Is this really what is wanted?
I'm sorry but this seems like development for development's sake. JUST MAKE THE MARK AS PATROLLED BUTTON MORE ACCESSIBLE!!!! But that's too hard to code properly and ruins performance, but a fancy new interface is the answer. KISS principle, please. If every time you (maybe just opt-in, maybe just autoconfirmed, maybe everyone) read an unpatrolled page you saw the little "Mark as patrolled", then I'm sure the backlog would disappear in days. At least have a trial.The-Pope 12:38, 20 September 2011 (UTC)
I understand your frustration on the issue of the bug request. Asking for seemingly simple things and having them closed or delayed for performance reasons is not fun. But please understand that not only are some of the responses are years old, they're not made by the same people who have worked on the very early stages of this draft design. We're not a hive mind (yet) at the Foundation. ;-) The truth is, this interface might also be rejected for performance reasons too, we won't know until we ask someone to code it. At this stage we're just trying to think of a variety of solutions.
P.S. Thanks for dropping your comment here as well as on English Wikipedia.
To be honest, we need to build interfaces like this for several reasons above and beyond New Page Patrol. We have a large set of what we call "curation" problems, some of which are only just now beginning to surface. Rapid scan-and-tag systems will be needed for many, many things - not the least of which is image curation for commons.
A button alone will not be of much help if the primary problem is that people are not educated as to how to patrol, or what constitutes a patrol.
A button would be very useful if it made more efficient use of volunteer time and brought experienced editors in on articles they understand. Quite a few of the articles that reach the back of the queue have actually been categorised or otherwise cleaned up, especially if the back of the queue is running to thirty days. The people doing that are often very experienced editors, they may even be members of wikiprojects looking at articles that have just turned up in the high level categories for their project and which need checking for hoaxes and often a more appropriate category applied. If those editors also had the option to just click the appropriate ones as patrolled it would take some of the pressure off the end of the queue, and save everyone time.
Please don't underestimate the importance of that Bugzilla request - there are a few medium sized opportunities to improve NPP, but this is the only uncontentious big improvement on the table.
I welcome the signs that Brandon is slowly but surely warming to the idea that the transient mass of New Page Patrollers is basically uneducable. The next thing for software developers to understand is that there is a limit to the amount of human experience that can be replaced by scripts and filters on a php driven website - it's probably not such a good idea to be constantly searching for electronic solutions just because Wikipedia happens to be an electronic encyclopedia. As I've intoned before, sophisticated tools are only of any use when in the hands of experienced users.
That remains to be seen ;) There appears to be some kind of consensus (here at least) that NPP should be a user right for competent patollers. It is hoped that the current survey will shed more light on this. Whichever way the survey itself goes, the need for a user right would appear to be inescapable.
I am currently working on the development of a video tutorial for NPP, but much of the completion of it depends on the development of the excellent Zoom project that IMHO so accurately addresses NPP issues already, that all it requires now is coding up and putting on TestWiki for us to tinker with and suggest minor changes and/or additions.
I suggest we consult the Village Pump on the affected wikis before proceeding further with the idea of making a new userright. It might be more controversial than the conversations on this page show.
There is absolutely no intention to impose new user rights without consensus from the community. The overall community feeling regarding an NPP user right is hoped to be demonstrated by the results of an NPP survey which will take a couple of weeks to complete and analyse. Development of the Zoom tool should continue, because it will be used whether or not a user right is installed.
I've struck the bit about PagePatroller rights as that was opposed by a majority in the survey. I think that a detailed proposition with a clear explanation as to how it would work might be able to get consensus support on EN wiki, but you'd need an RFC for it. More importantly ThePope's point is valid - making the mark as patrolled box visible to any autoconfirmed editor who opens an unpatrolled page would solve the problem and in a far more wiki way than Zoom intends. Currently you only see that box at special newpages so all the people categorising articles and looking at new articles that have turned up in their wikiprojects categories don't get the opportunity to patrol them.
(I'm not sure where to put this; I really dislike Liquid threads.) I see two really major problems in the proposed implementation which to me would make everything very much more difficult: One. the list view is much more expanded. the way I patrol, the way i think most experienced people patrol, is to scan an entire large portion of list for something which strikes us as worth our noticing. By this I mean , not anything that needs attention, but something that needs the particular attention that I in particular can and want to give it at this particular time. There is no way i can reduce this to a single algorithm, because I know I look for a very wide range of different things at different times, depending on my degree of alertness, the time available, and what I am most concerned with that day. Some of the things I look for could be made into an algorithm, but not all. I rely of my sense of what seems weird to me; I normally scan a screen's worth and pick out a single particular questionable item. Sometimes in a half-minute,sometimes slower. (I do not even attempt to go systematically--I find what I find; there's too much to see everything.) Now, perhaps it is intended that the expanded list version displayed will be collapsable, but unless the default view is as compact as the present one, I'll have no use for it. The best thing you could do to help me patrol is to maintain the present interface exactly as it is--as an alternate. I say as an alternate, because others probably would do better with a different display. But for me, there is no change in the display that would be an improvement. Two, you seem to have defined New Pages as pages that have not yet been patrolled. That's of little value to me. What I am really looking for is errors in the pages that have been patrolled. I'm not actually patrolling new pages; I'm patrolling the patrollers--both at the explicitly patrolled pages and the autopatrolled ones. Again, maintaining the exact same current interface as an alternate would let me do what I want to do, and what I know how to do. (I assume others think it of some value, whether or not they want to work the way I do).