William, responses in-line:
On Fri, Sep 11, 2015 at 9:34 AM William Green
wrote:
I've been following the past few emails and I have a few suggestions
which you might find useful.
Awesome!
With Javascript, you can get the user agent of the device which tells you
what computer their using, what browser version, and more.
Yep, as mentioned I'm using platform.js to handle the OS detection at the
moment. This can be tested on the downloads page:
http://static.gimp.org/downloads/
The text in red is the result of the that bit of js running.
A responsive design is also definitely a good idea. 50% of web users are
on a mobile device of some type, which means that it is likely that half of
the web users will be on a phone of tablet.
Yep! This is exactly why the site was built to be responsive from the
ground up.
For my projects, I use different pages for each device which gives you
more flexibility even though it might be more work.
Not sure what we would gain having a different page depending on device,
mostly.
It offers several advantages in that the actual content can be easily
changed. If you do responsive web design, you are still limited since you
are just changing the styling for the elements and not the content itself.
The content can be easily changed with the infrastructure that's in place
now. Since we don't really have a need to serve different pages for
different devices, styling changes are sufficient to accommodate the
various screen sizes we may encounter.
If you were to 'create' the page on the server, you could make the page as
flexible as you wanted. Just detect the user agent in PHP and then 'piece'
together the page on the server. Each page could use a separate stylesheet
and separate content to allow a better mobile experience. For content and
styles that you wanted to remain constant among all devices, just insert
them with PHP for each one. Hope you understand what I am saying.
Yes, a dynamically generated site. I was specifically avoiding this as I
don't see a compelling reason to have it this way, personally. It also
removes server requirements, as well as minimizes the possible attack
vectors.
It also has the benefit of requiring less server resources, and handles
heavy loads much better as all we have to do is serve the requests.
Another thing that I've been wondering is when I posted my message to
the mailing list, the page said that it would be waiting for moderation. I
posted my message right *before* signing up of the mailing list. Should I
post the message again or should I wait for the moderation process? How
long does the moderation process usually take?
I am not 100% sure how this works with the ML, but I am going to be
forwarding this back on-list - so just reply all to keep the conversation
on the list. :)
pat