Feedback on javadoc UI changes

[Jonathan Gibbons]

To give some context for all this work ...

The original motivation was to fix a mostly-internal issue, such that 
without some internal hackery and limitations, when clicking on a link, 
especially on a link to earlier in the page, the target of the link 
could end up "under" (obscured by) the floating header, leading to 
confusion and bug reports.

The idea behind the initial solution was to move the floating header out 
of the way when it was not in use, so that we could do away with the 
internal hackery and limitations.  But as the work progressed, Hannes 
found a different way to fix the original problem, using some newer CSS 
constructs, without needing to animate the top header out of the way.  
So, this allows us to separate the design issues, to get rid of the 
internal hackery and related limitations (which we definitely want to 
do), and the optional ability to move the header out of the way, which 
might be useful on smaller screens, but which does not otherwise seem 
very popular.

I also note some of the comments in this thread about bugs and 
less-than-optimal focus handling for the search box. Those are different 
issues, and which can/should be handled separately.

-- Jon


On 10/11/2019 02:09 PM, John Rose wrote:
> On Oct 11, 2019, at 8:08 AM, Hannes Wallnöfer 
> <hannes.wallnoefer at oracle.com <mailto:hannes.wallnoefer at oracle.com>> 
> wrote:
>>
>> Hi Stuart, thanks for addressing these finer usability issues.
>>
>> Yes, the animated transitions can feel weird when triggered by a 
>> small scrolling movement. The behaviour that you prefer is called 
>> sticky positioning. I did try to use it, but wasn’t able to combine 
>> it with the requirement of making the header appear on demand (it is 
>> possible but requires quite a bit of DOM trickery).
>>
>> The take-away message is that there are people who have a strong 
>> (negative) reaction to this kind of animated transition.
>
> FWIW I had the same reactions as Stuart.  It felt like the top of my 
> static webpage was floating away from my (virtual) fingertips.
> I tried to like it but the floatiness didn’t feel like it gave any 
> advantage.
>
> Also the thing Stuart describe about the top-left mouse control 
> “clicking itself” happened to me a couple of times.
> My interactions with that self-clicking thingy were almost always 
> non-decisive.  Again, control was floating away…
>
>

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://mail.openjdk.java.net/pipermail/javadoc-dev/attachments/20191014/144d8d7d/attachment.html>