On Mon, Nov 14, 2016 at 1:34 PM, mwall <[email protected]> wrote:

> thank you for the feedback!
>
> what features/functionality do you consider must-have for the standard
> skin?
>

​It should be useful out-of-the-box on all platforms, and all screen sizes.
The goal is to make sure a user is successful on his/her first try. Once
successful, she/he will be encouraged to start modifying.

​


> what other attributes should the new standard skin have, e.g., "easy for
> new users to understand", "easy to extend", "show all sensor data",
> "illustrate core weewx capabilities"
>

Five years ago, I would have said "simplicity, simplicity, simplicity." It
was better to have an easy-to-understand, but plain, website that could be
easily customized. Now, I'm not so sure. The state-of-the-art has moved on
and so have expectations.

When I look back at what modifications users have made to the Standard
skin, they seem to fall into three areas, of decreasing popularity:

   1. Adding new sensors;
   2. Playing with CSS and the <div> blocks to make a customized layout;
   3. Developing a responsive website.​

​The first is easy: users can just follow the pattern established by the
Standard skin to add new plots or stats.

The second is being attacked by your new skin. By making liberal use of
.inc blocks, you're making it easy to move things around and to attach CSS
to the blocks.

The question is whether we should try #3. That would require jQuery and
something like bootstrap. I think if we structure the logic carefully, we
should be able to factor out most of the UI complexity, so the user need
only add new .inc blocks to extend functionality.

Over time, I would like to offer what you're calling "JavaScript plots,"
but, at this point, I think that's a bridge too far.  I don't think the
tools are up to it yet: D3 is waaaay too hard! Eventually, somebody will
write a good D3 library, but it hasn't happened yet.

-tk

Reply via email to