4 ms·
You're scattergunning. What exactly about the 300ms tap delay removal do you disagree with?
by jaffathecake 13y ago
You're scattergunning. What exactly about the 300ms tap delay removal do you disagree with?
- jblow 13y agoI disagree with the fact that it happens inconsistently and creates a more-complicated situation for me to internalize; and even if I know the rules (which most users don't), it may not always be clear which are happening unless I engage in experimentation -- can I tell just by looking which pages are going to have delay and which aren't? Sometimes it will be obvious, but will it always? Of course not. So sometimes I have to tap before I know. And that is if I am an extremely experienced and knowledgeable user, which most people are not. For most people this will be just one more small piece of inconsistency in the pile of confusion that is Android. Removing the delay 100% of the time, always, would be a big win.
- jblow 13y ago... And what about the actual action of double-tapping? On some pages it zooms, and on some pages it does nothing, and in fact the first tap fires off a link. How is that simple, predictable, or solid-feeling? It isn't.
- untog 13y agoSo you're objecting to the idea of non-scalable mobile web pages? Because the behaviour all stems from that.
- jaffathecake 13y agoThis is why we have the beta, we're interested in cases where the user expects double-tap to work, but it doesn't. However, with non-zoomable pages, this isn't a new problem at all.
- KerrickStaley 13y agoI agree (a) inconsistency is bad and (b) latency is bad, but double-click-to-zoom is very useful (for operating the phone with your thumb when your other hand isn't free), and I think removing it would be bad overall. Perhaps people will be more inclined to write mobile-optimized sites that don't need to be zoomed if they can get reduced latency as a bonus.
- jaffathecake 13y agoDouble-click-to-zoom is useful, but not on mobile-optimised sites, which is why we only disabled it there. Although, if you know of a mobile-optimised site it's useful on, let us know, it'll affect how the feature lands in stable.
- tjoff 13y agoYes it is useful on mobile-"optimized"-sites. The fact that mobile sites somehow assume that zooming is bad is the exact reason for why they are useless and most people, that know what they are doing, are not browsing mobile sites at all but going to great lengths to get the full blown version that lets you get an overview and zoom in on the interesting parts instead of endless scrolling to find what you are looking for. I've yet to experience a mobile site I prefer to the real thing, on any of my mobile devices.
- krzysz00 13y agoDisabling zoom is also not good for visually impaired users and other people who need things to be bigger sometimes.
- girvo 13y agoLol. "Most people that know what they're doing", don't pull up the desktop version, because they're pretty okay with mobile sites. But I guess I've no idea what I'm doing when browsing on my iPhone, and my analytics showing far better engagement with a mobile site vs desktop site on mobile just shows me a lot of people who don't "know what they're doing".
- jaffathecake 13y agoYou're talking about badly built sites here. The change to Chrome discussed in the article doesn't make the situation you describe any better or worse.
- tjoff 13y agoI'm talking about mobile-optimized sites, but yes, I guess you could say that all mobile-versions are badly built. Which also makes all browsers that allow a page to disable zoom badly built.
- jaffathecake 13y agoThis assumes that if we leave things as they were before, every interaction takes the same amount of time. Because of the network, and JavaScript processing beyond our control, this is not true. Not true on any mobile device.