3. Install the Termux api CLI tools via apt. You now have access to clipboard.
4. Install Emacs within termux.
Termux offers an apt-based package management system, and a substantial, though partial, set of Debian-based packages. (592 packages as I write this -- a full Debian archive is >60k packages now.)
There are other tools installable through other package-management archives -- notably python's "pip", which provides tools such as youtube-dl (I've previously mentioned this for offline viewing and listening of YouTube and other video/audio content).
I'm not particularly enamoured of the Android app "ecosystem" on multiple points. Including privacy and quality.
This is a great setup. However, on a small screen phone without a physical keyboard I tend to prefer a touch-based application. I rarely make any edits to my org files, but I often want to have a quick look.
Fair enough, though a larger tablet (8-10") and bluetooth keyboard makes for an excellent mobile platform.
IMO the phone form-factor is too small for any really effective use, and a dedicated dumbphone, or more stationary use of IP-based voice comms is generally a preferred option.
I've almost completely abandoned voice comms, as it happens.
How's the speed? I mean, I use use-package and a bunch of init file optimizations to get Emacs startup time as low as possible, but I can't imagine it being tolerable on my phone. And opening the agenda across my standard set of files would probably be awful.
That is a really funny question. I started using Emacs a relative short time ago on what I still consider a fast machine - 2001 on a dual-processor Pentium Pro at 200MHz with 192MiB of RAM. My phone is one of the lowest-end Android models available today and has a quad-core 1.2 GHz ARM processor and 1GiB of memory. Emacs has not really bloated like other applications over time, it just started out with a lot more feature than workstations in the 1980s (https://en.wikipedia.org/wiki/3M_computer) could handle.
Sad but true, I guess, haha. My Emacs process currently has a RSS of 201 MB, and that's just with my everyday packages loaded. It's amazing to think about the Lisp Machines...
I've only been making occaional use of Emacs (see comments elsewhere -- I've been using vi / vim since the mid-1980s, and have far more muscle-memory there). But I've had no complaints.
I was wondering just a few days ago whether building a nice Mac desktop app or maybe a web app, akin to Ulysses, but for Org file writing/editing, exporting etc. would be a viable idea and whether it could eat some of Markdown's marketshare by not being tied to Emacs, which is why I feel most people don't pick it up.
Would that be something the HN crowd would be interested in?
This will be a tough sell. Org-mode is a fairly fast moving target, due to how easy it is to extend anything in emacs. For lots, you will actually be battling against all of the other extensions in emacs. (For me, as an example, you will have to replace flyspell, beacon, tramp, ace-jump, ido, and ggtags.)
All of this is doable, of course. But it will be a tough sell.
That said, I'd say go for it. Worst case, nobody uses it and you just learn some of the difficulties of implementing. Hard to see that as a failure.
Yeah, it would be fairly difficult to build, which is why getting feedback would be useful, but there's a catch 22; people who already know what Org is are fairly comfortable using Emacs and I don't know how to find out if people not that found of Emacs itself would still find Org useful, or if combining apps like Ulysses and Omni Outliner is good enough for them - the worst part is that because they don't use Emacs, they can't tell me.
This really may be the 'just build it and see' sort of scenario.
I've often about some sort of app that uses Org as a design basis, but stores data in a way that makes it easy to access remotely without having an Org parser, to be easy to sync and query programmatically, and is designed for the modern world in terms of connectivity and interface.
I think that could absolutely kill it.
Imagine writing extensions in js or lua or whatever, and having those extensions have clear DOM-like access to the outline, have full access to their stdlib, have access to built in services like notifications, etc.
Imagine being able to programmatically access and manipulate the outline from wherever: webhook to add an outline node from IFTTT, twilio access your notes? etc. Yes you _could_ do that with Emacs, but IMHO it would take an incredible amount of work to not suck, partly because all the various pieces necessary feel like hacks on top of hacks.
Sort of how Atom and VSCode provide _much_ better interfaces (both code and UI) to your source code, and are starting to make honest inroads against the established players (like Emacs and Vim).
I've been wondering about this, as well, and so have been working on an Emacs configuration that makes Emacs work as close as possible to a "normal" text editor (hopefully allowing learning org-mode without the barrier to entry of Emacs itself):
It's way easier for someone to hack up an acceptable new UI for org-mode (whatever that looks like) using the existing underlying Emacs org-mode code/runtime than it is to hack up a new UI and completely replace all the Emacs code & org-mode.
You can run Emacs headless, it has C extension support now, your Mac app could start it headless and communicate with it over a socket or as a network process. Multiple Emacs modes do this to make Emacs itself talk to other programs (e.g. mu4e <-> mu, gud <-> gdb), there's no reason you can't do it the other way around.
It strikes me that if you want "org-mode lite", you would use Workflowy. This isn't meant as a criticism of Workflowy at all, they do a great job making a lot of the org-mode ideas accessible. But if I'm using org-mode, it's because I want the extensibility of emacs.
Too bad orgzly hasn't solved synchronizing. Notepad for Google Drive is a text editor with no features except it gets sync right and has become my favorite way.
Print out your todo list and carry a pen also works well.
http://www.orgzly.com/
https://github.com/orgzly
https://f-droid.org/wiki/page/com.orgzly
It was a little annoyance we did not have anything as good as this before.