Eine weitere Update-Runde für #kontra:

1. Nachdem @woxx@www.woxx.lu nun ein Update ihrer Seite vollzogen haben und die unterschiedlichen Kategorien nun besser ersichtlich sind, wurde die in Kontra eingepflegten Feeds auf eben diese beschränkt um Dopplungen zu vermeiden.

2. Das #SurplusMagazin hat seit einiger Zeit ebenfalls einzelne #RSS Feeds für unterschiedliche Themengruppen. Der in Kontra hinterlegte Gesamtfeed wurde also durch die Themen-Feeds ersetzt.

3. Außerdem neu sind Feeds des @volksverpetzer@digitalcourage.social und des österreichischen #Falter

Bereits seit mehreren Wochen gibt es übrigens auch ein Preview für ausgewählte Feeds, sodass man vorab schon einmal sehen kann, wie ein Gesamtfeed aussehen könnte, wenn man ausgewählte Nachrichtenquellen in seinen #NewsAggregator einlädt.

Für die nähere Zukunft sind noch Maintenance-Verbesserungen geplant, beispielsweise automatisierte Tests, ob die in Kontra hinterlegten RSS-URLs noch funktionsfähig sind. Kritik und Anregungen wie immer gerne per Mail, wie im Git-Repository beschrieben.

Version 1.3.11ist ab sofort online unter kontra.jayvii.de.

PS: Einen Träger / eine Partnerorganisation für das Projekt suche ich leider noch immer. Gerne teilen #pleaseboost.

#Nachrichten #Medien #depol #FOSS #OpenSource #PHP #WebDev

@toddsundsted thank you! It is actually quite minimal #PHP. There is a single config file (JSON) in the background, which the PHP scripts generate the sites from. This makes it quite flexible, actually. If I get around to it, I might extract a little more of the hardcoded stuff (mostly text) to config files, which would help to re-purpose the whole thing for other uses (or other countries).

Since today, it even has a bare bones preview-mode which fetches the selected RSS feeds and shows them as simple HTML, before one decides to export them to OPML.

The largest chunk of work was to actually gather all the RSS Feeds... Unfortunately, the European media landscape is quite scarce, when it comes to leftist or progressive media. Could be worse, but should be a lot better IMO

It has been a while since I worked on #serci, but here is a new feature, I've been thinking about for a while but didn't have time to implement it until today: Extended options for search engines.

https://src.jayvii.de/pub/serci/commit/4a990c43d256913923eed1acc1c206e470b54df2.html

This is actually quite a small (but IMO useful) feature that let's serci parse additional text to the search engine, besides the search term. Or, as another way to think about it, it gives the user more options to tell serĉi your intent. Options are added to the keyword with the ":" separator.

Say, you want to search explicitly on the German version of Wikipedia. You can do that now with !wp:de

Or you want to translate from German to Italian with DeepL. Just use !deepl:de/it

#webdev #PHP #searchengine #selfhosted #foss

I did some basic explanations and screenshot-showcase of my new bookmarking / note taking #PHP web application #rememori: https://src.jayvii.de/pub/rememori/index.html#features

Suggestions for new features (and coding style...) are very appreciated!

Been working on a simple bookmarking client ("rememori") completely written in #PHP, without any data base in the back, but instead based on simple .txt files.

The features include (for now):

1. tagging (only one per note),
2. a simple search through all your notes,
3. automatic title-fetching for URL-notes,
4. a simple API endpoint for programmatically adding new notes (I imported all my linkding bookmarks via three lines in bash and #curl),
5. #RSS feeds for individual categories (so you can use it as read-it-later application together with an RSS reader

Reworked my tiny #tw2html webtool that let's me check and "follow" streams on #Twitch. Checking is now done client-side via #JavaScript instead of the much slower server-side #PHP procedure. The new procedure checks 30 streams in well under 1 second.

Feel free to try it out here: https://twitch.jayvii.de/

On yesterday's #serci Update post (https://social.jayvii.de/objects/MVKQo4c6huo), today I finished a massive overhaul of the web-frontend for this Meta² search engine.

Besides looking a lot slicker, it should be a lot easier to understand for new users now. The list of search engines can now be searched and filtered and your personal default search engine can be chosen easily with a single click! I also added descriptions for each search engine, to make the list easier to browse and understand.

Feel free to try it yourself! https://search.jayvii.de

Oh, and the re-factoring increased performance by 33%! Searching on average takes 0.12s before redirect (same as before), but loading the web-frontend takes only 0.18s instead of 0.28s now. Really happy with how it turned out. Still: any suggestions are more than welcome!

#webdev #PHP #webdesign #searchengine #selfhosted 

I have been adding quite a few more search engines to #serci today, including #Fireball, #SearX/#SearXng, #eTools, #TigerCH, #WolframAlpha and #DBLP today.  All new additions are already live at https://search.jayvii.de.

With 64 search engines in the default configuration, I feel like the webfrontend has to present all the options in a better way. But I am really out of ideas on what that could look like... Suggestions are very welcome! #webdev #php #foss #search

The move to a different VPS actually increased the performance for the HTML-building of #serci substantially from 0.28s above to 0.19s over 100 trials (also, this might have to do with #PHP updates and smarter Apache configuration). Redirection does not differ. This is likely the minimal overhead I can achieve.

Migrating all my tools was quite a hassle, eventhough most of my self-written ones require not much configuring (if at all) and need no databases, etc.

I switched my rss reader from #miniflux to #selfoss, let's see how that turns out. I am also thinking about trying #ttrss again.

I do need to limit server load on this tiny 1-core, 1 GB RAM VPS. But that is part of the fun! Now, in the coming weeks I plan to go over my tools again and improve their performance (especially server-load) substantially. I think there is still room for improvement!

For example, I would love for yt2html to be pure #PHP instead of the old #Rstats script. Also, I would like to write a really simple file-based notes and bookmarking tool in PHP and retire the linkding service I am running, as it is a bit overkill for my uses.