icodestuff

Sure, but weight gain in adulthood doesn't cause you to get taller.

Kinda seems like it is mentioned. I can't believe they wouldn't try to upsell you to the that.

Maybe Apportable will eventually have Swift support.

libdispatch is Apache licensed.

Substantially faster than 2, but somewhat inconsistently. Marty's brother started fading after only 9-10 hours. Marty doesn't start fading until nearly the week is gone. Old Biff starts fading after an hour or so? The tombstone fades immediately, and the matchbook and newspapers change within a few minutes of Marty stealing back the almanac for the last time, or possibly immediately after it's burned.

How is he supposed to get there without a passport?

Has the FCC had price-setting power since the Bell days? Particularly when it comes to surcharges, and as applied to state prisons, I imagine this is going to be overturned judicially.

> How can a millennial ’trep with no lofty urban-design credentials even consider proposing such a bold idea?

What is a "'trep"?

I've had 2 taxi drivers steal my credit card info in the past 2 years and taxis are still more expensive than Uber or Lyft, even through Flywheel.

Polite is great and all, but I've never had a Lyft driver steal my credit card info.

There are a number (16?) of vtable lookup-based message-sends for some of the common selectors, but in general, this is the fast path.

LAX is not, Tegel-Berlin definitely is. It's not universal either way.

Apportable basically used a heavily modified fork of GNUStep originally, but yes, then they did switch to a correctly layered architecture.

It looks like they took an approach similar to GNUStep, with its inverted layers compared to Apple's implementation. CF on top of Foundation on top of UIKit.

Looking a little deeper, it's missing some things... KVC/KVO are entirely absent. libdispatch looks to be pretty broken. NSNull is wrong in about 3 ways. There are several classes which are just stubs. Most of the NSLock subclasses are missing.

It's a good start, but I don't see it becoming the ObjC standard library by any stretch of the imagination.

The motivation for this was trying to parse a binary file with a discriminated union, with the tag being a single byte representing one of several ASCII characters. Since Swift doesn't have single byte character literals like single-quoted characters in C, UInt8 raw value enums required decimal or hexadecimal integer literals, which really hurts readability; 0x42 is strictly worse than 'B'. As a StringLiteralConvertible, "B" is an acceptable value for a UInt8 case.

Why does the author use lets to make closures rather than functions here? I don't see anything that would require it.

Bizarre, but to be fair, I distinctly remember watching a version of Star Trek Generations in theaters which was apparently never released in theaters - it included the full scene of Soran torturing Geordi, and explaining the nano probe (so Dr. Crusher's explanation of removing it actually made sense), the last fight scene going differently (Kirk escaped the breaking bridge only to get shot in the back and fall onto it, causing it to fall with him), and the missile flying into the ground instead of exploding. But I know this wasn't just mentally interpolating the deleted scenes I saw later, because it didn't include the orbital skydiving scene, for instance, and I remember on watching the VHS being very confused that the torture scene was missing.

The only remotely reasonable explanation I've come up with, besides sliding between universes, is that my dad and I actually went to a test screening instead of the actual theatrical release, but I think the last time I asked my dad about it, he _didn't_ remember those additional scenes, and there's no reason that a test screening should have been happening in my hometown. Another less likely explanation is that the theater was shipped a test screening print rather than what was actually released. Neither is particularly satisfactory, since I definitely hadn't seen the orbital skydiving scene which was ostensibly in the test screening prints.

Diet Coke is New Coke with artificial sweetener instead of sugar. Coke Zero is Coca Cola Classic with artificial sweetener instead of sugar.

I find myself completely unable to care that "floating point" is not in the title, particularly since it _is_ in the very first sentence.

You're reading something into the title that isn't there. The author says "7 bits aren't enough for 2-digit accuracy" and you're reading it as "7 bits aren't enough to represent 2-digit numbers." The post is about converting base 10 floating point numbers (with 2 digits accuracy) to base 2 floating point numbers and back, not representation of integers.

Oof. I've got a NAS with 4 3TB Seagate Barracudas, same model number as the drives with the 46% failure rate. One's already failed, and I replaced it with the same model drive.

Anyone have a recommendation for how to replace all the drives with more reliable models without wrecking the RAID? Downtime is okay, but I don't have spare space for the contents to make a giant backup. It's currently set up as a "Synology Hybrid RAID" that looks an awful lot like a RAID 5.

That might be better in one case, but then if gcc starts optimizing other functions which look like stdlib functions, should you really have to implement them all in assembly before being able to test anything?

Huh, I thought they did the falling/time dilation experiment on Mythbusters and found that the subjects could read the digits that were previously flashing too fast.

One question that's not well-answered in this article is how new bdelloid individuals are created. Do they bud like yeast? Do they lay eggs? Do they just cleave in two?

I played around with it a few months back (I'm working on something similar but substantially less complex), and while it does sort of work, the output is hideous, and is more of a starting point than usable code. As of when I tried it, running it on itself produced an uncompilable mess.

> The result is 89 terms which have the same meaning in a conversation (unless your votes/comments say otherwise). And that doesn’t even consider whether to say someting[sic] is O(n²), has O(n²), or is of O(n²).

That has more to do with how the rest of the sentence is constructed, and to some extent the selection of pronunciation of O(n²). e.g., selecting "order n-squared" makes "is of" the natural choice, while "big-oh n-squared" is less grammatically awkward with "is" (arguably, the former example could be "of order n-squared", in which case "is of" is not actually different from "is"). To completely disambiguate, one should specify what the growth is in respect to: time, space, or something else. For example "In the worst case, quicksort is oh of n-squared with respect to time." "Has" should be reserved for when using O(n²) as an adjective, e.g. "Quicksort normally has oh of n log n growth in time."

Of course they're outraged, in the same way that a serial burglar is outraged when they discover that someone's backdoor, which has been wide open for years, is suddenly locked. What's their outrage got to do with anything?

It was disappointing, I thought, that while the article did concern itself as to whether noticing the phenomenon was affected by confirmation bias, the author did not attempt to disprove the notion. Surely there must be a list of programming languages which have made it to version 6; an exhaustive search could at least determine if there are counterexamples.

And the diagram did get the terminology right. The author may need a (better) proofreader, but it wasn't impossible to see what the author meant. They only conflated the names, not the concepts.

It's an interesting choice, but I think I'd have rather had a sensible default for each type category than one (internal) for everything. For instances: public classes, public enums, private or internal ivars, public methods, etc. Just to cut down on the boilerplate.

Would tidal locking really happen fast enough for that to be a plausible reason?