tkfu

I'm not sure I buy the embedded/IoT use case; OSTree is a really good model there and is more featureful. The "well, if your filesystem image delta happens to be in the form of a lot of very small files it's not so great for CDNs" doesn't strike me as a terribly good reason to give up everything OSTree gives you (especially with stuff like the meta-updater [1] Yocto integration).

[1] https://github.com/advancedtelematic/meta-updater

(Full disclosure: I work for Advanced Telematic, the creators and maintainers of the meta-updater Yocto layer.)

Asciidoc is wonderful, especially with Asciidoctor, which has now become the de-facto standard renderer. (It's what github uses, for example.)

It's as easy as Markdown (and the syntax is so similar it's not hard to switch), but just has much better design, along with supporting the features you need once you start having enough documentation that you need an actual tech writer. Plus, it's still just plain text, so it plays nice with Git and is thoroughly readable without being rendered.

How did you manage to get something written in Go into the iOS app store? I was under the impression that foreign binaries weren't allowed.

> Uber is proving that car service is not a natural monopoly. If the dominant provider is taking too many liberties with pricing, someone else will come in to right the market.

This is complete nonsense. Car services have never been a monopoly, natural or otherwise; in every major city there have always been multiple car services. Uber's valuation is premised almost completely on the hope/expectation that they can become a monopoly and start exploiting their market position to the detriment of everyone.

Seriously, check out the Naked Capitalism series linked above. It lays out in great detail why Uber is not in any way an example of someone "coming in to right the market". They're wronging the market.

I detest this essay. I won't go too far into the reasons why, because David Beaver has already done an excellent job of that: http://languagelog.ldc.upenn.edu/nll/?p=992

I'd also agree with Geoff Pullum's characterization of it as "a smug, arrogant, dishonest tract full of posturing and pothering, and writing advice that ranges from idiosyncratic to irrational" (http://chronicle.com/blogs/linguafranca/2013/04/04/eliminati...)

But apart from it being a very poor source of writing advice, I don't believe it's accurate in its diagnosis of how language is deployed for political ends—the question is a much more complicated one than he claims here.

There exist actual good books that can teach you how to write better, instead of patting you on the back and allowing you to tut-tut at those plebians who write things that are "outright barbarous". Pinker's The Sense of Style is one of those; Ann Lamott's Bird by Bird is another. There also exist much better (and more accurate and scientific) resources about how language actually affects the way we think about things. Benjamin Bergen's Louder than Words is an excellent start, and virtually anything from Lakoff's long list of publications is worth reading—Metaphors We Live By is the classic, but Women, Fire, and Dangerous Things in particular is excellent from his more recent work.

Also, "I talked to somebody currently working for this company while they're at their job, and they told me it was great" is not exactly a very reliable way to get data.

I think it's safe to assume that drivers working 70 hours a week are already working during the surge times, and so the benefits are already accounted for in the average.

To me all his assumptions seem pretty reasonable. Since Uber isn't providing a detailed breakdown, all we can do is estimate. And if we're willing to accept the basic assumption that people respond to economic incentives, remember that Uber's incentivized to be as misleading as possible in their favour. Even if it turns out they're completely lying, there's a reasonable chance that the economic benefits from building network by doing so are significantly greater than potential future costs if someone found out they'd lied. Given that context, I think the author was actually being overly generous and deferent to Uber's stated numbers.

My father-in-law's an NYC taxi driver, and before that (in the 90s) worked for a black car service. So I can tell you from inside experience that you make more money in Manhattan than anywhere else, period. The fact that Uber drivers have the freedom to go to New Jersey or Far Rockaway or wherever doesn't mean that they actually get any benefit from doing so.

Also, while those coverages might be difficult-to-find-but-available now, that could change in a hurry. Once most insurance companies catch on to the fact that people are paying for regular insurance and then operating essentially a taxi service, they'll add clauses to their contracts specifically forbidding it, and charge something more in line with commercial taxi insurance. It's a lot like what we see with AirBnB, where as it caught on more and more, landlords started fighting back against tenants who were using it.

For (1), how do you think it applies? The situation analyzed is taken from Uber's numbers, and is for an "average" driver that works 70 hours per week. If you're working 70 hours, you're not gaining a huge advantage from surge pricing.

For (2), again, this is a vehicle being used 70 hours per week, 40,000 miles per year. If you were looking at the case of a part-time driver it might not be appropriate to allocate all the expenses as business expenses, but in this case I think it is.

Is there any particular reason you didn't just add your domain to the pool at https://freedns.afraid.org/ ?

The turnstile is actually a great analogy for DRM, including all of its flaws. And the turnstile isn't what allows subway service to exist.

A turnstile isn't very hard to jump over. If the turnstile is the only enforcement mechanism, anyone who wants a free subway ride just has to vault over (or perhaps also glance around to see if there are cops nearby, google whether there are monitored cameras, or pull a hoodie over their face).

There's an alternative, though: most places in Europe, there are no turnstiles, and nothing preventing you from getting on public transit without paying. It's a "trust but verify" system: you are required to validate your ticket before you get on the train, and occasionally someone comes around the train to check if you have a valid ticket, and issues a fine if you don't have one.

With this method, the users aren't inconvenienced, enforcement costs are reduced, and everyone's happy. With the turnstile, you get massive queues at busy stations during rush hour, and by making this one mechanical device your sole point of enforcement, you've actually made it more likely that people who don't want to pay will cheat the system.

I think the problem with offering both options is that you can't count on your customers to choose the right one, and they'll blame you when they choose the wrong one.

Customer (via SMS): hey, the whole internet on my phone crashes when i use feature x. wtf???

Customer Support Rep A: Hey, we'd be happy to help you with that. What platform/version/subscription do you have?

Customer (7.5 hours later): its the latest version, and i have a verizon. why can't you guys fix this???

CSR B: [lists some possible troubleshooting steps, asks customer to call]

Customer, the next day on Twitter and Facebook: [Company] has the worst support! They broke my phone, and they say they'll help you by text, but then when you ask them anything they just ignore your problem or tell you to call anyway. FUCK [COMPANY] AND DON'T BUY THEIR CRAP!!!