People are shooting themselves in the feet with the Universal Installer and our Installation instructions. Just see this thread from yesterday and today: Working Installation Documentation - #7 by zopyx
I posted the links to the installation instructions in docs.plone.org in that thread partly as an experiment. In a slack discussion last week with other community members one remark was 'but we do have installation instructions' . Yes we do, but they're confusing to people.
Jens' response there was the most direct 'solution available at the moment if the platform is not Windows: use the system Python, install dependencies, run minimal buildout, start.
But even Windows is getting there with WSL(2) to support installation from source in a unixy-container.
Over the years, my experience was that the UI was most used by the least experienced. The main target was never developers -- who can usually figure out how to do it themselves -- but rather inexperienced integrators.
Let's talk about the goals and the target group:
We 'the community' need to have a 'political incorrect' discussion about what we want to achieve with Plone in 2020 and beyond. Do we still want to put Plone in the hands of as many people as possible to democratise publishing on the Internet and intranets by giving them a 'works out of the box' Plone Application? Is that still necessary and if people think so: can we find sponsors for that goal?
I'm aware it sounds very elitist and entitled coming from me and other developers/sysadmins/integrators, but maybe we should protect inexperienced integrators from installing Plone and seemingly getting it up and running as a production environment. When in fact it was sheer luck and/or they have no clue what they are doing. And cannot solve basic issues when something or anything goes wrong on that production system. Like restoring the backup that wasn't made.
I'd love to give everybody who wants to their own 'Plone in a black box' with 4 buttons (start, stop, backup, restore) that runs on any possible platform, but we would need significantly more resources for that to still wrap Plone as an 'end user' product.
Also the technical bar has been raised significantly in web/frontend development on knowledge on how to run command in a terminal. 10-12 years ago a UX designer or frontender could still be proud to say they were a brilliant Photoshop wizard and therefore didn't know how to descend into a dark terminal. It's 2020 and you can't do much modern frontend development without a bash/zsh/whatever prompt.
Maybe we should demand at least minimal terminal fluency from our target group to set up a Plone site locally from source and accept that the learning curve involved serves a purpose and is not 'hostile' towards power users anymore in 2020.
There's a really good acceptable alternative nowadays if you can't install a server/service yourself or don't want to spend time tinkering/learning on it. SAAS.