User:KnowledgeOfSelf/That Whole Admin Bit

These are my personal views on administrator actions, mainly on the ones that I myself partake in the most, although I will include commentary and opinions on areas I hardly ever explore. This is not a policy page, it is rather more like an essay. I'm going to be using it more or less as a reference for myself so I do not get dragged too far away from the way I apply my use of the adminship tools.

Using rollback
I try not to use rollback on anything but vandalism, the reasons being varied but the main one being, it's incredibly rude to do so against an other editor. Rollback doesn't leave a proper edit summary instead it has a generic "reverted edits by "xxxxxxx" to last version by "yyyyyyy". So if someone is adding what they believe to be viable content to an article when in reality it's unencyclopedic, a violation of what Wikipedia is not, and orignal research, using rollback would be quick and easy but it's still rude. (I will not lie I do use it once in awhile for things other than vandalism - especially when I'm on vandal patrol and the vandalism is heavy). If you are going to revert the edit, use the undo tab while adding an edit summary, or do it manually. (The undo feature allows you to input an edit summary, rollback does not.)

Blocking and unblocking
I'm sure everyone knows this, but blocks should not be used to "punish" an other editor. For example you block an editor for the three revert rule they contact you, apologize and assure you that it won't happen again. This is where assume good faith comes in! (With common sense of course&mdash;if the editor that is blocked has been blocked for 3RR numerous times, or has been disruptive in the past&mdash;it may be time to just let the block finish its course.) Go ahead and lift the block, (it is their first offense) on the condition they not touch the article (for at least 24 hours) they violated the policy on in the first place. Make it clear that they would be reblocked if they do so. Remember blocking is not a punishment, in a situation like that it is used to help cool off the editors involved, and to prevent edit wars.

Blocking for vandalism
This is my exception to the above, I will block repeat and long term vandals, even if they have not been warned to an immediate level 4. For example an school IP address that has been blocked 12 times before, whose talk page is covered with last warning and blocked messages, is obviously causing harm to the Wiki, and I see nothing wrong with blocking them. Especially now that administrators can block anonymous users only, which makes it possible to block school IPs, and other shared IP addresses for longer periods of times. I almost never long term block IP address's with both account creation blocked and anonymous users only checked. Regardless of templates such as schoolblock I just personally do not believe that is right. You never know what innocent user is blocked because their school mates don't know how to behave, or because the country of wherever has a ratio of of 20 vandals for every legit user new and long established. When blocking long term I prefer to use anonblock and disable anonymous users from editing while allowing accounts to be created from that address. Sure a vandal may take advantage of that, but with the numerous editors on RC patrol vandals usually get caught rather quickly be it from an IP address or a registered account.

I also rarely ever go with the full 4 warnings before a block. If it's obvious or obscene vandalism they get an automatic level 4 warning (bv,test4,Test4im) if they vandalize again it's a block. If it is not obvious or profane vandalism, I usually start with a level 1 warning (test1 test1a), then a level 3 (test3,test3a), a level 4, and then the block. Sometimes level 3 is skipped if it is a mixture of obvious and subtle vandalism. Those are judgment calls, every administrator has to make them. I also will block a vandal who has had a recent level 4 warning or a recent block, (usually up to 2 weeks old, or if they have more than 1 level 4 warning in the same month) assuming good faith is a very valuable policy, but it should not be followed blindly. (In a way that sort of falls under ignore all the rules but that's a different story) I will not block a vandal listed at AIAV unless they are currently active. I will consider 2 week old level 4 warnings or recent blocks, but if they are not currently active I will not block and I will de-list the report.

Closing AfD's
AfD's are not a vote! Well yes they are, it is foolish in my opinion to deny that fact, it is also foolish not to take into consideration that it is about discussing if an article should be deleted. If 20 people comment on an AfD, 15 say keep, 4 say delete and 1 thinks a merge would be the best way to go, how do you determine what is right, what is the consensus, and what you should do without blindly counting the votes?


 * First off make sure you as the closing administrator read the article that is being discussed, that will help you right away in determining an outcome for the discussion.
 * Then read the nominators reasoning for deletion, is it sound? What policy does the nominator believe the article does not adhere to? Next read the on going discussion, if the nominator thinks the article violates our no original research policy, and our notability policy, decide for yourself. As an administrator who is closing the debate, you should be familiar with the deletion policy, as well as the policies concerning the articles deletion. (If you read the article and the two policies the nominator cites, then you should be alright)
 * On to the discussion, if 12 of the editors who want to keep the article give reasons such as "Keep, per the above", or "Keep, don't delete this article, it's the best one on Wikipedia", et al. Those comments should be ignored, or at least taken with a grain of salt as they do nothing to promote consensus. Now if one user gives a valid reason why the article should be kept, and an other user says "Keep, per "xxxxx's" rationale, I couldn't have said it any better myself." In that case it is different, there is nothing wrong with agreeing with someones rationale the key is to determine whether or not they are voting blind.
 * Next determine why the 4 editors believe the article should be deleted. If they cite the same reasons as the nominator, then examine the article further, (can you determine why the article fails the policies yourself?), if the article obviously fails the cited policies deletion may be the way to go if no one attempts to fix the problems. (Barring a merge not being feasible).
 * Now if the one person who believes the article should be merged, uses a rationale such as "The contents of this article while alone may not be notable for it's own article, the information is somewhat relevant to "article X", I think merging is the best way to go." If that is the case, and you have established that while the article as it stands fails the cited policies, the information itself is not worthless, then perhaps a merge is the best way to go. If you are unsure of how to perform the merge, then read this page. Make sure that if you go with the merge, you state the rationale for your decision, at the top of the articles' AfD page as it may be contested and even over-turned.
 * If you think the merge may be a viable way to go&mdash;but do not believe it can done with current consensus, (as is most likely the case with only 1 person !voting that way) then you can always re-list the AfD. (Again explicitly state your reasoning for re-listing&mdash;but also know that, that decision may very well be overturned, especially on controversial nominations. Don't be scared to be bold, but don't be overly rash either!)

Speedy deletions
These can be tough, sometimes an article so blatantly violates our speedy deletion policy that all you really have to do is find the relevant criteria to use in your in your log when you delete it. The tough ones fall under A1, and A7 as they can be difficult to evaluate. What exactly constitutes "no context", if the article is one good sized paragraph (e.g. 7-9 lines) is that a enough context? Sometimes yes, sometimes not, it depends on the content of the article. Is it encyclopedic? Is it verifiable? It's a good idea to check against other criteria the article might fall under. Is it notable (A7)? Perhaps it's in an other language (A2), if that is the case tag the article with notenglish one of the many helpful Wikipedians who are polyglots are bound to take care of it. What do you do when you come across an article that has more than 1 criteria for speedy deletion, you can use the most obvious criteria as your reason, or you can put done all of the relevant ones, no big difference either way. When you delete an article without specifying a reason, your log will instead display something such as "content before blanking was: xxxxxx xxx xx x x xxxxxx xxxxx xxx xxxxx", or "content was: xxx x xxx xxxx xxxxxxx xxxxx xxxx xxxx xxxx xx x xxx xx xxxxx". It basically will display the first sentence of the article you just deleted. While obvious speedy deletions will sometimes be deleted this way, it generally is not something I try to do.

Images
I almost never delete images unless it is blatantly a copyvio that can not be claimed for fair use. As I do not partake in this particular area of maintenance I don't have much to say on it. And as my understanding of our policies on images and media is rather limited (blasted copyright laws are confusing to me) I don't have much of an opinion on any of it.

Full protection
Using full protection is rather simple, it can be used to cool off an edit war, in which case the article should not be touched by anyone. (Unless they are removing vandalism, obvious copyright problems, or for office actions which are only done by WMF staff members and User:Cary Bass) It is also sometimes used to prevent abuse from trolls and other users who are mis-using their talk page. Most templates that are "high use" templates are always protected to prevent abuse, they should not be edited without consensus and they should never be unprotected unless they are scheduled to be redirected. (Again they must have consensus for this to happen)

Semi-protection
Is generally used to combat heavy or very obscene vandalism by IP's and new vandal accounts. It is sometimes placed on user pages including user talk pages, though this should be avoided unless the page is being attacked relentlessly. If semi-protection is needed on a talk page I try to keep it between the 1-5 hour range depending on the situation. It is very wrong to permanently semi-protect your talk page, as it prevents users editing from an IP address from contacting the user whose page is under protection. The only time semi-protection is in place on an article on a permanent basis, are high profile pages well known to attract vandalism, e.g. George W. Bush. Some policy pages are also permanently semi-protected to prevent abuse.

Move protection
Move protection should be applied on a permanent basis to very few articles in main space, or to other highly visible pages. Some examples include AIAV, or the Main page. I believe it is acceptable to protect user and user talk pages from moves, as they never have to be moved to an other name.