yahelc

The purpose is so that Twitter is seen as the source of the traffic. A lot of Twitter-sourced traffic comes from native apps, so when people click links from tweets, they usually don’t send referrer information.

If the redirects were server side (setting the Location header), a blank referrer remains blank. Client side redirects will set the referral value.

From Twitter’s POV, there’s value in more fully conveying how much traffic they send to sites, even if it minorly inconveniences users.

How hard was acquiring the domain twenty.com?

I thought about it, but I hadn't originally built it to make money. It was a weekend hack project that I never expected to do that well, and suddenly the prospect of centering my life around a product I didn't really feel passionate about felt like a nightmare. I had a lot of anxiety about whether my external dependencies I'd built on top of would continue to work, and reached the point where a large payoff was worth it to de-risk myself from $23k/mo becoming $0/mo with no exit.

I built a social media analytics API. At its peak, it brought in about $23k/mo, with server costs at about $3k/mo. When I got a job at Instagram, they decided it was a conflict of interest and told me to sell it, which I did for a very healthy six figure sum.

Easy to imagine why this didn’t capture peoples’ attention in late March 2020…

This unnecessary condescension is made all the funnier by the fact that the quote the expression alludes to wasn't actually written by Mark Twain, but in all likelihood by a French writer named Nicolas Chamfort.

https://quoteinvestigator.com/2013/04/03/eat-frog/

Probably because they include HIBP data https://www.troyhunt.com/were-baking-have-i-been-pwned-into-...

The author, Virgil Goode, is best remembered for this incident. It may help in understanding his motivations in opposing high-skilled immigration. https://en.wikipedia.org/wiki/Quran_oath_controversy_of_the_...

In addition to properly escaping inputs, Content Security Policy Headers to restrict the hosts that the browser executes JavaScript from (e.g., script-src). https://developer.mozilla.org/en-US/docs/Web/HTTP/Headers/Co...

I’m well aware. But every other map handles this correctly except Google Maps. It’s not that hard.

Also a pain-point -- 2.6 million people live in Queens, NY, but Google Maps gets the city wrong for the vast majority of them, and it causes major headaches for residents (since so many services rely on Google Maps), and they have yet to fix it, and reject any edits that attempt to do so for them.

Picking a random example -- "1880 Willoughby" is in the neighborhood of Ridgewood in Queens. Either "Queens" or "Ridgewood" are acceptable as the city.

However, Google Maps calls it "Flushing", which is a totally different neighborhood 10 miles away. Apparently this has to do with the history of where the central post offices were located for given areas in Queens, but no other online map provider struggles with this distinction.

The city doesn't run the subway, the state does.

I've done this a few times and have found it to be an extremely effective way to do simple pixel tracking (for custom emails and the like).

The piece explicitly credits them:

> Workers have also started to replace faulty signals that trigger a train’s emergency brakes at low speeds, a problem investigated by The New York Times and The Village Voice that has also led to slower service.

You're close.

The bank isn't "a" bank, it's a Central Bank.

And it's Swedish, not Swiss.

No, cookies can be restricted to paths, it's just less commonly utilized.

https://developer.mozilla.org/en-US/docs/Web/API/Document/co...

Can't be this, since they don't have on-page ads.

Interesting that String padding made it in -- sort of jumps out as the simplest of these additions. I wonder how much of that had to do with negative PR for JS-land due to left-pad-gate.

The HackerNews repo is gone (and thus the issue), but at one point this was a Github issue I successfully filed to get HN to add it so their referral data for non-SSL pages wouldn't be lost:

https://github.com/HackerNews/HN/issues/68

(Background: https://news.ycombinator.com/item?id=7197520)

IIRC, I think the developer was open to adding a referral, but not have it leak any on-site information, so this was a compromise.

Voter files being public data depends on which state you're talking about -- you're looking at ~50 different sets of laws and regulations. Some states restrict access to only political parties using it for political uses and prohibit distribution to non-parties/candidates. Some states (Florida, for example) will let anyone just apply to get it.

To your latter question, some states, like California, do include email addresses on their voter file, but the coverage tends to be poor.

You can probably assume that, if you're a registered voter in the United States, you're on this dataset, as is your age, gender, party affiliation if applicable, and race/ethnicity info if your state collects that, as well as modeled information projecting things like party support, race, age, and likelihood to turnout.

Twitter loophole works: https://twitter.com/search?q=https%3A%2F%2Fwww.wsj.com%2Fart...

Someone actually built a model back in 2013 to predict risks of various Google APIs shutting down:

https://www.gwern.net/Google%20shutdowns

>$30 just for the delivery fee, a daily income of a well off family back in 1998

$30/day is $10,950 per year, which in 1998 was only $100 more than the poverty line for a household of 2.[1]

I agree it's a lot of money for a delivery fee, but it's just not true that that's the daily income of a well-off family in 1998.

[1] https://aspe.hhs.gov/1998-hhs-poverty-guidelines

Oyez does this for all cases. They also publish the oral arguments (which the Supreme Court itself releases) as a podcast feed, which I highly recommend. It's great listening even if you're not legally trained.

scotusblog.com also blogs about every case.

Perhaps I've been reading the wrong newspapers, but I can't recall Kraft, Coca Cola, Nestle, P&G, Coca Cola, et al ever being considered media companies.

Randall Munroe has a response to your lamentation: https://xkcd.com/603/

This is amazingly well-timed. Just yesterday I was complaining that the Python SQL libraries are annoying and verbose, and that I wished there was a library as good as requests is for HTTP.

From my testing, it seems like Gmail will cache via its proxy, but that cache respects the cache headers, so you can instruct Gmail how often to refresh the data.

(That, and you can cache-bust on a per user basis using a unique identifier in the query string.)

Is there a better migration path for MySQL 5.5 -> MySQL 5.6 than the existing recommendation, which is ultimately "Create a new database and port your data over"?

https://cloud.google.com/sql/docs/migrating

Useful but dangerous:

If you have (1,2,3,4,NULL) as data in column foo, and you query

    select foo from tbl WHERE foo!=1

Returns:

    2
    3
    4

Which, makes sense, but can be very confusing if you're not explicitly looking out for it.

Redshift (and I think Postgres generally) will also, very confusingly if you're not aware of it, return no rows if your subquery returns any NULL values in a NOT IN scenario.

If bar is (1,2,3,4,5)

    SELECT * FROM bar WHERE foo NOT IN (SELECT foo FROM tabl)

Will return no results (instead of a single row with 5).