So, I’m tentative to announce this project, as the server it’s running on is a bit of a potato which will probably fall over pretty quick if it gains any traction, but…
Introducing RSS Temple!
If you’re interested in a free RSS reader which attempts to mimic some of the more useful features of the big players (Feedly and Google Reader, in particular), including full-text search, hotkey navigation, small footprint interface, and sharing to both Lemmy and Mastodon (among others), I would love if you gave RSS Temple a try.
I’ve been working on this project for ~7 years now, and I alone cannot find any more bugs or usability issues, so I hope it’s ready for the community to see. Any feedback is appreciated!
The code is open at:
https://github.com/murrple-1/rss_temple (server, Python)
https://github.com/murrple-1/rss_temple_ui (landing page and web app, Angular and EleventyJS)
https://github.com/murrple-1/ansible-collection-rss-temple (Ansible scripts to deploy one’s own instance)
I wanted to check this out but I did not find any documentation in how to set this up. The readme files are mostly empty.
As in, you’d like to run your own instance? Fair, I didn’t expect that would be the first thing people would want to do, but I can make some better README docs for self-hosting
No, I did not see how to see any of the UI. What am I missing?
Ah, I see it now. rsstemple.com. I missed that earlier.
Wow, that captcha is impossible.
Ugh, first impression matter immensely, and I dropped that ball, eh? :(
For what it’s worth, I’ve set the captcha library back to its default settings, so hopefully the defaults are less obnoxious. The library I’m using is pretty popular, so I was guessing that bots would have fully trained their way around it by now, and so I chose some slightly less obvious fonts to put up a tiny bit of a barrier, but ya know, the balance of “legible for humans” and “hard for bots” is very hard. And I am no expert on captcha design
Yeah, the fact is today that the more advanced bots are probably better at captchas than humans. There isn’t really a great solution here. But I tried at least 10 times before I got one right.
Aye, I’ll stick with the defaults for now, at least until I have a bot problem (which is in some ways a good problem to have 😋)
I tried to upload a opml file and got a 400 response.
{“detail”:“JSON parse error - Expected object or value”}
I don’t suppose you would be able to share the OPML file with me?
One thing that I noticed is that reading items is very difficult because basically as soon as they appear on the screen they are marked as read and become light-grey on white. Either the “read” style needs to be more readable or the marking of read should be deferred until I am done reading the article.
Noted, and I see what you’re saying. The text needs higher contrast in the read state
- I would expand the ‘No Category’ category by default or better yet make an option to not use categories at all.
- The unread article indicator gets pushed off to the right if the title is too long. I’d make it so it sits at the right next to the main window and does not move.
Both issues have been noted, and I will investigate them ASAP
Sounds neat! I’ll check it out later!
I’m basically tied to newsboat now (and I’m not complaining), but something where I don’t have to hack it up with git would be nice, I suppose!
Might want to fill out your privacy policy :)
Damn, I hoped no one would notice that yet.
Informally, I am saving the bare minimum - even IP addresses are transient in my logs, no tracking, and profile deletion is total and immediate, but I don’t think that’s “legalese” enough
I would also move the currently selected article to the top of the page. Right now the article you are reading can jump around a little as you hit the J key to select the next one to read.
