Timeline for Let's get rid of the 10K flag queue
Current License: CC BY-SA 4.0
23 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Jan 18, 2021 at 11:45 | history | edited | CommunityBot | replaced http://blog.stackoverflow.com with https://blog.stackoverflow.com | |
| Apr 13, 2017 at 12:48 | history | edited | CommunityBot | replaced http://workplace.stackexchange.com/ with https://workplace.stackexchange.com/ | |
| Apr 13, 2017 at 12:47 | history | edited | CommunityBot | replaced http://hermeneutics.stackexchange.com/ with https://hermeneutics.stackexchange.com/ | |
| Mar 20, 2017 at 10:30 | history | edited | CommunityBot | replaced http://meta.stackexchange.com/ with https://meta.stackexchange.com/ | |
| Apr 24, 2014 at 14:02 | history | edited | CommunityBot | Migration of MSO links to MSE links | |
| Apr 24, 2014 at 14:01 | history | edited | CommunityBot | Migration of MSO links to MSE links | |
| Mar 22, 2014 at 0:01 | comment | added | Gilles 'SO- stop being evil' | I'm surprised that you find the VLQ queue to be a waste of time and the 10k flag queue useful. I have the opposite experience, both on SO and smaller site: the 10k queue is full of stuff that I can't remove from the queue, so I'm wasting my time, whereas the VLQ queue lets me participate in dealing with the item. | |
| Mar 20, 2014 at 21:55 | comment | added | Shog9 StaffMod | 20 votes from /tools on The Workplace in the past week. 3 of them yours. Only 4 users involved total. Probably worth noting I'm pulling all these numbers straight out of the logs, so there's no "debouncing" going on - if one of those voters happens to vote-unvote-vote-unvote-vote, that actually gets counted as 5. And yes, folks do that. | |
| Mar 20, 2014 at 21:53 | comment | added | Monica Cellio | Interesting data. (So I'm, like, 5-10% of the network-wide vote-from-flag-queue hits? Wow.) SO is kind of a different beast from other sites; per-site breakdowns of actions from the 10k and LQ queues for the sites that are reporting problems might be interesting. (I wonder what this looks like on Workplace.) | |
| Mar 20, 2014 at 21:47 | comment | added | Shog9 StaffMod | To put that in perspective, I tracked 114 votes from Low Quality reviewers clicking through to the question page, just on Stack Overflow in just the past 22 hours. Really don't think we're losing anything here. | |
| Mar 20, 2014 at 21:42 | comment | added | Shog9 StaffMod | That would be terrible UX, and rather unfair. We did mock this up back when we were designing these queues, but discarded the idea. Something just occurred to me though - not all "votes" are actually votes. So I went and broke down those 896 votes by vote type: 629 delete votes, 184 down votes, 68 up votes, 9 favorites (yes, really), 2 undelete votes, and... The rest I'm not sure; possibly someone screwing around. Anyway, that's 252 actual votes over 7 days across the entire network. | |
| Mar 20, 2014 at 21:33 | comment | added | Monica Cellio | Huh, interesting. How did your answer over there work out in actual use? (And while this would probably be terrible UX, I'll just throw this out there: for the LQ review queue it's sufficient to enable downvotes. I guess if that were one of the orange buttons, alongside "edit" and "looks good" etc, rather than up/down arrows, that might solve the UX problem.) | |
| Mar 20, 2014 at 21:28 | comment | added | Shog9 StaffMod | The cost of combining multiple reviews and voting - we could possibly 'fix' this by not making votes actually "count" as a review, but somehow I don't think that would satisfy anyone either. | |
| Mar 20, 2014 at 21:25 | comment | added | Monica Cellio | I don't want to keep the 10k tool; see my first sentence. I want to keep the ability to vote and flag in the replacement queue. In the LQ review queue you're already showing the post (and IIRC its score); I'm asking for the vote buttons and the "flag" link. I like your proposal; it just doesn't go quite far enough. | |
| Mar 20, 2014 at 21:18 | comment | added | Shog9 StaffMod | Yes, extremely expensive. I'll elaborate more on this when I'm able to, but the short of it is that this page has been slowing or blocking badly-needed fixes to the (much more heavily used) moderator flag queue. So we can invest serious time in fixing a badly broken tool that sees relatively little use, or get rid of it and fix a critical tool. (slightly less than twice as many people using /tools/flagged as there are moderators, many times fewer actions taken as a result - it is effectively dead weight) | |
| Mar 20, 2014 at 21:17 | history | edited | gnat | CC BY-SA 3.0 | pearls += http://blog.stackoverflow.com/2011/06/optimizing-for-pearls-not-sand/ "'answers are the real unit of work in any Q&A system..'" |
| Mar 20, 2014 at 21:12 | comment | added | Monica Cellio | So is providing that affordance for those who do (and those who we might be able to encourage more, as we build community) expensive? | |
| Mar 20, 2014 at 21:08 | comment | added | Shog9 StaffMod | Maybe not to you, but it's still another click - and the vast majority of folks using the queue don't click. | |
| Mar 20, 2014 at 21:07 | comment | added | Monica Cellio | Oh right; should have realized that. The AJAX load doesn't feel like a new page load (even if it technically is), so at least to me the click to add some inline context is very different from a click to go to a new page, @Shog. | |
| Mar 20, 2014 at 21:04 | comment | added | Shog9 StaffMod | The flag queue is the only place where you can cast votes, @Monica - you have to click through to another page elsewhere. Technically, even voting from /tools/flagged is another page load (and the same # of clicks as it would be from /review), but since it's an AJAX load the referer is still /tools. | |
| Mar 20, 2014 at 21:01 | comment | added | Monica Cellio | That's up to 896 more votes than would have been cast otherwise. (We don't know how many would have been cast through other means, of course.) When you talk about flags from /tools do you mean all tools (including stats, close, and delete), or just the flag queue? | |
| Mar 20, 2014 at 20:57 | comment | added | Shog9 StaffMod | In theory, yes: the current flag queue provides a lot more in the way of tooling. In practice, it's still at least a click away (you have to expand the post to access it) and almost no one uses it. In the past 7 days, only 896 votes have been cast from the flag queue - that's across the entire network. If I exclude Stack Overflow, we're down to 322 votes. To put this in perspective, 24826 flags were raised from /tools during the same time period. Same problem as the old review system: lots of tools, no users. | |
| Mar 20, 2014 at 20:21 | history | answered | Monica Cellio | CC BY-SA 3.0 |