After some digging I did indeed find what I believe is the article. As mentioned I read about in American Scientist (IIRC), but based on the abstract this must be it. From the abstract:
By the time of the test, children of all ages had acquired most of the vocabulary necessary to describe the target event. Despite this, they did not translate preverbal aspects of their memory into language during the test. In no instance did a child verbally report information about the event that was not part of his or her productive vocabulary at the time of encoding. We conclude that language development plays a pivotal role in childhood amnesia.
I also found this follow-up study by someone else[2], again from the abstract:
This research suggests that children cannot independently translate preverbal memories into words even with extensive task support. Therefore, language acquisition may indeed play an important role in the offset of childhood amnesia.
However this[3] recent paper argues that this effect might only be relevant for strategic recall, where the kid is asked to recall the memory, but not spontaneous recall which happens involuntarily.
Of course, I imagine spontaneous recall isn't very useful for recalling what you've studied in your home country...
So you want to learn electronics? I suggest that you take the time to actually learn the theory behind it and learn how resistance, capacitance, inductance affects the system. It is hard but helps a lot. If you want to understand digital systems, then learn how digital basic components work and how you combine them to create more complex systems. Had electronics as my major in my bachelor's degree, and the mental model is really important when it comes to electronics, and the only way to really understand it and get the correct mental model is by repeatedly analyzing the networks and the change over time (resistance,capacitance,inductance) and all of sudden it makes perfectly sense!
Underlying theory is important, but seeing practical examples that _explicitly applies theory_ has been critical in my particular electronics adventure. There are more ways to apply Ohm's Law (or Kirchhoff's Laws) than you can shake a stick at and seeing someone explicitly do the math and apply it while explaining a circuit that I am interested in has helped me get better at applying it.
I think the hardest things for me so far been (a) getting my head around the idea that everything in a circuit is happening all at once (rather than iteratively like an algorithm) and (b) "input" and "output" are convenience terms, e.g. an opamp can sink current through its "output". Both of these insights have come from seeing real circuits analyzed on YouTube.
Funny thing, I did the same thing (used Jellyfin for several years, then Navidrome for probably a year) and now moved back to Spotify. I think the discovery of music in self-hosted apps is worse. I like to listen albums mostly but I always happened to listen to the same handful of albums, only because those were the first on my mind. I think Navidrome's smart playlists have potential though. Did you come up with a good way to rotate the music?
I am fond of setting my "album" tab in Sonixd/Finamp to "random" sort. I just refresh or scroll until I see something interesting.
I also use last.fm (still) for recommendations. I've been scrobbling there for over a decade and I still get decent recommendations. Add in recommendations from friends and I far prefer the experience to algorithmic discovery.
If Spotify invested a bit in client UI improvements like making album listening easier (they push playlists far too much for my liking) and fixing the buried "offline" and "private listening" switches, I could be tempted back. Maybe.
Out of curiosity, did you manage to implement auto crawler that can always find the correct html element for any site and remove rest? Or did you have to create per-site rules for it?
I needed a tool for searching emails and was unhappy with existing solutions. I found out about Meilisearch and realized it would be the perfect engine for emails. So I wrote a tool to index mails (from local Mbox, useful with Thunderbird) and query them in CLI. This is initial release for the tool Meilindex. Search Gui is very simple. I have been using this tool for a couple of months now and I'm very happy with it. Check out Meilisearch for more info on the engine as well!
I too have created command line terminal application specifically for this. My solution is to be able to add any metadata to bookmarks with full text search. No release yet, there are bugs but it's perfectly usable (have been using it for months now). https://github.com/tryffel/bookmarker
I'm excited to see what's the Batch api going to look like. I have a few projects that definitely need it to make operations atomic.
Also, readonly/hidden fields are a great addition!