FreshRSS on the Synology 🔗
I can make do with just about any RSS service provided it:
- Has a clean web UI
- Allows third-party clients (e.g. Reeder, Unread)
- Has some sort of provisioning for filtering.
There are plenty that do the first two. That last requirement is a little harder, and it’s why I’ve largely stuck with Feedly despite a small break, when their management team demonstrated some bad ideas they eventually walked back.
But I’ve been on this self-hosting kick, and I learned about FreshRSS while I was poking around with PikaPods, so after a quick OPML export of my Feedly list and a little experimentation I’ve moved FreshRSS into the Synology:
- Clean web UI with keyboard shortcuts
- Extensions framework
- Sharing configuration
- Works with Reeder
The filtering works a couple of ways:
You can set up a “mark as read” filter with a simple syntax that includes operators like inurl, intitle, and keyword. That will handle new articles as they come in. That’s the best option if you’re using a client.
You can also use those operators in the web UI’s search field then save them and apply them to current articles just to clean up the view. That’s okay for cleaning up your current view before your filters start working on new stuff, or if you want to be able to create a one-click way to narrow a feed.
And your saved queries can also be referenced with the search:QUERYNAME operator, so if you model one out and like it you can make it more proactive.
You can guess what I don’t care to read about from The Oregonian based on my initial list:
inurl:/advice/
inurl:/ducks/
inurl:/betting/
inurl:/nfl/
inurl:/beavers/
Give it time and it will grow. I’ve only been using FreshRSS for a couple of days, so I haven’t had time to be exposed to everything, like the many, many, many articles that landed in the /casino/ section between starting this draft over breakfast and coming back to wind things down after dinner, making wish The O would do like other newspapers do and segment its RSS feeds at least into the major sections.
As it is, they don’t do that, and have a wildly flat and very granular IA, so you can’t just filter out inurl:/sports/ in one go, but at least they have reliable category URLs. At some point I will have to do the work of filtering out deal posts from all the tech sites, who are not at all incentivized to make it easier to avoid their affiliate link farming. Feedly has some sort of “AI” that creates meta-categories to filter on, so I could just tell it “no deal posts” and clean up a few edge cases.
I did look at a few others. TinyTinyRSS was a prime candidate, and even has a native Synology package. Performance wasn’t great, though. A few others had no filtering.
Kitty and the Synology 🔗
Also on the topic of the Synology, if anyone has gotten kitty’s advanced features working with one I would love a guide or insight into how. The environment is just a tad too locked down and non-standard, and a recent skim down a GitHub issue suggested that there’s not a lot of hope on that front. You can get as far as the Syno understanding that there’s such a $TERM as xterm-kitty, but the stuff like kitten ssh has been a lost cause. Or rather, I am was running out of patience with it because there’s a lot of magic on the happy path and then it gets pretty manual if you have to go off-road, and this is a Linux-based appliance we are trying to get this thing wedged onto, not some plain old box.
I changed that to was because I found this thread and it helped me figure out the small step I needed to take to get kitty working with my root account on the Syno.
And what a thread. It ends with the maintainer saying:
In any case, as I have explained before, if you want to perpetually stagnate the terminal ecosystem by insisting every terminal be xterm (the biggest boat anchor on the entire ecosystem), please just use xterm. kitty is not for you.
I kinda get where he’s coming from by the time the whole thing is over: Open source project owners are all over the map in terms of their approach to community interactions and management. Just yesterday I saw one who pleaded with a user who’d decided to just go use another app to share some sort of exit feedback. People in this thread were doing the whole “if this isn’t addressed I will try something else” thing that I imagine gets pretty grating if you’re just some person making a thing and giving it away and are over whatever things in the ecosystem you are over. Like, there are a lot of projects devoted to strict compliance with What Has Come Before, and I don’t think the *nix world has any shortage of adherence to What Has Come Before.
Anyhow, I usually don’t do anything particularly complex over on the Syno, but I was having issues with tab completion, backspace, and command history going screwy whenever I was operating as root so it seemed worth fixing to just keep using kitty and not remember to use another terminal app for this particular hobby.
E-readers vs. Tablets 🔗
Jason Snell on eink e-readers vs. tablets.
I’ve been through the “maybe I want to just consolidate down into an iPad mini” thing with my assorted e-ink devices, and I always come back to “the reading experience is worse on a tablet.” I’ve been back and forth on which ecosystem, too, and will stick with Kobo, less as a features thing and more as a “it’s hard to be rid of Amazon but I can at least be less into Amazon” thing.
The not-Nook/Kobo/Amazon space is interesting. I don’t know if I am up to being that much of a resister to Big Ebook. I have a little Onyx Boox Palma and super wanted to like it, but it’s just an e-ink Android tablet about the size of a “Max”-sized iPhone, and the e-ink/Android app experience with Kobo wasn’t very good. There are other reading apps for it, but I don’t know how location sync works in that ecosystem. Now that I have calibre-web set up I could get my books into anything, but I’m being way too much of a perfectionist on questions like “but what if I leave my e-reader behind and just want to use my phone for the train ride downtown” or whatever.
On the flipside, there are some very good e-reader apps for Android that can talk to calibre-web in OPDS mode. You don’t get the location sync stuff – it’s just a way to browse and retrieve your books – but if you can upload an epub to calibre-web it can come back down to a bunch of these apps, some of which do have for-pay location sync backends as an app feature.
Kinda academic right now: I have a nice and relatively new Kobo, I have a slightly more worn out but functional old Kobo, and I have them talking to my calibre-web install plus access to the Kobo store, Overdrive, and my Pocket articles. I don’t need to go out and buy another e-reader just to be a little more indy.