All my rss on one page

About a month ago, i wrote about trying new ways of listing rss channels. I said in that post that i planned to use a static page listing posts from the rss channels i follow, and as of today i got that setup. I’m using a little piece of software called tinyfeed, and it’s not bad.

Over the past month, i have been experimenting with four ways of reading rss: the first is akregator, an rss reader for the kde plasma desktop; the second is fraidycat, a local rss reader that runs as a web page and has a unique design; the third is capy reader, a fairly traditional rss reader for android; and the fourth is sfeed, which is a suite of tools for fetching data from different rss channels and using an intermediary format to export them to different formats. I probably had the best experience with fraidycat, but for just two thing: it significantly increased the power usage of my computer, and it kept getting stuck on channels that were blocked by the great firewall. I don’t know what there is to do about the former problem, except to find some software that replicates the interface but works more efficiently.

The latter problem was unfortunately not unique to fraidycat. Akregator also seemed to have a lot of trouble, and would turn half of my subscriptions red on a regular basis, and make their favicons disappear. Sfeed and capy reader would give up quite quickly and move on which was nice.

Akregator was so close to being really nice but, like a lot of slightly older kde software, it had a few too many rough edges that made things a bit frustrating. I don’t know why every desktop rss reader on linux insists on embedding a web browser, for example, because none of them work very well. As i mentioned, favicons would disappear when i site couldn’t be reached, but also for some sites they simply never showed up at all. Also, you can’t select multiple channels to move them into a folder at the same time; channels have to be moved one by one.

Capy reader is not bad. It is fast and has a lot of quality of life features. Rather annoyingly it still uses a sidebar, which does not play well with the new gesture-based navigation on android, and the sidebar also holds multiple different interface elements which would be better placed in their own separate positions. For example, there are buttons at the bottom of the sidebar which act as filters for the currently selected channel (all / unread / starred). It does not really make sense for this filter which affects the currently seected channel to be in a list with the channels it is possible to select.

But the biggest issue with capy reader is that it is an android app, and i don’t really want to be increasing the time i spend on my phone, even if it is to read nice blog posts. The screen is kind of small for comfortable reading anyway. At this point i realised that it was extremely unlikely that an existing solution would tick all my boxes, so i decided to look at sfeed due to its flexibility.

It’s true, sfeed is flexible. But to be honest, working with really really really long-lined tsv files is not fun. There’s a certain level of jank that something like this has to embody, but the sfeed developers’ decisions as to where to put that jank did not bring me any joy. Rather than messing with it and hoping that i could get something i kind of liked, i though i would for now find some software that was almost good enough, and use that until i have the time and motivation to make something that i actually like.

And as i said at the start of this post, that something is tinyfeed, a little go program that takes a list of rss urls and makes a page of rss entries in the style of one of those nerdy link aggregator sites but without the comments and toxicity. Again, there are some strange design decisions, but it’s easy to get around them so i can’t complain. In particular, the tinyfeed solution to an optional stylesheet is to embed an inline stylesheet as well. This meant that the stylesheet i wrote was getting overridden by some inline styles. Luckily we are able to also define a go template to use to generate the site, so i created one and removed the inline css, and also took the opportunity to change some other parts i didn’t like. I’m very grateful to the developer for giving the end user these possibilities.

At the moment the program is running via cron on a public access unix server. I’m happy with the design and layout. Maybe in a month or so my opinions will have changed but we’ll have to wait and see and in the meantime, i’m going to enjoy looking at everyone’s beautiful websites.