Thinking aloud here. We know that policy on external linking is, well, not in a settled state.
With a meta-Arbitrator's hat on, I say this: ArbCom can in the "attack sites" case possibly explain something about implementation of existing policy, mainly WP:HARASS. ArbCom doesn't write policy; what it typically does is to explore the amount of "stretch" in existing policy and custom.
Policy is not showing up too well here; turn to "process".
We just have this fragment at the moment: links to ED can be speedily deleted from the site. We are not even very advanced in templating links for "proposed deletion"?
By the article deletion analogy: the following might work, restricted to namespaces other than the article namespace. Have a three tier process:
- speedy deletion, restricted to ED and any sites explicitly put alongside it (so these are the "attack sites"); {{hangon}} only on the grounds that the page is a clearly reliable source and the link is in context.
- proposed deletion by template, to remove junk
- LfD process, to handle contested cases of proposed deletion, and also any mass deletions of links from one "site" (mirrors etc. - what is a site?).
An obvious drawback is that the discussions in the contested cases would attract attention (and might make ArbCom Workshop pages look like a tea party in comparison).
How would the speedy criteria work? Perhaps three-time losers under a mass deletion, plus ArbCom remedies in cases. In other words Slashdot, Slate and so on would usually only come under consideration after a long history.
How could an editor get "banned"? For being too out of process, or persistent recreation of deleted links, editors could be sanctioned.
Charles
----------------------------------------- Email sent from www.virginmedia.com/email Virus-checked using McAfee(R) Software and scanned for spam
It would be interesting to experiment with a process to force discussion and consensus, rather than allowing single-actor edit-warring over links. But if Arbcom can't create policy, can it create process?
It is also the case that deletion neccessarily require some kind of process in part because any editor can desire deletion but only admins can perform it. Any editor can edit article content including removing links, and creating a LfD process could really enable wikilawyering by people who support links that already have broad support for removal, like personal web sites, myspace, Youtube copyvios, and so forth.
If you think that Arbcom actually can CREATE a new process, I would suggest something like this (this is a pretty huge set of ideas and I wonder if ArbCom could pull it off even if it wanted to)
1. In the event of a dispute over an attack link, the link is removed pending discussion. 2. Discussion held at WP:LfD (so that if the outcome is "remove" there isn't a huge archive of content on the talk page needing to be deleted). 3. A flag put on the article talk page pointing to LfD. 4. Closed by an uninvolved admin after 5 days. 5. For links judged as "removed" the closing admin will obfuscate them. 6. Editors who disrupt the article by adding or removing links while the discussion is ongoing or against consensus after the discussion is closed can be banned from editing that article or talk page for a reasonable period (a la article probation). 7. Process to be reviewed in 6 months to see if a stable consensus on links has developed making the process no longer needed.
For this to work well, I think ArbCom's principles in the attack sites case would also have to clarify the following (I haven't been following the case so I don't know if you are close to this yet) 1. Wikipedia has an obligation to protect its editors from harassment. 2. Interactions between editors are generally covered by the NPA and harassment policies. 3. Notwithstanding #1 and #2, article content is generally covered by a different set of policies (NPOV, reliable source, verify) and only in extreme cases should policies designed to cover editor interactions intrude into article space.
With remedies like this: 1. Links added to project or talk pages with the intent or effect of harassing or intimidating other editors may be removed under the existing NPA and harassment policies, and repeat offenders may be briefly blocked by an uninvolved admin. 2. Links added to article pages should be considered under article content policies. 3. Disputed links in article space to be discussed at LfD (and removed during discussion). 4. Editors who disrupt the encyclopedia by removing or adding links during LfD or against consensus decided at LfD may be banned from that article (and related articles if necessary) for an appropriate period of time by an uninvolved admin; ban enforceable by blocking.
On 10/12/07, charles.r.matthews@ntlworld.com charles.r.matthews@ntlworld.com wrote:
Thinking aloud here. We know that policy on external linking is, well, not in a settled state.
With a meta-Arbitrator's hat on, I say this: ArbCom can in the "attack sites" case possibly explain something about implementation of existing policy, mainly WP:HARASS. ArbCom doesn't write policy; what it typically does is to explore the amount of "stretch" in existing policy and custom.
Policy is not showing up too well here; turn to "process".
We just have this fragment at the moment: links to ED can be speedily deleted from the site. We are not even very advanced in templating links for "proposed deletion"?
By the article deletion analogy: the following might work, restricted to namespaces other than the article namespace. Have a three tier process:
speedy deletion, restricted to ED and any sites explicitly put alongside it (so these are the "attack sites"); {{hangon}} only on the grounds that the page is a clearly reliable source and the link is in context.
proposed deletion by template, to remove junk
LfD process, to handle contested cases of proposed deletion, and also any mass deletions of links from one "site" (mirrors etc. - what is a site?).
An obvious drawback is that the discussions in the contested cases would attract attention (and might make ArbCom Workshop pages look like a tea party in comparison).
How would the speedy criteria work? Perhaps three-time losers under a mass deletion, plus ArbCom remedies in cases. In other words Slashdot, Slate and so on would usually only come under consideration after a long history.
How could an editor get "banned"? For being too out of process, or persistent recreation of deleted links, editors could be sanctioned.
Charles
Email sent from www.virginmedia.com/email Virus-checked using McAfee(R) Software and scanned for spam
WikiEN-l mailing list WikiEN-l@lists.wikimedia.org To unsubscribe from this mailing list, visit: http://lists.wikimedia.org/mailman/listinfo/wikien-l
As an alternative, Arbcom could turn the focus of the current case away from policy and towards editor behavior in the link removal disputes that sparked the case. Start with the same principles:
1. Wikipedia has an obligation to protect its editors from harassment.
2. Interactions between editors are generally covered by the NPA and harassment policies.
3. Notwithstanding #1 and #2, article content is generally covered by a different set of policies (NPOV, reliable source, verify) and only in extreme cases should policies designed to cover editor interactions intrude into article space.
With remedies like this: 1. Links added to project or talk pages with the intent or effect of harassing or intimidating other editors may be removed under the existing NPA and harassment policies, and repeat offenders may be briefly blocked by an uninvolved admin.
2. Links added to article pages should be considered under article content policies.
3. Disputed links in article space to be discussed on the talk page. The normal dispute resolution processes (third opinion, RFC, mediation) apply, and the link will be obfuscated or unlinked during the discussion.
4. The following editors are briefly blocked or desysopped for edit warring over link removals:
Eh?
On 10/12/07, Thatcher131 Wikipedia thatcher131@gmail.com wrote:
It would be interesting to experiment with a process to force discussion and consensus, rather than allowing single-actor edit-warring over links. But if Arbcom can't create policy, can it create process?
On 10/12/07, Thatcher131 Wikipedia thatcher131@gmail.com wrote:
- Wikipedia has an obligation to protect its editors from harassment.
"Obligation" is too strong a word. But whenever it is possible to protect somebody from harassment without causing greater disruption to the remainder of the project and community, it's a very nice thing to do.[1]
- Interactions between editors are generally covered by the NPA and
harassment policies.
If you mean interactions on Wikipedia, I'll agree. Interactions elsewhere are not covered by en.wikipedia policy[2], instead falling into the frontier jurisdiction of common sense, [[Judge Roy Bean]] presiding.
- Notwithstanding #1 and #2, article content is generally covered by
a different set of policies (NPOV, reliable source, verify) and only in extreme cases should policies designed to cover editor interactions intrude into article space.
"Only in extreme cases" should mean "almost never", but in practice every case involving one or more external links criticizing Wikipedia users[3] is considered extreme.
With remedies like this:
- Links added to project or talk pages with the intent or effect of
harassing or intimidating other editors may be removed under the existing NPA and harassment policies, and repeat offenders may be briefly blocked by an uninvolved admin.
Let's consider the slight practical differences between the following: A. Writing a lengthy personal attack on somebody's talk page, and: B. Writing a lengthy personal attack on your blog and pasting it on somebody's talk page.
When the most noticeable factor is server load[4], you might as well be splitting atoms.
Beyond that, option B does make it possible for one to tone down the "attack" after the intended "target" has become quite offended, but before anybody else sees it, which could easily degenerate into "he said, she said", but it private communication would always be more effective for that.
Also, option B creates a BADSITES situation which will immediately make you the newest cause célèbre of two diametrically opposed factions who have absolutely no idea what is actually going on between you and the person you are apparently trying to attack.
Meanwhile, back on the mailing list, there will be a handful of reasonable people who don't really care one way or the other about this particular incident, and would prefer to meta-discuss it ad nauseum rather than take action, eventually passing around a new draft proposal because you've just proven that the last one wasn't good enough.
Either way, "remove personal attacks"[5] seems to have been for better or worse "rejected by the community", but people do it anyway (again, a "common sense" thing, for "extreme cases"), making this distinction more or less moot, as the second half of the suggestion merely provides that "repeat offenders can be blocked", which is already (and has always been) the case for personal attacks.
- Links added to article pages should be considered under article
content policies.
The key issue here should be whether the article contains (or could contain, upon further expansion) information relevant to the topic and obtainable through the link(s) in question. I doubt actual "content policy" for external links enjoys this level of brevity, but this is what makes sense to me.
- Disputed links in article space to be discussed on the talk page.
The normal dispute resolution processes (third opinion, RFC, mediation) apply, and the link will be obfuscated or unlinked during the discussion.
I suppose "obfuscated or unlinked" would be better than "referred to so vaguely that nobody knows what the hell we're talking about".
- The following editors are briefly blocked or desysopped for edit
warring over link removals:
Blocks for edit-warring are common enough, but has anybody ever been desysopped for it (i.e. for edit-warring on a non-protected page)? That strikes me as an unusual "remedy".
—C.W.
[1] http://www.usemod.com/cgi-bin/mb.pl?DefendEachOther [2] http://en.wikipedia.org/wiki/WP:NPA#Off-wiki_personal_attacks [3] (other than Essjay, that is) [4] http://en.wikipedia.org/wiki/WP:DWAP :p [5] http://en.wikipedia.org/wiki/WP:RPA
Hi, Charles. Thanks for this! It's exciting to see the possibility of progress.
charles.r.matthews@ntlworld.com wrote:
By the article deletion analogy: the following might work, restricted to namespaces other than the article namespace. Have a three tier process:
speedy deletion, restricted to ED and any sites explicitly put alongside it (so these are the "attack sites"); {{hangon}} only on the grounds that the page is a clearly reliable source and the link is in context.
proposed deletion by template, to remove junk
LfD process, to handle contested cases of proposed deletion, and also any mass deletions of links from one "site" (mirrors etc. - what is a site?).
An obvious drawback is that the discussions in the contested cases would attract attention (and might make ArbCom Workshop pages look like a tea party in comparison).
I think this would resolve most of my concerns with the current approach. In particular, it fits in well with Wikipedia's tradition of resolving disagreement through open discussion, hopefully with involvement from a selection of disinterested parties.
Although it would be a contentious area for some links, the good part is that it would be focused contention. Right now the contention for this seems to be frustratingly diffuse. I'd much rather have it in one place.
William