User:Pathoschild/2010 global sysops vote summary

From Meta, a Wikimedia project coordination wiki

Pathoschild

This is a summary of the 2010 vote to create global sysops, posted to Talk:Global sysops/Vote#Final summary.

Final summary[edit]

This is an unofficial summary, not the closure of the vote. I myself don't feel strongly about either implementation or rejection.

The proposal has attracted roughly 1880 eligible voters, with a 1446:434 or 76.9% support ratio. While this suggests reasonable support, counting votes does not accurately represent the opinions of those commenting. Following is a summary of the points raised throughout the discussion, their prevalence and importance to the participants, and a few possible changes that address some of these concerns. (The summary is long but densely packed.)

I will not distinguish between users immediately affected or unaffected by the proposal, because this is more a global proposal than a multi-wiki proposal: affected wikis can opt out, and unaffected wikis can opt in. Serious points raised by users ineligible to vote are included, but not counted above.

Arguments in favour: Most supporting users added no comment or repeated the benefits described in the proposed text. The most common advantages cited were reducing vandalism or helping smaller wikis (7.3%), reducing stewards' workload (2.1%), and the ability of wikis to opt out (1.5%). Several supporters saw no disadvantages (1.3%), or noted the decreased need for more stewards (0.6%), well-defined limits to global sysop rights usage (0.6%), and the availability of steward oversight (0.5%). Less common points were forging a stronger crosswiki community (0.3%), multilingualism not being needed to spot vandalism (0.3%), the possibility of abolishing global sysops if the system fails (0.3%), global sysops being a quick and easy solution for a shortage of stewards (0.3%), and the inactivity clause (0.2%). Several more rare opinions are listed in my rough notes, but not mentioned here.

Many arguing in favour did have reservations. They stressed the importance of carefully selecting candidates (14.5%) and avoiding overbearing superadmins or dictators (7.6%), were skeptical about global block permissions which extend beyond affected wikis (0.2%), and were concerned about possible misunderstandings due to the language barrier (0.1%), the need for transparency (0.1%), and the need to recognize local sysops' priority in making local decisions (0.1%). Minority concerns included avoiding interfering with larger wikis, insufficiently inclusive automatic-opt-in criteria, the possibility of global sysop burnout, the need to minimize global block lengths, and the need for term limits.

Anoopkn summarized many commentators' attitudes when he said, "More policemen might not imply a more peaceful world, but perhaps thats one of the better things we can do".

Arguments against & discussion: Not surprisingly, opponents of the proposal were much more verbose. The most common concerns were that global sysops could not recognize abuse in languages they can't read and that different wikis have different standards and policies (12%), and the introduction of unneeded bureaucracy or hierarchy or management (9.9%). Many feared overbearing superadmins or dictators or crosswiki meddlers, or felt global sysops could not be trusted not to abuse their access (8.8%). Many saw the proposal as a concentration of power into an elitist group (8.3%), and objected to the inclusion of the 'global block' right which affects every wiki (7.3%).

Many felt that local communities should be allowed to take care of their own wiki without interference (6.9%). Several were concerned that the opt-out process was undefined (5.5%), though a proposal to require five voters was rejected. Many suggested closing wikis outright if they couldn't take care of themselves (4.8%). Several preferred election more stewards (4.4%), felt global sysops should be opt-in only (4.4%), felt they had far too much power (3.4%), or felt there was no need for global sysops (3.2%).

Several felt global sysops would be a form of imperialism on the part of larger, mostly English wikis (2.5%). Some wanted less inclusive auto-opt-in limits (2.1%), felt "uncontroversial maintenance" was too vague (1.3%), thought it should be discussed more widely before a vote (0.9%), felt it would drive away smaller-wiki users (0.9%), or suggested the vote was invalid due to midway changes or biased layout (0.9%). Minority concerns included that this was being forced onto smaller wikis by the larger wikis (0.7%), that it would make local adminship more difficult to obtain (0.5%), that global sysops should be tracked to prevent abuse (0.5%), that there should be some way to resolve disputes or discuss global sysops (0.5%), and that term limits were needed.

Supporters' responses to these points, and various points only raised once, are noted in my rough notes.

A few quotes help summarise common sentiments amongst opponents: "I'd rather see little maintenance rather than poor maintenance" (Zhouf12); "The majority of the people attracted to such power are exactly the kind of people we do not need" (Trackinfo); "When in doubt, vote against" (ShawnIsHere); "It sounds like a conspiracy theory to take over the world almost" (Presidentman); and "I, for example, wouldn't recognize vandalism in Russian or Tagalog if it bit me in the butt" (SMcCandlish).

Proposed changes: Many of the concerns are about issues inherent in the global sysop proposal, and cannot easily be addressed.

However, global blocking was a concern for both supporters and opponents. This would allow global sysops to block IP addresses on all wikis, regardless of which wikis they are normally limited to. Global blocking must be used very carefully, and I have often seen potential candidates ask for 6-month global blocks when a 2-hour block was sufficient. Global blocking should be removed from the proposal; global sysops can continue to ask stewards when a global block is needed.

Clarifications are needed for opting out, dispute resolution, communicating with or discussing global sysops, and appealing or reporting global sysop actions. Tools to track global sysop actions on affected wikis would allow concerned users to watch for potential abuse. Although only mentioned a few times, term limits (as implemented by en-wikisource) and wider elections may ease many opponents' concerns of a perpetual, English-dominated oligarchy. Wider use of Melancholie's Global notifications or my Synchbot would allow smaller wikis to more easily keep track of crosswiki developments, by having regular reports placed on a specified local page.

Other ideas for addressing the objections raised are welcome. —Pathoschild 05:57:00, 02 February 2010 (UTC)

Rough notes[edit]

These are rough notes about Global sysops/Vote that will be compressed into a summary after its closing. (Notes complete.)

initial comments
From a strictly statistical viewpoint, the proposal has attracted roughly 1897 eligible voters, with a 1463:434 or 77% support ratio. While this alone suggests reasonable support, straight vote counting is a poor way of making community decisions. Instead I'll attempt to summarize the points raised throughout the discussion, their prevalence and importance to the participants, and propose changes to address concerns.
I won't draw distinctions between users immediately affected or unaffected by the proposal, because this is more a global proposal than a multi-wiki proposal: affected wikis can opt out, and unaffected wikis can opt in. Serious points raised by users ineligible to vote are included, but not counted above.
"Yes" section
On the most part, supporting users added no comment or reiterated the benefits put forth in the original proposal. The main points raised can be grouped into the following:
  • Benefits
    • reduce vandalism / help small wikis (103 + 2);
    • help stewards deal with crosswiki abuse (28 + 2);
    • wikis can opt out (22);
    • no disadvantages (19);
    • Steward light / reduces need for more stewards (9);
    • well bound by parameters or restrictions (8);
    • Steward oversight (7);
    • forge crosswiki community (5 + 1);
    • trust not a matter of languages, don't need to speak language to spot vandalism, etc (5);
    • Can be abolished/changed later if doesn't work out (4 + 1);
    • terms limited (4);
    • Not perfect solution, but good enough for now (3);
    • better to have more stewards, but we don't (2);
    • Can I be one? (2);
    • Improve PR image, reliability, etc (2);
    • Suggest limited trial; compare to can-be-abolished-later? (2);
    • tested on Wikia, ie janitors (1);
    • Reduce local countervandalism work, so they can concentrate on editing (1);
    • Creating new wikis less risky (1);
    • Suggest WikiProject to track GSs on every wiki, with RSS feeds etc (1);
    • Suggest global sysop must be a local sysop somewhere before nomination (1);
    • Users already involved in thankless task of crosswiki countervandalism/spam unlikely to abuse power (1).
  • Cautions
    • stressed importance of carefully selecting candidates (20 + 1);
    • avoid being overbearing superadmins / dictators (8 + 3);
    • sceptical about global-block permissions, which extend beyond affected wikis (4);
    • noted possible misunderstandings due to language barrier (2);
    • avoid interfering with larger wikis (1);
    • dislike most enwiki admins (1);
    • this shouldn't be a vote (2);
    • criteria for automatic opt-in should be more inclusive (1);
    • account highjacked = chaos! (1);
    • tedious, unrewarding job; burnout (1);
    • regular re-elections? (1)
    • bad name; maybe “Global Assistant Sysops” (1);
    • process should be transparent, clear (2);
    • local admins should be consulted, have ability to override GS (2);
    • minimize global-block lengths (1);
    • shorter limits (1);
  • Pithy quotes
    • "More policemen might not imply a more peaceful world, but perhaps thats one of the better things we can do"
"No" section / discussion
  • Concerns
    • they cannot recognize vandalism in other languages / wikis differ (50 + 2);
      • ...but it's often obvious (1);
      • ...but this also applies to stewards (1);
    • introduces more bureaucracy / hierarchy / tyranny (38 + 5);
      • ...but it's necessary (1);
    • avoid overbearing superadmins or dictators / don't give more power to crosswiki meddlers / can't be trusted to use wisely (34 + 4);
      • ...but their role is strictly defined to avoid meddling (3);
      • ...but wikis can opt out (1);
      • ...but abusive global sysops will be de-sysop'd (1);
      • ...but admins actually do a pretty good job, if we could just stop harping on them and help them out a bit (1);
    • concentration/centralization of power into the few (32 + 4);
      • ...but we're expanding power beyond stewards (2);
        • ...but stewards are elected by crosswiki community (1);
    • should not have global block (31 + 1);
      • ...but stewards will be watching (1);
    • opt-out process not defined / vague (24);
    • we should let local sysops/communities take care of their own wikis (23 + 7);
    • Wikis that can't take care of themselves should be closed/ignored instead (21);
      • ...but they may one day be viable, so should be kept open (1);
      • ...but having global sysops do (some of?) the thankless work will encourage users to participate (1);
    • we should have more stewards instead / stewards are sufficient (18 + 1);
      • ...but stewards need help (4);
    • should be opt-in only (16 + 3);
    • have far too many unnecessary permissions / power (14 + 1);
    • no need for global admins (14);
    • global sysops not elected by local/crosswiki communities (11 + 2);
    • English/large wiki imperialism (11);
    • includes too many wikis, auto-opt-in limits should be less inclusive (9);
    • "uncontroversial maintenance" too vague (6);
    • should be discussed more widely first / notify more wikis (4);
    • will drive away small-wiki users (4);
    • vote is broken due to rule changes midway through process/biased layout/biased authors (4);
    • being forced upon small wikis by large-wiki users (3);
    • makes it more difficult to obtain local adminship, since global sysops take care of everything (2);
    • need term limits (2);
    • global sysop actions should be tracked to detect abuse (2);
      • ...but they can opt out (1);
    • global sysops need to communicate more, have a central place for discussion/review/notification/dispute resolution/etc (2);
    • "inactivity" removal criterion vague, what is considered activity? (1);
    • should not have mark-as-bot permission (1);
      • ...but cannot mark own edits as bot, only rollbacks; part of regular sysop rights (1);
    • should not be able to nuke pages (1);
    • won't help get people engaged in Wikipedia (1 + 1);
      • ...but that's not the problem it's intended to solve (1);
    • should be opt-out only (1);
    • rules too arbitrary (1);
    • should be more discussion on how candidates chosen (1);
    • better little maintenance than poor maintenance (1);
    • "It's Brazil (movie) -- with the unhappy ending." (1);
    • too much work for global sysops, they should take care of their own wikis (1);
    • should mention that wikis not included automatically can opt in (1);
    • local admins should be asked to perform actions if any are reasonably active, with global sysops only acting as backups or in emergencies (1);
    • should explore technical solutions: no account unification until edit threshold reached (1);
      • easily circumvented (1);
    • candidates should be carefully selected (1);
    • creating global sysops will encourage vandals to attack small wikis to get attention / perpetrate petty war game (c.f. deny recognition?) (1);
    • global sysops can band together to outvote every opponent and be unimpeachable (1);
    • editinterface allows more users to perform XSS attacks (1);
      • ...but all Meta users can do it already (1);
        • ...but that's easier to catch because it's all on Meta (1);
    • vote is broken, anyone can vote! (1);
      • ...but voters are being verified (1);
    • tighten requirements to foo (1);
      • ...but requirements are already stricter (1);
    • why do users from unaffected wikis get a vote? (1);
      • crosswiki policy, wikis can opt in or out (1);
    • vote suffrage should be much more exclusive (1);
      • ...but this would exclude most small-wiki editors (1);
  • Pithy quotes
    • "I'd rather see little maintenance rather than poor maintenance."
    • "The majority of the people attracted to such power are exactly the kind of people we do not need."
    • "When in doubt, vote against."
    • "It sounds like a conspiracy theory to take over the world almost."
    • "fight elitist stupidity, not globalize it"
    • "I, for example, wouldn't recognize vandalism in Russian or Tagalog if it bit me in the butt."
    • "I have absolutely no clue what all this is about, I just got an invitation..."