Help talk:Tracking changes

Confusing
I don't know if this will be answered (not the best location) but I have to say that I found the Article Validation to be very confusing. Only after reading this page twice did I finally realise that each drop-down box is for a different thing. Until then I thought that you select your 'category' from the first box and then rate it according to Accuracy and Coverage. I think a simple rating of 1-5 for each type of validation would be far easier to understand.--ЗAНИA talk 00:21, 27 November 2008 (UTC)
 * We do have a simple 0-4 rating for each of three criteria. Is that perhaps not what you meant? &mdash; Mike.lifeguard &#124; talk 03:49, 27 November 2008 (UTC)
 * I'm confused with it too :-(. Why can't I rate some categories 0?  If I select poor for one of the three then the button gets greyed out.  The only way I can rate something poor is if I rate all three poor which then ungreys the button and gives me a comment box. --AdRiley (talk) 13:55, 27 November 2008 (UTC)
 * It seems that you can only leave a comment when you are rating positivly (i.e. you have to comment and say why you're being positive). I think it would be very useful to be able to make comments for negative ratings (i.e. to say why you are being negative).--ЗAНИA [[Image:Flag_of_Italy.svg|15px]]talk 21:16, 27 November 2008 (UTC)
 * The idea is that if it's bad you're not going to flag it. Why would you want to show a stable-but-bad version to anon users by default? However, I agree we should tweak things so comments can be added at any level. We should probably get a bunch of tweaks together over the next month and list them all together at bugzilla. &mdash; Mike.lifeguard &#124; talk 23:25, 27 November 2008 (UTC)

Adding comments
Furthermore why can I only add a comment when I say that the coverage is 'good'? I want to always be able to add a comment.--ЗAНИA talk 00:33, 27 November 2008 (UTC)
 * That is something we should probably revisit. I definitely like the idea of leaving comments for flagging at any level (though of course not always necessary). &mdash; Mike.lifeguard &#124; talk 03:49, 27 November 2008 (UTC)
 * Thanks Mike (and for the above). Also how can one view comments left by other reviewers? --ЗAНИA [[Image:Flag_of_Italy.svg|15px]]talk 12:34, 27 November 2008 (UTC)
 * Special:Log/review &mdash; Mike.lifeguard &#124; talk 23:23, 27 November 2008 (UTC)
 * Brilliant, thanks. Is there any way to add special pages to my watchlist?--ЗAНИA [[Image:Flag_of_Italy.svg|15px]]talk 22:44, 28 November 2008 (UTC)
 * No, since the special page doesn't itself change, only the data in (on?) it. Just like the block log, or deletion log you'll have to check it yourself - though you should (I assume) see patrol log entries for pages on your watchlist, just like every other type of log. &mdash; Mike.lifeguard &#124; talk 23:05, 28 November 2008 (UTC)
 * OK. I knew that the special pages weren't addable to the watchlist but I never knew why until now.  I'll experiment reviewing some of the pages on my watchlist and see if the patrol log entries show on the watchlist when I have time.--ЗAНИA [[Image:Flag_of_Italy.svg|15px]]talk 23:18, 28 November 2008 (UTC)
 * It seems that reviews don't appear on my watchlist. I just 'reviewed' the English_as_an_Additional_Language/Telling_the_Time page and the review didn't show on my watchlist.--ЗAНИA [[Image:Flag_of_Italy.svg|15px]]talk 23:20, 28 November 2008 (UTC)
 * Normally your own actions do not appear in the watchlist. You might try to show your own actions.  &mdash; Mike.lifeguard &#124; talk 00:23, 29 November 2008 (UTC)

Must not share an IP with other users
I'm not entirely clear on what this clause means in practice. Who does it exclude? Pi zero (talk) 18:20, 27 November 2008 (UTC)
 * This is intended to combat sockpuppetry. Any legitimate users who are affected can request manual promotion. &mdash; Mike.lifeguard &#124; talk 23:21, 27 November 2008 (UTC)


 * I've asked this poorly.
 * Isn't IP address a property of a particular wikibooks session, rather than a property of a wikibooks account? (Any further questions in this vein would be predicated on the answer to that.)  Pi zero (talk) 06:47, 1 December 2008 (UTC)
 * It's whatever your IP is right now. So, your IP can change during a session, or it might stay static for months. &mdash; Mike.lifeguard &#124; talk 17:20, 1 December 2008 (UTC)


 * Is the criterion intended to mean that anyone who has ever connected from a shared IP is disqualified from automatic promotion? (Or in what way does the intended meaning differ from that?)  Pi zero (talk) 20:35, 1 December 2008 (UTC)
 * My current understanding is it is the IP right now - not any other IP. I don't understand how that's not clear. &mdash; <b style="color:#309;">Mike.lifeguard</b> &#124; talk 02:35, 2 December 2008 (UTC)


 * I've had reciprocal difficulty understanding why the source of my puzzlement isn't clear.


 * Does the clause mean that automatic promotion to editor can only occur at a moment when the user is connected from a non-shared IP &mdash; regardless of whether the user connected from shared IPs throughout the qualifying 100 edits etc., and regardless of whether the user ever connects from a non-shared IP again?


 * (If so, I'll find myself back at my original question, which I'll hope to phrase effectively this time.) Pi zero (talk) 01:14, 3 December 2008 (UTC)
 * Yes. So, taking a stab at answering your first question again. It means that people on shared IPs will not be automatically promoted while they are sharing an IP. If that continues to be the case, they will never be autopromoted, but can request manual promotion. Since that's probably not the case, they will simply be delayed in being autopromoted. Provided they meet all other requirements, they'll be autopromoted when they stop sharing an IP with another user. &mdash; <b style="color:#309;">Mike.lifeguard</b> &#124; talk 05:47, 3 December 2008 (UTC)

Edit intervals
What are they? Conrad.Irwin (talk) 14:26, 3 December 2008 (UTC)


 * It's my understanding that an "edit interval" is a period of time that separates two edits (not necessarily consecutive edits). In order for the user to qualify for automatic promotion to editor, the user's edit log must contain at least 15 non-overlapping edit intervals of at least 3 days each.  If I'm reading that right (from general remarks at mediawiki here, and wikibooks settings here), this criterion cannot possibly be satisfied in less than 45 days (3 times 15), so that the "30 days since registration" would never actually come into play under the settings we're now using.  Pi zero (talk) 10:02, 4 December 2008 (UTC)


 * I'm curious, where are you getting your information that the spacing and benchmark settings refer to number of days as well, and that you should multiple spacing and benchmarking to calculating the amount of time needed to satisfy a requirement? --<span style="font: bold 10pt 'courier new', comic, sans, ms;"><font color="midnightblue">dark lama  14:04, 18 January 2009 (UTC)


 * On closer inspection of the FlaggedRevs page on MediaWiki, I think I see where you could be getting that info. I still don't see how spacing and benchmarking should be multiplied to get what the minimum amount of time needed is though. The description is a bit too vague or ambiguous for that. If its true though, I guess the settings should be changed to something like at least 4 edits that are at least 7 days apart &mdash; that would take at least 28 days to do, which wouldn't be so bad. --<span style="font: bold 10pt 'courier new', comic, sans, ms;"><font color="midnightblue">dark lama  14:24, 18 January 2009 (UTC)


 * You mean 5 edits that are at least 7 days apart; that's the fencepost error I'd made. If you make five edits, there are four intervals between them, hence at least 28 days.  I'm pretty sure that's the right way to interpret the benchmarks field (i.e., the number of edits rather than the number of intervals) because of the phrasing of the key passage in the sparse, terse documentation on that page:
 * The user must have at least X edits that are Y or more days apart, where X is the number of benchmarks and Y is the time spacing.
 * (emphasis mine). I too was a bit frustrated at first by the lack of an explicit statement that the test requires each consecutive pair of benchmark edits to be at least Y days apart.  I wasn't quite satisfied with the common-sense approach, which is that either "Y or more days apart" means that every pair of benchmark edits must be Y days apart, in which case the minimum time to satisfy the test is Y times (X minus one) days, or the whole complicated test degenerates into "edits were made on more than one occasion, with the two occasions separated by at least Y days".  So I hunted down some of the discussions that surrounded the creation of the extension.  (I'd have to hunt them down from scratch, now.)  It was &mdash; as I recall &mdash; clear from the discussion that the spacing/benchmarks criterion was originally meant to be a very strong criterion to make sure the user had devoted time to the project on a large number of separate occasions.  That means non-overlapping intervals.  Pi zero (talk) 16:40, 18 January 2009 (UTC)


 * Well, it is certainly extraordinarily confusing, and whatever the settings are we need a way of describing what it means to people who ask! <font color="#E66C2C">Unusual? Quite <font color="#306754">TalkQu 16:48, 18 January 2009 (UTC)


 * It is, isn't it? How about: We check to see that X of your edits are at least Y days apart each — we look back Y days from your most recent edit, and find your last edit before that, and then again, and again, until we have done it X - 1 times. This is simply a way of giving a relatively dumb machine a simple way to figure out if you have been editing for a while with reasonable consistency. Chazz (talk) 09:00, 2 February 2009 (UTC)

Unable to fix template now
Dropimage has modifications that need done but I notice that I cannot make changes to it. It affects every instance of its use. Is the sudden addition of hide/show text to do with flagged revisions? Any ideas? Repairman (talk) 16:26, 6 January 2009 (UTC)

Estimated turn around time
What's the estimated turn around time for edits to be approved? I made some changes to how_to_learn_a_language about 4 weeks ago, but they haven't been approved. The article seems to have evolved into a long series or largely unsubstantiated bullet points. I'm willing to to help start a re-write, but I haven't gotten a response in the discussion and don't want to put too much time in if it's a dead end. Bryantt


 * Book: How to learn a language.
 * What do you mean by "edits to be approved". There is no need to get prior approval for anything that you don't expect people might find objectionable. There are large patches here at Wikibooks that have only a single author at a time. Give issues as long as you reckon it will take for active contributors to see your comments. You can also contact them directly. If you want input from more people, you can always try advertising your efforts on books of related content (in this case, the language textbooks. --Swift (talk) 18:46, 6 January 2009 (UTC)


 * The key point to remember is that this is a wiki, and as such is decentralized. There are administrators, but it is not their job to vet changes; if you are writing a book, you are the authority on that book, and if you are changing an article, you equally may be the authority on that article. If you have made enough edits, you will have the ability to "approve" the page you are working on. The Wiki has built into it the assumption that you will be honest in your editing, and equally can be honest in your assessment of the changes. Feel free to approve your own pages, if you feel they are of sufficient quality; if you don't yet have the tools (drop-down boxes at the bottom of the page), edit a bit more, and be sure to include reasons for your changes, and you will, after a while, get those tools automatically. Chazz (talk) 19:02, 6 January 2009 (UTC)


 * Maybe I'm missing something, but it seems someone with Editor status has to approve the changes before they'll appear to anonymous users ('10 changes awaiting review.'), which makes sense if there is author who has already invested a great deal of time in creating the book. I'd just like to be sure this person exists and discuss possible changes before I start writing.  Thanks for both of your responses.  Bryantt


 * The key point that you are missing is this: you are, or will be, an editor, once you have made enough edits. It would help greatly, f course, if your edits were under a name, rather than an IP address; sign up for an account, and edit when you are logged in, so that your edits are correctly credited. Chazz (talk) 21:50, 7 January 2009 (UTC)


 * Understood. Thanks Chazz.  Bryantt


 * Is there a way to turn off this Flagged Revisions feature for a particular book? It really cuts down on the number of people who just want to fix one typo here or patch up the grammar there without necessarily wanting to edit 10 things over however many days in order to get promoted to an editor.  There are some books where there appear to be no editors willing or able to approve changes, and people just get frustrated after a few edits that don't get approved for months, so they give up. —Preceding unsigned comment added by 41.241.89.88 (discuss • contribs)
 * By default, it's only enabled the first time someone cites the page. After that, the page is considered at least "stable" just in case it's necessary to roll back to an earlier version.  If there's a backlog of over 15+ changes ready for review, you can register for an account and be an active editor - you should automatically get the privilages to mark pages if you make enough edits across the wiki, but if it doesn't autopromote, you can post on WB:RFP.  --Sigma 7 (talk) 15:25, 11 February 2009 (UTC)
 * Moreover, Special:ValidationStatistics shows that 75 percent of anonymous changes to stable pages are reviewed in 2 hours, 95 percent are reviewed within 18 hours, and across main, Cookbook, and Wikijunior namespaces, only two stable pages have revisions awaiting review. So it appears the weeks-long backlogs are behind us now. --Damian Yerrick (talk) 03:31, 19 August 2010 (UTC)
 * Check that: As of today, the average review time is up to 16 days and the 95th percentile to 97 days. --Damian Yerrick (discuss • contribs) 17:22, 22 April 2011 (UTC)

Autopromote
http://noc.wikimedia.org/conf/highlight.php?file=flaggedrevs.php

I'd recommend stetting 'recentContentEdits' to 0 and 'email' to false. It just slows down the promotion rate probably. Voice of All (talk) 05:38, 25 January 2009 (UTC)
 * so you're suggesting that people who haven't done anything recently, and people who haven't seen fit to give us a valid email address, should become editors? I would have to disagree. In particular, having 10 recent edits is certainly not an onerous requirement. Chazz (talk) 20:03, 25 January 2009 (UTC)
 * As of thirty seconds ago, one of the editor criteria was worded as follows: "10 or more [of a user's] edits are in recent changes at the time automatic editor promotion is checked." But recent changes on this site appears to default to only the 50 most recent edits. Where I come from, having one-fifth of recent changes on a widely edited public wiki is considered flooding and indicates that either the user isn't using the preview button or the user is editing so fast that he should have sought a bot flag first. If it is not the intent to encourage a user to flood in order to become eligible for "editor tools", I would appreciate a clarification of this rule. --Damian Yerrick (talk) 02:37, 26 November 2009 (UTC)
 * This remains true... but you are looking at two different definitions of "recent". By definition, you can set the number of edits, or days, that are "recent" in the list that you view; mine is set to one hundred over seven days, for instance. However, the meaning of "recent" as far as editor criteria go is different... though I can't now locate what that value is, I suspect that "recent" means "within the past week" for purposes of autopromotion. I would be delighted to have an admin correct me, though. Chazz (talk) 07:48, 26 November 2009 (UTC)
 * This admin can't find further information on the particular variable that defines this setting beyond the comment in the extension's PHP file. As an aside, discussions on Wikipedia regarding this extension have led the community to decide against autopromotion at all due to concerns of gaming the system. -- Adrignola talk contribs 12:56, 26 November 2009 (UTC)


 * 30 days is the upper limit for Special:RecentChanges, so I'd figured that was probably the definition of recent for autopromotion. I don't think I've ever been aware of a case where this criterion alone prevented autopromotion.


 * As another aside, one often hears the claim on English Wikinews that they have an "easy come, easy go" attitude toward the editor bit, but in fact they don't use autopromotion, because the project's status with Google depends on editors' ability to enforce high publication standards &mdash; greater ability than mere activity can imply. So subjective judgments are required.  I think autopromotion, with some moderately nontrivial criteria, is legitimate here, but that's because we're trying for a modest level of trust combined with a modest level of understanding of Wikibooks; with a higher threshold for either, we probably wouldn't be autopromoting, either.  --Pi zero (talk) 15:39, 26 November 2009 (UTC)


 * Not to drag up an old topic but that actually did prevent me from automatically being given it, at first. I had all the other criteria met but had been inactive, once I had made enough recent edits it apparently set it off and I got it. So I doubt it's an actual issue in practice, and absolute worst case they can post on RFP. Xerol Oplan (talk) 10:01, 25 May 2010 (UTC)

Sighting and validating
Umm... is it true that an editor can only sight, and not validate? I'm an editor, and apparently I can validate (mark as level 3) &mdash; check the history of this page, for instance. A reviewer can mark as level 4, which is one better than validate... but I don't know what it's called, because of course I can't do it. Chazz (talk) 09:05, 2 February 2009 (UTC)


 * A sighted revision is any revision that has a consistent style, good or average accuracy, and has great coverage. I believe those are all within the range in which an Editor can review a page. So editors can sight pages and review pages. The only mention of "validate" that I was able to find at all in FlaggedRevs' documentation is in terms of Editors can spot check for nonsense/graffiti and Reviewers can spot check and proofread a page. I think both descriptions are based on what the default criteria used for reviewing pages are for FlaggedRevs. Wikibooks uses custom criteria for reviewing pages though. Editors on Wikibooks do the spot checking for nonsense/graffiti and proofreading instead. All a Reviewer on Wikibooks can do is say that a revision's style, accuracy and coverage is of "featured" quality, which is a reference to Featured Books. So I think "validate" is either meaningless on Wikibooks or just means that when an Editor reviews a revision of a page they have validated the revision. --<span style="font: bold 10pt 'courier new', comic, sans, ms;"><font color="midnightblue">dark lama  17:02, 2 February 2009 (UTC)


 * You're right. I guess the confusion comes in because the basic settings seem to only have "unverified", "sighted" and "good". Wikibooks has more than that.
 * As far as I can tell, "sighted" just means the lowest setting. meta:Help:FlaggedRevs is a useful page (and will be more so as it develops further). I'll update the page to reflect the fact that editors can review pages and that the only thing reviewers have beyond that is the highest setting, required for featured books. --Swift (talk) 20:50, 2 February 2009 (UTC)

Page title
I've just reverted part of an earlier edit. I removed a section header titled "Stable revisions" that spanned the entire page, and bumped the sub-section headers. If a section really spans the entire page, then the page should probably be renamed to its title.

I wasn't sure what the most proper page name would be so I checked mw:Help:FlaggedRevs which states that stable revisions are only revisions of a certain limit. On Wikibooks that limit is currently at the lowest (sighting). It might be more useful to keep a more general title for this page in line with the help page currently being developed over at MediaWiki.org. --Swift (talk) 07:57, 3 February 2009 (UTC)


 * Don't think of it as a section heading, but as a page heading. Books do that all the time, its considered part of Wikibooks' style. This goes back to my suggestion on Help talk:Contents to make the help pages follow some basic book style guidelines. --<span style="font: bold 10pt 'courier new', comic, sans, ms;"><font color="midnightblue">dark lama  14:47, 3 February 2009 (UTC)


 * The remainder of this discussion was moved to Help talk:Contents after it moved on to a more general discussion about the sectioning of content in the help namespace. --Swift (talk) 04:06, 11 February 2009 (UTC)

Merging and Reviewers and moving to project namespace
These two project pages currently contain very little and are largely redundant with this page. I'd like to propose that we merge them in here.

Also, after going back and forth on this for a while, I think this page belongs in the project namespace as it is rather descriptive of the project than a description of how to do something on Wikibooks. If we need something on the latter, we could put it up at somewhere like  or similar (though I think the drop-down boxes don't need much explanation apart from the annotation on this page). I decided to hold off moving since this page has already undergone a few renames and some might want a different name (see discussion above). --Swift (talk) 08:06, 3 February 2009 (UTC)
 * I am in favor of this proposal. This is all related information, let's put it all in one place. Chazz (talk) 08:24, 3 February 2009 (UTC)


 * [[Image:Symbol comment vote.svg|15px]] Comment I've been trying to untangle the issues involved here, and I'm inclined to think that
 * whether or not we merge those pages into this page may affect whether or not this page should move to project space, and
 * if the one would in fact affect the other, then we shouldn't merge those into this.
 * As this page now stands, its most substantive content is the assessment criteria, which one would consult when one wants to know how to assess pages for review; certainly that has been my reason for consulting them recently, pursuant to the thread in the reading room, to compare the criteria here with those at Using Wikibooks/Reviewing Pages &mdash; which is primarily a how-to page. The other current content of this page seems rather ambiguous in its allegiance; it might expand, in future, in a how-to direction.  However, the most substantive content of the other two pages is the list of autopromotion criteria, which is not how-to information, and if merged into this page would shift its center of gravity.  If the shift would be enough to tip the balance for this page from help space to project space, then the difference between this page and those pages was really enough to justify keeping them separate (though it might still make sense to merge the other two pages with each other).  --Pi zero (talk) 13:47, 3 February 2009 (UTC)


 * I don't really follow your logic there. What if it is a positive effect? The main reason for the merger is to get rid of these akward pages. Both are stubs and adding the minimal content over to this page would, I think be justified by getting rid of two extra pages. More related in another comment with the same timestamp, below. --Swift (talk) 16:58, 3 February 2009 (UTC)


 * I agree with the merger. I don't think this belongs in the Wikibooks namespace though. This is help not a proposal for a policy or guideline. Pages in the help namespace, don't have to be project neutral. In fact I think the point is to make the help pages more relevant to Wikibooks' way of doing things. --<span style="font: bold 10pt 'courier new', comic, sans, ms;"><font color="midnightblue">dark lama  14:51, 3 February 2009 (UTC)


 * In reply to the last two comments (Pi zero @ 13:47 and Darklama above); I'm happy to leave this in the help namespace for the time being. The main thing is to have a useful page that users can find. I feel a bit as if I'm splitting hairs here (linking properly to the page is more important to users than which namespace it's in) but I'll still let my rationale follow.
 * I think about the project namespace not as being restricted to policies and guidelines, but for any page with normative information about the workings of Wikibooks (the PaGs are just the ones we need explicit community consensus for use in big impact discussiosn such as deleting content or blocking users). The namespace, on the other hand, guides users through processes or helps the solve specific problems.
 * In the case of this page, I'd say that it doesn't so much help users review pages (there is no discussion of how to use the interface, for example, nor do I think we need that) but the page does contain normative information designed to standardise the understanding and usage of this tool across editors and reviewers. --Swift (talk) 16:58, 3 February 2009 (UTC)


 * I agree that the assessment criteria are normative, and I think I agree that normative material is best placed in project space. So I guess I'm in favor of moving it to project space.  The content of the other two pages, now in project space, actually isn't normative, but it's also fairly small, and I suppose it could probably be made to sit comfortably here even if this page eventually evolves into a guideline (as I suspect it will, someday).  --Pi zero (talk) 17:29, 4 February 2009 (UTC)


 * I think this page helps people to understand how to use the tool to review pages through explaining the review criteria that Wikibooks uses. I think pages in the help namespace can describe norms, and there are some help pages that do this. I guess I'm not saying that the Wikibooks namespace is only for PaGs, so much as I am that I don't see the Help namespace as being limited to only certain kinds of help or information, and I don't see any particular need for this to go somewhere else. I've moved pages from the Wikibooks namespace to the Help namespace before because the pages seemed to fit better with the intent of the help namespace, and I think I've seen other people do the same for similar reasons. I guess differences in opinion over which namespace pages should go in could be another reason why the Help namespace has been in bad shape. --<span style="font: bold 10pt 'courier new', comic, sans, ms;"><font color="midnightblue">dark lama  05:33, 6 February 2009 (UTC)


 * They are, indeed, in bad shape. Another reason why I don't consider the existance of normative pages in the help namespace as setting any precedence. Could you explain your view of how best to logically seperate this overlapping Wikibooks-desrciptive content into these two namespaces? --Swift (talk) 06:49, 6 February 2009 (UTC)


 * I've just merged the two. I held off on moving it to the project namespace to give darklama more time to explain his view on the boundary between help and project. --Swift (talk) 05:30, 11 February 2009 (UTC)


 * I'm not sure what that boundary is, I think figuring that out is a bit of a touch and go. Like for this case perhaps a clarification of the difference might be whether the page is descriptive of project norms or of the community's norms. Thinking about what is mostly in the project namespace, I think I would conclude that the project namespace is for things related to the community, like policies that the community is expected to follow, community discussions, how the community works, etc. verses content that describes Wikibooks, how Wikibooks works, how to do things on Wikibooks. I would say this page describes how to use a tool on Wikibooks, rather than about the community expectation of how the tool is to be used. Wikibooks:Media describes community expectations for instance while Help:Uploading_Images is more in the context of how to use the upload tool. Help pages might also be described as a means of helping people to understand what to do to meet the community's goals or how people can meet the community's goals, while pages in the project namespace describes the community's goals. I hope that helps explain where the boundary might be between help and project. --<span style="font: bold 10pt 'courier new', comic, sans, ms;"><font color="midnightblue">dark lama  11:00, 11 February 2009 (UTC)

My proposals for catagories:
83.24.20.136 (discuss) 09:54, 26 May 2017 (UTC)
 * poor - vandalism.
 * minimal - vandalism-free.
 * average - typo-free.
 * good - 100% correct.

Watching All Pages in a Book
The second section of the page is "Watching Pages". "Watching All Pages in a Book" should follow that but no, it's over at Using_Wikibooks/How_To_Edit_A_Wikibook. What? That makes no sense and helps less than it could. Please add the section here. Thanks, PeterEasthope (discuss • contribs) 14:06, 16 June 2017 (UTC)

Patrol
"New pages cannot be reviewed by reviewers for pending changes, until that page has been patrolled by an administrator; an exception is pages created by reviewers and administrators." &larr; Really? I've never understood patrol, but was vaguely aware of it, and had never caught any hint that it had anything whatever to do with review. I'm tempted to experiment to try to witness this phenomenon, though it's a bit tedious (since I can't use my primary account, which has the admin bit). --Pi zero (discuss • contribs)
 * Yes, pending changes of pages in Special:UnreviewedPages can't be reviewed. "This page may need to be reviewed for quality." is shown. Minorax (discuss • contribs) 13:08, 3 March 2020 (UTC)
 * If a patrol is the first review of a page (as the text says), then how is that not just another way of saying the page hasn't been reviewed yet? --Pi zero (discuss • contribs) 15:04, 3 March 2020 (UTC)
 * I think it might be about having a "stable version" of the page. See this, which states "20:09, 7 July 2010 Adrignola set stable version settings for Wikijunior:How Things Work/Light Bulb [Default: Stable] (hist)". Minorax (discuss • contribs) 15:19, 3 March 2020 (UTC)
 * That's a different thing. Each page (in any namespace with flaggedrevs enabled) has "page stability settings", which can be set by an admin (there's a special page for it, that's linked from the "change protection" panel).  The first setting is "Revision displayed on default page view", and when a page is created that is set to "The latest version".  In the summer of 2010, we decided that for all pages on Wikijunior we should change that to "The stable version" &mdash; thus, if a vandal puts something obscene on a Wikijunior page, the kids won't see it by default.  It looks as if the page you linked had its settings changed on 7 July 2010; but the earliest review of that page was on 10 March 2009. --Pi zero (discuss • contribs) 16:45, 3 March 2020 (UTC)
 * Thought that the issue was with that. My bad. Minorax (discuss • contribs) 04:10, 4 March 2020 (UTC)

it turns out that reviewers can can review unpatrolled pages, but in an unexpected way: Well I tested this out on Sanskrit/Numbers and it did work, I was able to review it despite the page not having been patrolled (you can check this on the page log). But its all bit odd really. I will try to update the page with my discovery.
 * 1) Obviously reviewers can't patrol pages (except for autopatrol of self-created pages). Patrol is different to review since (a) all created pages can be patrolled (although there is also autopatrol from reviewers) while only some are subject to pending changes; (b) there is a separate log at Special:Log/patrol, while review is at Special:Log/review.
 * 2) Reviewers can't review unpatrolled pages from Special:RecentChanges (but see 4. beneath). There is no [review] tab displayed for these pages for reviewers. An example would be Real Analysis/Continuity.
 * 3) Reviewers can't review unpatrolled pages from the page history (but see 4. beneath). e.g. doesn't show any review options for me.
 * 4) However I have discovered a way! Taking Real Analysis/Continuity as an example, at the bottom of the page is an option that says "Review this revision - You can advertise to other users that you are reviewing these changes. - Quality: poor/unrated  minimal  average  good - Comment: ". Now if you press the "Accept revision" button it actually reviews the page, rather than just advertising a review (as you might have thought). After that the page can be reviewed as normal

Btw I used my test account User:Mrjulesd(alt) during testing. If you wanted to see for yourself you could (a) create a test account; (b) give it reviewer privilege (uncontroversial since you are an admin). You might find it interesting to see it. -- Jules (Mrjulesd) 20:36, 3 March 2020 (UTC)
 * Just going to add: another oddity is that only manual patrol shows on the patrol log. Autopatrol doesn't show up as such. I don't understand why this is the case really. -- Jules (Mrjulesd) 20:48, 3 March 2020 (UTC)
 * I still don't understand all this. As I say, I've never understood patrol (and still don't); and I haven't-and-don't understand how they relate to each other.  I'm not even sure which of them predates the other, although one might guess patrol came first since it's a simpler/lesser device. --Pi zero (discuss • contribs) 12:59, 4 March 2020 (UTC)
 * The chief difference between pending change review and patrol is that pending change review is a process that needs to be done each time it is edited, while patrol is only done once per page creation. I think I'm pretty familiar from my work on English Wikipedia. There, hardly any pages are under pending change protection; they are listed at w:Special:StablePages, and there is only a few thousand I think. So the vast majority of en.wp articles don't have pending changes, although they may have other protection (e.g. semi-protection); while all books on en.wb seem to be under pending change protection. However all created pages at en.wp are patrolled; pages needing patrol are listed at w:Special:NewPagesFeed. To patrol pages there you need a user permission called New Page Reviewer (also called patroller), so it is quite restricted. So in summary on en.wp pages are hardly ever reviewed for pending changes, but all pages are subject to a single patrol some time after creation (although there is a backlog of about 7000 unpatrolled articles there currently). -- Jules (Mrjulesd) 13:23, 4 March 2020 (UTC)