26

I am glad to notice that the massively annoying refusal of the 10k flag counter to disappear when I'm not needed has been appropriately dealt with via the extreme decapitation method of simply eliminating the 10k flag queue. As of a few days ago, the orange-ish counter on the top bar now links to the /review page and displays some measure of the pending review tasks on the site. Overall I find this to be a Good Thing, but in its present state it also has a massively annoying feature.

  • The counter can be nonzero in situations where there are no actionable review tasks.

enter image description here

This came up recently in Physics, and it turns out it is . The rationale, as Shog9 explained there, is that even if I have completed all the review tasks that are actionable by me at a specific point in time, I should still be worried if there is a large number of review tasks to be done (e.g.: ahem).

Now this is a patently reasonable thing to say, but it confuses two aspects of the user-interface utility of this counter, and this confusion is much to its detriment. The counter is primarily a call to action, and this new feature adds an informative aspect. In other words, it is now acting as both a dial and a warning light.

This is to the detriment of the UI feature, because if the big bright orange number does not reflect the fact that there are actionable tasks for me, I am much less likely to jump in whenever it says there are things to do. A warning light is not useful if it is turned on all the time. This actively diminishes the utility of the counter. Fortunately, though, this has an easy* solution:

  • Dim down the color of the counter, preferably to some shade of gray, when there are no more actionable posts for me.

Once you do that, the counter retains its 'dial', informative usefulness, and it regains its 'warning light', call-to-action usefulness.

*Yeah, yeah, I know, performance and so on. Work something out, cache if you need to. It's worth it.

Improve this question
3
  • From Anna Lear's response to the original feature request: It is also cached, so may or may not be very accurate depending on the amount of ongoing activity in /review at any given time.
    – Martijn Pieters
    May 5, 2014 at 16:25
  • I beat you to it on this feature request! I may consider removing or reducing the overlapping part of mine though, as I made other alternative suggestions.
    – Nick Stauner
    May 5, 2014 at 16:45
  • 1
    @NickStauner Using the top-bar tag would have helped getting picked up in searches. Or maybe there's a duplicate-tag problem there.
    – E.P.
    May 5, 2014 at 18:24

1 Answer 1

Reset to default
3

As Shog9 states in a comment on the original feature request

Super-expensive to personalize this - so the alternative is simply not having an indicator.

Because the indicator in the top bar is used on all pages it is very expensive to personalise it and is heavily cached, by necessity. As Anna Lear states:

It is also cached, so may or may not be very accurate depending on the amount of ongoing activity in /review at any given time.

As a result, I don't think we are going to get even a dimmed version, as that'd require querying on every single page if you have any items to review.

Improve this answer
5
  • So the problem is programming a check of whether there are any review tasks I can complete at all, not how many, right? I may have to update my feature request if so.
    – Nick Stauner
    May 5, 2014 at 16:48
  • 1
    I am speculating based on my own experience as a web dev, I don't have any insider knowledge of the Stack Exchange stack here. But to keep that number current per user would require a massive pub-sub cache setup that's not worth the effort, I'd say; if someone else reviews suggested edits before you, then that affects your personal 'outstanding reviews' count as well. You can't go query that for every page you render, or with caching, every N pages. You reserve that cost for just the /review page when someone actually comes and visits.
    – Martijn Pieters
    May 5, 2014 at 16:58
  • 2
    Now, I'm not even close to a web developer, but it occurs to me that below 10k, the indicator works very well as a warning light and above 10k it works well as a dial. Is it truly so expensive to attach the below-10k procedure to the colour of the indicator and use the current above-10k procedure for the number? I have less than 10k and I generally don't see the indicator. When there is something for me to review, I see it. If you can do that (with numbers too) for me, why is it massively harder to continue that for 10k+'s colour?
    – Jim
    Jun 30, 2014 at 14:07
  • Because is easy to generate two or three numbers and cache those. Creating individual numbers for everyone is a whole different kettle of fish.
    – Martijn Pieters
    Jun 30, 2014 at 14:18
  • 1
    Maybe I'm misunderstanding both @Jim and \@MartijnPieters here, but why can't we have a finite number of caches, one per reputation tier? It wouldn't be "individualized," yet it'd solve the problem of people feeling like there's always a count of review items they can do nothing about.
    – Andrew Cheong
    Jul 1, 2014 at 21:26

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .