dbieber

Yes, you can do this with OBS. The virtual webcam feature can be added as a plugin. And chroma-keying is a built-in feature. I use a Mac, and I've been toying with this a bit over the last weeks (e.g. https://davidbieber.com/snippets/2020-04-09-exploring-obs/)

Consider "Sign in" and "New user"

It would be great to see the video durations in the rankings.

Cool ideas. I’m definitely interested in adding browser history into the mix.

Transaction data, I already collect automatically (and semimanually label because the existing auto labeling is insufficient) and find quite valuable.

Love this question. I've written up part of my strategy for doing this here:

https://davidbieber.com/post/2019-12-29-track-your-life-in-a...

The short version is: (1) I manually enter the activity(s) I do each day in a spreadsheet, one row per day; (2) the spreadsheet automatically figures out how recently I did each category of activity; (3) [bonus] a chat bot uses this data (and much more!) to help me keep my life on track.

The decision has already been overturned: https://toronto.citynews.ca/2019/11/09/quebec-overturns-deci...

You may like Google Apps Script e.g. https://developers.google.com/apps-script/reference/gmail

I haven't used rust, so can you explain this to me:

If I do the rust equivalent of:

    def add1(x):
      return x + 1
    
    x = 1
    y = add1(x)
    z = add1(x)

then will x have been deallocated by the first call to add1 and will the second call to add1 fail?

[You can ignore the fact that I'm using numbers and substitute an object if that makes more sense in the context of allocating / deallocating memory in rust.]

You can read the text of the law here: https://leginfo.legislature.ca.gov/faces/billCompareClient.x...

I used to have a scrabble bot playing itself three times a day, posting its games to tumblr: https://scrabblebot.tumblr.com/

IIRC simulating an AI-AI game only took a short period of time (maybe a few seconds to a minute), so replicating the study with this scrabblebot wouldn't be too difficult.

Source: https://github.com/dbieber/ScrabbleBot

Each data point is a bunch of features x_1, x_2, ..., x_n. You can make new features for your data points using whatever functions you like -- it doesn't matter if they're linear. Let's say we add two new features x_{n+1} = f(x_1, x_2) and x_{n+2} = g(x_2, x_3).

Now if we train a linear model on the new expanded set of features, it's linear in those features. It's not linear in the original data though, because of the new features that we introduced: x_{n+1} and x_{n+2}.

I hope they're collecting data on how long it takes someone to answer, not just whether people answer correctly or incorrectly.

Author of Fire here. Was super happy to pass the 10000 star mark on GitHub today thanks to this HN post :) -- that boosted my spirits on an otherwise long and rainy (and now delayed) bus ride.

Hope you find Fire useful!

What's next for Python Fire? We're working on improving the help screens and usage outputs so it feels less like a developer tool and more like a professional grade CLI. So stay tuned as it gets better!

^I get that error with just `tangent.grad(np.sin)`.

So, it's not a nested differentiation problem, so much as a problem with "ufuncs". If you wrap np.sin in your own function then it takes the gradient just fine.

def sin(x): return np.sin(x)

negative_sin = tangent.grad(tangent.grad(sin))

An excellent question. Certainly switching from argparse to Fire will usually only save you some small amount of code.

But if your code is changing, that's when you'll start to see larger time savings. With Fire, your CLI will stay up to date without you having to update any parsing code.

I think the rules are implemented incorrectly. Multiple times I've seen a move get made that seems to violate the rules.

As an example, I just saw team Twitter place a piece in the center square of the top left board (the center board had not yet been won). The Github team then proceeded to place their piece in the bottom right board rather than the center board. It may have been the case that there was a split vote for center and bottom right in Twitter's move that resulted in this bug manifesting itself.

Yep, my bad. Added that too now. (changing not allowed)

Ask and you shall receive http://hnplays2048-democracy.herokuapp.com/

Edit: Now integrated into the OP's app

I added democracy mode...

The pull request is here: https://github.com/grant/hnplays2048/pull/10

You can play it here: http://hnplays2048-democracy.herokuapp.com/

Are you familiar with Brainfuck? http://en.wikipedia.org/wiki/Brainfuck It has just 8 symbols and looks like gibberish when you actually write it. I think you'll find it to be an amusing example of a very small Turing complete language.

I'm a big fan of the Facebook git creator www.in-a-gif.com

Wishful thinking.

My Facebook usage patterns have become significantly healthier since I installed News Feed Eradicator. I still use Facebook chat and other FB features extensively.

News Feed Eradicator: https://chrome.google.com/webstore/detail/news-feed-eradicat...

One's Wikipedia browsing history is very revealing. For instance people often use Wikipedia to learn about diseases they (or loved ones) have or are concerned they have. One should be able to research such diseases without revealing to advertisers or insurance companies the association between the user and the disease. The amount that can be garnered about an individual from their Wikipedia search history is kind of unbelievable. You can figure out things such as a person's interests, to some degree various aspects of their intelligence, some things about their medical history, if they are pregnant, their religious and political affiliations, if they may be suicidal, if they have surgery coming up, where they live and how old they are, what sort of pastimes they have, the list goes on and on. You can't figure out all of these things about all people, but the tracking has the potential to be very dangerous.

The most common way in which this data is used is targeted advertising, which probably doesn't concern you. Historically this type of data has also been used for price discrimination (which you may care about) and this data has the potential to be used in much more nefarious ways.

Even if we could somehow ensure the data would only be used in acceptable ways, the fact that the tracking goes on without most people's knowledge or approval seems unethical to me, and would be particularly troublesome on an encyclopedia of everything.

Introducing ads on Wikipedia would likely do more than show you ads and have ad providers pay Wikipedia; it would also introduce third party tracking.

Adding ads to Wikipedia would give ad providers the ability to trace your Wikipedia activity. This would be horrible because the average person would not understand that their wikipedia history was being revealed to third parties and this history is often highly personal. For nontechnical users who do not know about this tracking, this would be deceitful. For nontechnical users who do know of the tracking but do not know how to stop this tracking, having ads would severely compromise their ability to use Wikipedia since they must then become cautious of the way they browse.

Huh. It's funny you say this, because I got the opposite impression from the way the post was written. To me it came across as being written by a deeply positive person putting all their negativity in one place. The way they got to where they are now was by staying positive in the face of many small persistent negative things.

Think about how someone else who felt the same way as the OP would write their thoughts -- it would probably come across as a complaint, whereas this post comes across as a mix of introspection and advice.

Don't forget Google Shopping Express!

I used these when interning at Khan Academy this summer for an internal tool for video creators. Really very easy to get up and running, and looks good to boot.

>Experience creating programs to reproduce speech and musical style

Would love to see more about this https://ahrussell.com/projects/fifthseason/run Would be esp. awesome to hear the generated music.

niggler is referring to a low level of talent, though the word choice was less than ideal.