It mentions uv at the end and rye at first (which use uv internally).

( https://en.wikipedia.org/wiki/Dining_philosophers_problem )

* A social graph is built from various sources <-- this is where I wonder what the sources are, what is extracted?

* If I'm connected to Bob, I will see content related to his interests : It appears that the system tries to pick video channels related to me, but the video selection is skewed towards Bob's topics. It seems that Bob's topic embeddings are mapped to the text embeddings of video thumbnails (same for the video titles but with less weight). Since the context is small, sometimes it's off.

* If you refresh the page, it may return to your original recommendations, unless new data from external sources is fed into the algorithm. What's scare me: I wonder if the feed is in near real time. If this is the case, that's explain why my recommendation are as usual and sometimes completely off track for me but related to some people I know.

* It might buffer some topics : if you connect again with Bob but there are no new topics, it gives back the old topic related to Bob.

Once again, this is a pill of a lot of gueses.

---

I really wish to go back to the previous algorithm which one of the reason I subscribed to YT premium, now it's just a minefield.

Synology Lost the Plot with Hard Drive Locking Move (servethehome.com) https://news.ycombinator.com/item?id=43734706

Now, I strongly feel it's based on a wrongly inferred social graph. It recommends videos according to what's happen in the group. It's really unsettling.

I wish there were a setting to disable this "feature." and to stick on my history.

My hypothesis

* The algorithms have linked my account to some others.

* They then serve me the embeddings extracted from those profiles. The near-real-time nature of this has crossed my mind more than once.

It's really unsettling, and afterwards I feel uneasy about any recommendations (all Google services, Netflix seems problematic too, not Amazon).

YouTube seems to have some hidden knobs for tuning this behaviour: after multiple negative feedbacks, the problematic content disappeared from my front page. However, the recommendations on the right-hand side of individual videos remain problematic, and the automatic playlists of YouTube Music are still strangely disturbing (even after multiple negative feedbacks).

However since few months, all recommendations became off track. Then I started to realize these recommendations were related to the life of people I know. At least most of them, some subjects are private and personal, so this is really unsettling and I can't say if the algorithm is halucinating or not (I want to know if people tells me, not like this...). I have the impression that this is a Plato's cave illuminate with the private lives of those around me.

Technically, my understand is this:

* YT suggests channels closed to my interrests

* YT picks videos with thumbnails very closed to this "life of others" feed = the thumbnail leaks more than the video title.

I ended up to send a feedback to YT, no response, but most of these recommendations disapeared, and then it went back here and there, so I don't use the recommendation page anymore. YT music has more or less the same issue with automatic playlist.

The analytical part of me wonders why YT subscribes me to these "feeds", how these "feeds" are fed (maybe it goes beyond Alphabet / Google), but that's strange and disturbing enough.

Note: perhaps the fact I use ublock, blokadda, Firefox, etc... most probably makes my profile focussed on technical stuff, which makes other recommendations more visible.

https://github.com/searxng/searxng/blob/f1a148f53e9fbd10e95b... https://github.com/searxng/searxng/discussions/970

I'm not sure if this person is still active on HN, but I'm really curious about the results.

The intent of SearXNG is to be stateless (with no sessions on the server) and to work without JavaScript.

However, this approach limits certain features because of the restricted size of cookies (and other forms of browser storage require JavaScript).

The other one, I was not able to "teach" the algorithm what I like even after 3 months.

YouTube because this is one of the algorithms that provide good suggestions with minor work on my side.

Position because I can go back in time when I travel (I try to turn it off otherwise).

(there are other posts with one or very few comments)

[1] https://github.com/BurntSushi/aho-corasick/

[2] https://github.com/BurntSushi/aho-corasick/discussions/136

[3] https://github.com/intel/hyperscan

[4] https://github.com/OISF/suricata

[5] https://github.com/VectorCamp/vectorscan

There was the Unreal Mode aka Flat mode. With few lines of assembly code, an memory allocation using HIMEM.SYS, it was possible to access all the memory: the segment limit is 4GB instead of 64KB in this "mode". Of course it was not compatible with Windows, so the unreal mode was more for fun than anything else.

https://en.wikipedia.org/wiki/Unreal_mode

Post from 2022: https://news.ycombinator.com/item?id=32134038

> A file is further divided into blocks with variable lengths. We use Content Defined Chunking algorithm to divide file into blocks.

> This mechanism makes it possible to deduplicate data between different versions of frequently updated files, improving storage efficiency. It also enables transferring data to/from multiple servers in parallel.

I use it on old PC without issue. Drawback: since the files are not stored in clear, in case of data corruption of the Seafile repositories, I need backup (never happened to me).

* https://manual.seafile.com/develop/data_model/

* https://pdos.csail.mit.edu/papers/lbfs:sosp01/lbfs.pdf

* os.pipe and serialization (pickle or whatever): https://peps.python.org/pep-0554/#synchronize-using-an-os-pi...

* immortal object, but I don't see a way to create immortal object from Python (only from C). https://engineering.fb.com/2023/08/15/developer-tools/immort...

I guess it will more iteration to get a better way to communicate between the interpreters.