Gentoo |
89 |
> projects. They might leave Gentoo entirely, or work on something |
90 |
> completely different. If having a MIPS team were actively harmful to |
91 |
> Gentoo it might still be the right call, but that is of course not the |
92 |
> case. |
93 |
|
94 |
Are you saying there are people within Gentoo that volunteer to get in |
95 |
other peoples' business and might leave Gentoo if they were removed from |
96 |
that role? Do we really want a culture of nosiness? You're right in that |
97 |
we're all volunteers, but perhaps we should stop for a moment and |
98 |
consider whether we *want* that sort of volunteer. We're strapped for |
99 |
manpower, sure, but not for that sort of thing. We need package |
100 |
maintainers, arch testers, recruiters, and mentors; not overseers. |
101 |
|
102 |
>> Large problems can be pushed to the |
103 |
>> council, as they are already elected officials who are expected to make |
104 |
>> impactful decisions on Gentoo. |
105 |
> |
106 |
> The way it works is that large problems get pushed to Comrel, and if |
107 |
> the people involved are dissatisfied with the results they can appeal |
108 |
> to Council. |
109 |
> |
110 |
> I don't have a problem with Council handling Comrel the way it handles |
111 |
> QA. I don't think it will improve things if the Council takes over |
112 |
> the role of Comrel entirely. The Council is a governance body for |
113 |
> just about everything in the distro, it isn't staffed to run |
114 |
> everything directly. If you let the Council delegate, then fine, but |
115 |
> now we basically are back to Comrel. |
116 |
> |
117 |
> And in the end any really big decision that Comrel makes ultimately |
118 |
> makes its way through the Council if the person impacted thinks that |
119 |
> they're likely to disagree with Comrel. |
120 |
|
121 |
As I mentioned a few quotes up, the system does not work as intended if |
122 |
there is co-membership between Council and Comrel, because it blurs the |
123 |
lines between the two. If one group is beholden to the other, and there |
124 |
are people who wear both hats, then clearly more influence lies with |
125 |
those individuals than either of the groups separately. It may not be |
126 |
outright corruption on paper, but the fact that it allows it to begin |
127 |
with is what I'd consider a bug or loophole. We generally patch that |
128 |
sort of stuff out. |
129 |
|
130 |
>> If an issue is not big or impactful |
131 |
>> enough to bring to the council, then the parties involved need to |
132 |
>> solve their conflict in other, hopefully more constructive ways. |
133 |
> |
134 |
> That is already what they're supposed to do before bringing it to |
135 |
> Comrel. From the few appeals I've seen that usually does tend to |
136 |
> happen to some degree, but if you're the victim of harassment and |
137 |
> somebody doesn't stop when politely asked to stop, then I don't think |
138 |
> we can expect much to come out of that. |
139 |
|
140 |
A fair point. Sometimes people just don't get along. Those are the edge |
141 |
cases where someone is actually being a social threat to Gentoo. To |
142 |
bring up my upstream metaphor again, we should tweak that only to the |
143 |
extent that it protects Gentoo. Of the incidents that Comrel has become |
144 |
involved in, how many were actually warranted? Has that information |
145 |
even been shared with you? Those are rhetorical questions, as I know |
146 |
policy dictates (at this time, anyway) it's kept quiet. I'd wager the |
147 |
involvement has been more than what should have been, and the |
148 |
organizational failure to audit that is telling. |
149 |
|
150 |
>> Third |
151 |
>> party platforms already have their own guidelines, so if someone is |
152 |
>> dissatisfied with Gentoo's action (or inaction), they can go to the |
153 |
>> platform moderators and we (Gentoo) don't need to be involved. |
154 |
> |
155 |
> And what if the platform moderator doesn't do anything, and somebody |
156 |
> is the victim of harassment on a medium that bears our name and logo? |
157 |
> What happens when people start branding Gentoo with the sorts of bad |
158 |
> behavior that is occurring on that third party site? |
159 |
|
160 |
That's not the Council's concern. Things having to do with trademarks |
161 |
are meant to be handled by the Trustees. The platform is notified with |
162 |
the trademark notice, and they have the choice to act on it or face |
163 |
legal action. Most platforms won't give it a second thought and will |
164 |
have the logo removed without hesitation. "Gentoo" is a species of |
165 |
animal and has uses in other industries, so the removal of the logo is |
166 |
likely enough to remove legal liability. I'm not a lawyer, but it holds |
167 |
up to logic imo. |
168 |
|
169 |
> And if you continue to do nothing, then good luck ever stopping |
170 |
> anybody who sets up an "Official Gentoo Website" or whatever with our |
171 |
> logo on it, because they'll just point to the various other third |
172 |
> party sites where we fail to take action, and a court can decide that |
173 |
> we've failed to protect our trademark and the Gentoo name is worthless |
174 |
> legally. Then somebody gets upset and sues the Trustees for failure |
175 |
> to carry out their fiduciary duties and spoliation of assets. |
176 |
|
177 |
Right. But again, the council doesn't handle that sort of stuff. As |
178 |
a councilor, you're obviously free to alert the trustees (like the |
179 |
rest of us), but ultimately a councilor worries about mostly technical |
180 |
or organizational things. I'm aware I suggested the Council handle |
181 |
social disputes, but when it reaches the point of the Council, there |
182 |
should be evidence of harm against Gentoo; so clear that it doesn't need |
183 |
investigation. Without that sort of undeniable evidence, no group in |
184 |
Gentoo has business doling out punishments or bans. |
185 |
|
186 |
>> |
187 |
>> tldr: we should treat social disputes the same way we treat upstream |
188 |
>> software: modify it only when Gentoo functionality depends on it. |
189 |
>> |
190 |
> |
191 |
> Well sure, and that is why people get to appeal to the Council. I |
192 |
> can't think of any Council members who have been supportive of |
193 |
> significant Comrel actions when somebody hasn't done something |
194 |
> seriously wrong. Now, I can't say who that sort of action has been |
195 |
> taken against, or what they've actually done seriously wrong, and I |
196 |
> get that this is going to make people suspicious. I just can't think |
197 |
> of a way to do that without creating liability both for Gentoo and |
198 |
> myself. |
199 |
> |
200 |
> There has been talk of reviving the Proctors to deal with more minor |
201 |
> stuff, basically empowering forum mods, IRC ops, and so on. |
202 |
> |
203 |
IRC ops and forum members should be part of Staff, if they aren't |
204 |
already. Again, we do not need a group of people whose sole purpose |
205 |
is to get into peoples' business. As developers, we're all partly |
206 |
responsible for Gentoo and are expected to act in its interests where |
207 |
relevant. If a dispute is not negatively affecting Gentoo, why do we |
208 |
need to get involved? |
209 |
|
210 |
-- |
211 |
Daniel Campbell - Gentoo Developer |
212 |
OpenPGP Key: 0x1EA055D6 @ hkp://keys.gnupg.net |
213 |
fpr: AE03 9064 AE00 053C 270C 1DE4 6F7A 9091 1EA0 55D6 |