5 Years with FreshRSS
Time to read
Did you know: Google Reader has been dead longer than it was ever alive?
Actually, we passed that milestone a while ago: being shut down on the first of July 2013, GR lasted just shy of 7 years and 9 months, and it’s now been solidly over a decade since the party ended. In the intervening period a lot of news sites have stopped maintaining their RSS feeds properly, while the old Blogosphere has withered and in many cases been replaced with paid email ‘newsletters’. Still, there are those of us that never got over the format (seriously, you want to get this stuff in your email client?) and had to search for a new place to read our remaining feeds.
Five years ago now I wrote Adventures in the Twilight Years of RSS about my journey to that point—from alternate web services to self-hosting, via an extended digression in an ereader-based system. I had just spun up a FreshRSS server a few months earlier and was still feeling things out.
Well, I’m very happy to say that I’m still using that same system, five whole years later. It’s going great!
Fundamentally, it’s a web-based feed reader (functionally meaning that you can access it, synchronised, between devices) while lacking the subscription restrictions and ads you tend to get on centrally-hosted services. I’ve enjoyed watching development over on the github repository—it’s one of very few pieces of technology I look forward to improvements in anymore—and if only I knew any PHP I’d chip in.
The UI hasn’t changed much in all that time, but why should it? Actual features have certainly improved—including running requests in parallel, more powerful searching, and better extraction of full text—and that’s what we’re here for. The community is fine,1 there are a large number of compatible apps due to supporting two common APIs, and you can follow as many feeds as you like (I have more than 500). Ultimately these are all the features I strictly need, though others are still appreciated like the one that powers my shares feed.
Performance was perfectly fine on the rather oversubscribed linode I used to host it on; now on my home ‘server’ (a now decade-old mini PC under my monitor) it may as well not even be running for all the CPU it seems to use.2 On the client side Firefox tells me that a FreshRSS tab with an article open uses 64MB of RAM to render, which may seem like a lot but it’s only about three times what a page on my entirely non-interactive website needs so I’ll call it a win; the actual page size is in the tens of kilobytes so this is definitely ancient MacBook compatible.
Nothing is perfect of course: I’d like some better tools to manage subscriptions for one, but that’s improved over the years and I’m sure will further. And while there are so many apps, I’ve grown used to the specific ergonomics of the no-longer-developed Android app Readably3—but that’s not the fault of FreshRSS.
There is one extra drawback though to using any self-hosted solution: I want to recommend RSS readers to almost everyone, but a setup like this (while neat) is not something I can suggest to almost anyone. If I had a dollar for every time I’ve had that problem… well, I could pay for a month or so’s hosting, for sure.
Recently though I’ve been using Inoreader for a few things that I follow for work, and which therefore need the mental cordon away from the rest of my feeds. I picked this service pretty much at random, there’s plenty of others out there: Feedbin, Feedly, The Old Reader, etcetera (and of course there are local programs, but not on a work laptop). Inoreader, on a free account, is… fine!4 Absolutely usable in a way I’d be prepared to recommend to someone else. I strongly suspect the others are similar.
This is great, because I’m only more convinced that RSS readers are more important than ever in these unprecedented times.5 People love to scroll, and this is a better scrolling.
It’s also better for a web kept out of the hands of big companies run by what seems to be a crystallising class of (primarily) American oligarchs. In the name of spam prevention, email is controlled by a few big enterprises—being subscription based the same incentives do not apply to RSS, making it vastly easier for independent tools to survive. Blogs backed by RSS are much more portable than “newsletters” backed by emails, and that is going to matter.
I signed off the earlier post with I hope that RSS will continue to linger in its perpetual twilight. Now I need it to do more than that, and so I need us to get it there. We need to encourage writers to include proper feeds, and we need to encourage readers to use them. And we also need to catch up on reading the sites we follow, so I should probably get on that.
This is a problem in open source.↩︎
FWIW the only “migration” I did was to download the OPML backup and uploaded it to the new server. Backing up the SQL database itself is an option, but not necessary, as favourites and other feed-specific settings carry over without it.↩︎
I’m baffled that swiping back and forwards between articles seems to be so rare.↩︎
Between draft and publication of this post they seem to have launched an “Inoreader Intelligence” tool that summarises articles for you, which as an LLM-minimalist I’m not super fond of, but we’ve reached the point in the hype cycle where companies have realised that these things cost money and so it doesn’t seem to affect the free tier—save for a large and useless button at the top of every article. If I decide it’s too annoying I’ll just switch to a different reader.↩︎
No, not the These Unprecedented Times of five years ago, these ones.↩︎