I’ve been looking at self-hosting (on managed servers rather than a Raspberry Pi beneath the stairs) an alternative to Feedbin. I always like the idea of self-hosting, and Feedbin costs $5 a month. It’s a great service, but $5 could get you two or three low-resource apps on Pika Pods, such as Miniflux, FreshRSS, Wallabag and a nice range of other FOSS options.

There’s a real mix in terms of quality – finding a web-based RSS reader that can match Feedbin’s features and usability is nigh on impossible, although I may try Miniflux, especially if I can wrangle its appearance into shape. I’ve used – and liked – Wallabag before, but, unlike Instapaper, it lacks text tweaking options. While I’d rather not host my read-it-later articles with a third party – when you think about it, there’s a lot of personal info in there – Instapaper has a good UI, a free tier and is promising Kobo integration later this summer. That is a killer feature for me – be interesting to see whether they bundle it into the premium tier.

(Recommend Pika Pods, though. Good range of apps and they offer a $5 credit. Even better, they don’t ask for card details upfront, which means you can try it out properly without having to remember to unsubscribe. It’s that sort of thing that makes you well-inclined towards a service.)

I took a tour of a few other apps. Etherpad has a nice minimal editing interface, but I couldn’t figure out how to save a document (or whether you’re even meant to). Readeck is a very slick web clipper which probably required a lot more processing power to import my Instapaper articles. HedgeDoc is worth a spin. Lots of interesting stuff, but nothing that made me think I’d definitely commit to it.

I did like the look of Haven, a sort of Facebook clone. Groups are private and invite only, and don’t talk to each other. That’s a really good idea for groups that want a simple, familiar interface for privately sharing photos and updates and couldn’t care less about ActivityPub.

Unfortunately, Haven suffers from the usual one-developer problem, where their back end skills are obviously good, but they can’t write accessible, semantic HTML that’s easy to style. I didn’t need to discover another web app that doesn’t make it easy to add alt text to images. Still, I find this an interesting enough idea to think about raising issues, or even making a pull request. Can’t be that hard to add a main element to an erb template, can it?