blap With the way that Flarum's extensions work right now, a completely Composer-less installer would exclude the possibility of being able to add additional functionality to Flarum by way of extensions. Taking a look at Pockethold, for example, a community-provided installer for those without SSH access, even it includes composer and attempts to run it via the php application instead of via the terminal user. The completed Flarum it installs also includes Extiverse Bazaar (available for anyone, of course) which runs Composer in a similar manner in order to facilitate the installs. If you'd prefer this method of installation, it's pretty close to the single download method of other software, too.
Composer itself really isn't that extensive, it's literally one file and it does not place extra working folders or files in the system. When it's used to manage a project's dependencies, it only works in that project's folder, it does not stray outside of it. It's an extremely light system compared to something like NPM (which Flarum only needs, btw, for developing, if you are just running as an end-user you do not require NPM at all). Our developers have gone back and forth on keeping Composer as a requirement, and while Flarum is in beta, this is the system that best supports Flarum's consistent development.
You'll note that on our roadmap, Flarum is aiming to address the installation experience for 0.2, its first major version after stable (0.1). What form that takes, we don't yet know. Franz has done some work on simplifying the installation procedure in the past, so I expect that his experience there will come in handy when Flarum looks to tackle its future after 0.1. An idea similar to a cron job would be to include a composer.lock file in the Flarum repository upon release, which would fix the dependencies in a similar manner to a zip file.
I hope the use of Composer doesn't stop you from trying out Flarum. We'd love it if you gave it a try anyway, and could help give us some further critique as an end-user of the product, too. All of this is valuable, and these kinds of discussions only strengthen the Flarum team's resolve to get to a stable version of the product so we can focus on a better experience for users like this.
