What depresses me most about Jolla is that even now, when they appear to be at death's door, they talk about providing an alternative to iOS and Android. They do not have, never have had, and will never have any chance of doing that. If a third consumer phone OS ever gains a significant market share, it will come from a giant company like Microsoft or Samsung, because a company like that can afford to lose money making phones for many years, until they somehow crack the market. Even if Jolla comes up with some fantastic features for Sailfish, which are so clearly superior to iOS and Android that Jolla phones start selling briskly, those features will quickly appear on Android phones. If Apple can't keep Samsung from poaching its IP, then Jolla certainly can't. For Jolla, trying to crack the general consumer market for phones is futile.

I would suggest that they should try to find a much smaller, but potentially profitable market which is not large enough to interest the big players. I think they should build a device that is targeted at computer professionals. The key differentiator would be that this device is meant to be a second mobile device. The use case would assume that the user has with him a primary phone, iOS, Android or whatever, and the Jolla device is a second mobile device which provides enough utility to be worth having with you even though you have a regular cell phone too. To make it worthwhile to carry two devices, you optimize the Jolla device software to make it an excellent device for managing remote computers. This is the main reason I still carry my N900 around - for my job, I occasionally get phone calls telling me that there are problems with our computers or some software at work. I am "on call" 24 hours a day in this sense. And if I get such a call when I'm at a restaurant, or walking to work, or in a park, I need to be able to interface with remote machines, see what's going on, and fix the problem. The N900 is good for that, and had it been designed explicitly for that purpose, it could have been even better.

If you make a device that is squarely aimed at system manager type individuals, many of them will be able to get their employer to buy one for them. This should help sales. Computers are absolutely ubiquitous today, so there must be at least hundreds of thousands of computer professionals worldwide who could find such a device interesting. These people have money, and their employers have more money, so you could build a fairly high-end device that had a healthy profit margin even at modest sales volumes.

You want to use the "second device" usage case to differentiate it as much as possible from regular cell phones (the primary device). Don't include a camera - the user already has one in his regular cell phone. No accelerometer, magnetometer etc. Don't pointlessly duplicate hardware that the user already has in his primary phone. Instead, put a big battery in the Jolla device, and make sure that the Jolla device can be used as a charger for the primary device. Many manufacturers sell portable chargers, so there must be a market for them. Users will not need a portable charger for their phone if the Jolla device has that capability, and it's a very nice feature for persons who travel a lot. Give the Jolla device two microSD slots, and make them hot swappable. You can consider making the device WiFi-only (assuming tethering with the primary device's WiFi signal), but if you do include phone hardware, make it easy to insert the SIM from the primary phone into the Jolla device. The phone use case is as a backup phone in case the primary phone dies.

Optimize the software for serious computer work, not fun. Forget Facebook, make sure ssh, vnc etc. work very well. Consider dropping Wayland for X11, but if you go with Wayland, make sure it can work as an X11 server. No fancy UI. No animations, minimal compositing etc. Those things just pointlessly piss away your battery. Try to build a phone your niece would *hate*. When you add fancy software features, make sure they are fancy *nerd* software features, like having the phone automatically mount its microSD cards on your home and work computers, via sshfs, when it detects the appropriate WiFi signal.

This approach means you don't have to worry about not having a competative app ecosystem. There's no need for half-assed Android compatibility. No need to sing the old song about "web apps" being as good as native ones now. Your killer "apps" are things like vnc which are already available for linux. You are building a pocket workstation primarily, not a game player.