ralph87

It of course depends entirely on what kind of tool you are trying to build. PDFs capture a ton of visual information lost in plain text. I know the parent article is discussing construction of a text model, but what about say, if you were attempting to build the ultimate AI page reflow tool? etc.

The entirety of Library Genesis (about 2.7 million books, fairly poorly curated) can also be downloaded, its somewhere around 40 TB altogether of significantly more recent books.

This idea is common but a strange one at least to me. Going back to GSM would of course be a regression due to their repeatedly broken crypto, but modern LTE networks have many desirable properties absent from Internet-based messaging.

Perhaps it matters less to a US user, but as a European the fact an intra-country communication need not traverse any border nor be observable by any non-domestic entity is a huge deal for me, due to much stronger data protection laws in these parts covering both message content and metadata

Then there is the trustworthiness of the communication medium itself. WhatsApp is being advocated in this thread, but WhatsApp is a massive chunk of unaudited code running on your handset that is published by an advertising company, code that by design steals your phonebook without telling you every time you run it. At least in the UK, telecom providers generally act for the most part as commodity infrastructure, we don't seem to have quite as many scummy behaviours as are reported in the US. Both approaches require a central counterparty with either absolute trust to protect your cleartext message in transit, or provide otherwise unaudited code claiming to protect the same.

The bottom line for me is probably who gets to keep the metadata, a US advertising company or a local telco. I am certainly happier with the latter.

Just wait until you try out Gstreamer

(and suffer a few weeks learning it! I don't even remember C++ being so painful)

This style of embedding was popular at one point in Python's evolution before Django appeared. You can probably find 10 such systems in this list: https://wiki.python.org/moin/WebFrameworks#Discontinued.2FIn...

There are probably a few still around