semiotagonal

Git uses hashes to identify commits, but does it actually chain them, i.e. include the previous hash as an input to the current hash?

Sure. It's the world Americans already live in, except for certain specifically defined types of lies (e.g. fraud, perjury, libel of non-public figures, etc).

Of interest: https://www.theatlantic.com/politics/archive/2016/08/does-th...

In the US it is almost impossible for a public figure, especially a political figure, to win a libel lawsuit.

In theory, a deliberate, malicious false statement against a public figure would count as libel; but in practice basically everything is allowed. This prevents politicians from intimidating their critics with lawsuits.

EDITED TO ADD: In fact in the US there is pretty much a constant stream of false statements about every politician, and I can't recall any politician ever winning a libel case. Devin Nunes is trying currently, and basically making a fool of himself.

If you want a short-and-simple username, you need to do what it takes to defend it, even if that means logging into twitter once in a while. Even registered trademarks require active defense, or they lapse[0]. Why should it be as easy as just grabbing it first, especially if that pollutes the platform in a way that hurts the platform company?

If you want an easily defended but unique identity, pick something that isn't short-and-simple, and you'll have less competition.

[0] https://en.wikipedia.org/wiki/Laches_(equity)

I don't think a smaller number of dumb startups will have much negative impact, considering the huge numbers of highly paid employees the giant "fang companies" have.

There's nothing about GPS that says it should provide continuous information to others about one's location.

Also if you sit near the Android people, let them know I'm going to be irritated if I find that my parents have been tricked into activating location tracking on their phones again! It seems like I have to shut off all the various Google privacy settings every time I visit them. They don't want to be tracked.

Thanks for the response. I'm mostly familiar with Google Docs in the workplace, where it was apparent who was reading what document at a particular moment from a list of users at the upper-right.

Separately, on my personal account, at some point I viewed a publicly shared document from a well-known person, and that ended up in my "shared documents" (or "documents shared with you"?) for ages. I wasn't sure if that sharing was evident in both directions. I've avoided clicking on shared Google Drive documents since then.

If it's not the case that the owner of the doc can audit activity, I stand corrected.

EDITED TO ADD: I suppose the root of my confusion is that it's pretty hard to distinguish the behavior of Google Drive from Google Docs; and once the lack of privacy has been observed anywhere (e.g. using Docs the workplace), it's hard not to assume it's also happening with PDF's shared via Google Drive, simply because you can't know what the other person is seeing.

It's especially hard when there is an asymmetry (e.g. in a non-GSuite environment, I can see some publicly shared document in my "shared with me" list, but I guess the sharer can't see me).

I wish he'd keep it running, then write on lessons learned turning $200K into something much less, should such a loss be manifest.

I think you're missing my point that even if you willingly impose Google surveillance on yourself, Google Drive is still worse than Google Analytics because your personally identifiable activity is visible to the author and/or other readers, not just Google.

I should have made it clear I was assuming that people were authenticated to their Google account. I think most people who use GMail never log out, and might not check the URL and realize they were navigating to a Google doc.

Doesn't Google Drive show the identity of someone reading the document to other people reading it, or at least the owner? That's worse than just collecting analytics internally at Google.

> The good news is that there is one way to avoid this (the only way AFIK)

I managed to recover from a similar level of burnout/pissed-offedness by changing my domain of software development completely. I got out of high-stakes server-side software and into client-side mobile software. It took some effort to pivot, but in the end it was a great relief. In my experience, Apple's slow approval process is a much better problem to have than getting pinged at 3am for any reason.

If I were the sixth person at a five-person startup I'd be forever worried that I was just on the wrong side of the get-rich-if-it-succeeds line.

That's actually pretty minimal. It only looks cluttered because the desk is small. The shelves are only half-used, and mostly in proper stacks.

Learning on the job in an unforgiving domain can be costly.

I did about 0.1% of the amount of research you did a few years ago and was really struck by the irreversibility of mistakes. Other engineering domains can have irreversible (e.g. fatal) mistakes as well, but they don't have as many enthusiasts jumping in to the field.

Discovering new physics doesn't move us away from a unified theory, it just means we were wrong about how much farther we have to go.

This is why having a social safety net is so important. If society can put a floor under how badly people can lose, they'll feel more free to try low-probability high-payoff endeavors.

From a professional perspective, does it make sense to get on a train that is so crowded already? Step 0 is probably to take Andrew Ng's on Coursera, but as of right now, you'd be among "2,647,287 already enrolled!" [0]

[0] https://www.coursera.org/learn/machine-learning

A person who made his name manipulating unwitting people probably isn't the best spokesman for social responsibility.

Every laptop should ship with a charger and a roll of electrical tape.

On Reddit, several weeks ago, I strongly objected to an author singling out a presidential candidate with a nickname, while not doing the same for the others.

It's fair to say I was outraged since I threw an "F" word in there. Although I didn't do it to seek status, it got something like 2000 upvotes.

Anecdotally, I can see that grandstanding works, if "karma" is what you care about.

How hard would it be to run Slack in a special VM that brokered UI events from another application?

I.e. the latest Slack would be running in a VM that made everything seem normal to it, but in reality the UI input and output would be supplied by another application which was presenting a "Slack Classic" interface to the actual user.

I don't believe it will happen, but if it did, I think cities and outlying areas would return to a division according to desired excitement level, instead of who can afford what.

The number of people involved in the authentication would have to exceed the other party's ability to bribe them. Corrupt practices could extend to both companies involved in a deal.

Perhaps a deal could be worked out where the satellite companies also have to launch Atacama-scale telescopes, where Musk et al have to pay for the difference in cost between Atacama and orbit.

> ...the serial entrepreneur and investor.

I still don't understand the virtue of being a "serial entrepreneur", especially with no indication of what businesses were created. It just sounds like someone who bailed on several previous companies that aren't even worth mentioning.

I would add...

* Supplier relies on middleman because they can't write convincing sales pitch in the target market's language.

* Supplier relies on middleman so they can get into (and out of) business more quickly; don't need an end-to-end plan to get started.

The vast majority of the time I don't notice noise at all, but when I'm doing something at the edge of my ability, when I can't quite nail it, noise becomes extremely distracting.