How can collaboration without E2EE be considered "secure"?
(I'm aware there is a plugin but E2EE is definitely not built-in)
1,195 karma · joined December 13, 2015
Founder & CTO of reduck.ai
(I'm aware there is a plugin but E2EE is definitely not built-in)
It would be much more logical to do it with Messenger (for example by injecting JS with Tampermonkey) or to use a local model to perform S2T.
- How do you handle variable length data? Do you need to pad it?
- What is the memory overhead of the storage of encrypted data?
I think that at least for video data, the streaming scheme "leaks" the size of the encrypted data with the number of streaming packets.
I remember it was challenging to install because I tried to install with the 3 disks and it still didn't work after. However I found a torrent of a "portable" version that worked out of the box!
The "Scraping XHR" [1] explains how to inspect network requests and reproduce them with Python. I actually built har2requests [2] to automate that process!
[1]: https://scrapism.lav.io/scraping-xhr/ [2]: https://github.com/louisabraham/har2requests
On the same note, I wrote a program to generate Python code (requests) from a HAR capture: https://github.com/louisabraham/har2requests
I think using HAR captures is simpler for the end user than spawning mitmproxy as they don't require any installation and are extracted from the network tab of the browser devtools. Is there a reason why you didn't use them?
EDIT: I realized that mitmproxy can also get traffic from other devices like phones. Very cool project, I will think about modifying mine to support mitmproxy captures!
During isolation I bought my first mechanical keyboard (a cheap Qisan Magicforce 68, that I wholeheartly recommend). I quickly learned to use it and still love it at home.
A few months ago I bought a kinesis advantage for my office. I'm fully used to it (although it took me a while to stop hitting wrong keys).
When I bought both keyboards, I already knew how to touch type and went back to typeracer, keybr and monkeytype. My speed did not increase that much and I still get my fastest speeds on my macbook pro integrated keyboard.
The main reason why I use an external keyboard is to solve my back problems, by using a stand or an external screen.
In particular, I get my slowest typing speeds on my kinesis advantage. It is NOT built for speed, it is built for comfort. It allowed me to improve my posture considerably so it's still worth it.
PS: in my experience, using a laptop stand / external monitor helps with the lower back problems (weirdly). I solved my upper back problems by doing more exercise, and it's as easy as having some elastic bands near my desk and trying to stretch them horizontally behind my neck a few minutes a day (when I'm reading a long email or pdf or thinking or need a break).
PS2: a must-have on macOS is karabiner. I use it to have nicer keybindings for french accents (modifier keys like `) but also to configure function keys on my kinesis advantage and put navigation shortcuts on my mouse (logitech G502).
On a moderate use case (< 20 write operations per s) I kept getting timeouts so I first devised a quick fix system using RQ jobs passing writes to a single process and migrated to postgres as fast as I could.
Does this software help with concurrent write operations?
> For protection against enemy divers, dolphins will swim up to the infiltrator, bump into them and place a buoy device on their back or a limb using their mouth. The buoy then drags the outed diver to the surface for easy capture. When trained sea lions perform the same maneuver, they use a kind of handcuff with their mouths to attach the buoy
Imagine if we did this on land: monkeys or dogs handcuffing or tasing people! Surely they wouldn't be deployed against civilians and I know there have been exotic animals travelling with military (like bears) but training animals to fight humans with TOOLS looks incredibly hard.
I found https://gagarine.medium.com/use-sci-hub-with-zotero-as-a-fal... to get the pdf from sci-hub\
Thanks!
My typical workflow involves a big gdoc that I can share with my advisor. I paste the title, year, author and number of citations (to know which articles are the most influential). And, most importantly, a link to the pdf (sometimes through sci-hub).
Then I write notes under those articles, pasting screenshots of equations that I want to find easily.
When I have to write my paper, and this is the only painpoint, I have to paste the titles to google scholar and copy the bibtex code. I might write a simple bot at some point to do it for myself.
I'm taking every advice to improve this!
It reminds me of the hash maps that are said to be amortized O(1) but can be O(n) for some sequences of operations in various languages like Python and JS: https://blog.toit.io/hash-maps-that-dont-hate-you-1a96150b49...
It made me realize this is the purpose of PINs for some apps (eg Signal)
Signal erases that kind of information but I'm pretty sure that user must have had some messages delivered to them while signal was processing the subpoena. So pretenting they don't know anything else is just wrong IMO.
Just look at his previous blog post [3], in which he explains that the most cited neural networks all cite works by him. These papers cite dozens of papers, so a lot of other groups that are active in AI can claim the same thing...
[1]: For example, Turing published an independent proof of the Entscheidungsproblem, in the [TUR] article, just a month after Church, that the article forgets to highlight.
[2]: https://en.wikipedia.org/wiki/J%C3%BCrgen_Schmidhuber#Views
[3]: https://people.idsia.ch/~juergen/most-cited-neural-nets.html