* Posts by ThomH

2913 publicly visible posts • joined 18 Jun 2009

Apple sprays down bug-ridden iOS 11 with more fixes

ThomH

Re: I don’t get it.

It looks to me like a reskinning in which the programmer was sufficiently inexperienced not to be aware of the default behaviour when using Core Animation. The issue is that the animation applied to the orange buttons (i.e. the operators) blocks input until it completes. A one-line fix.

Maybe they should have done something crazy, like QA testing?

ThomH

Re: Does IoS stand for

Maybe they use each iPad like an abacus by sliding apps between pages of the home screen? It'd be inline with the rationality of almost every other IT project ever undertaken in Westminster.

Hey girl, what's that behind your Windows task bar? Looks like a hidden crypto-miner...

ThomH

Re: Finally, a reason to move the task bar

The apocryphal version I heard was that there were no Windows 3.1 apps that had an issue with screens being different sizes, there were some that had issues with the origin of the user-interactable area not being (0, 0), and the coordinate system was a shared and exposed resource with no coherent way to offer different versions to different apps.

So the start bar went at the bottom because there were too many significant apps that either assumed the top left was (0, 0) when maximised or had a bad habit of spawning new windows at (0, 0), no coherent way to lie to them about the coordinate system, and too many edge cases in every attempted kludge.

But unless and until I read it on something like Raymond Chen's excellent The Old New Thing, I'll continue to take that alleged version of events with a pinch of salt.

Used iPhone Safari in 2011-12? You might qualify for Google bucks

ThomH

Re: Actually

You're living in a backwards world if you've decided that the company that makes money by tracking and categorising people in order to show them appropriate adverts is a better privacy bet than the one that makes money by selling physical hardware that it in part advertises through PR attacks on its competitor's lack of concern for privacy.

ThomH

Re: Nobody caring that iOS didn't enforce this setting?

Google found and actively exploited a dishonest representation designed explicitly to circumvent what was then Apple's implementation of user privacy. Apple, like every other browser maker, attempts to use various heuristics to increase privacy without breaking web standards — in this case Safari was making a judgement about which cookies to admit to having based on the combination of domain and content being served, so Google started providing fake content in order to be able to track the user. Its HTML was custom crafted to look like one thing to the browser while looking like another thing to the user.

That's why for most of us it is substantially less concerning that Apple's attempt to provide additional security above and beyond web standards was flawed, than it is that Google methodically and deliberately decided to find a way to override the user's stated desire for attempted privacy, in order to earn itself extra money.

You mean Google updated its smartwatch OS and nobody noticed?

ThomH

Re: Any recommendations for a solid, inexpensive Wear watch?

I've had a much better experience with health watches, though only monetarily.

My 'shower-proof' 2014 [Intel/]Basis 1 suffered water ingress in 2016 and hence died. Despite being well out of warranty, I contacted customer support and was given a free replacement. That was nice momentarily, but then all the devices were recalled late last year. Full refunds all round though.

So earlier this year I bought a Garmin. After about nine months the glue that holds the screen on came loose along one edge, the screen slightly protruded and, once again, water got in and killed it. Also out of warranty (because I'm in the USA, so those don't last long), but a free replacement was provided regardless. Which is what is currently on my wrist.

Given that I had also picked up a second-hand Basis 1 for the wife very shortly prior to the recall for a pittance and then received a full original purchase price refund for it, I have now been wearing a sports watch for a bit more than three years and am almost £200 up on the deal. The current one even also does notifications and media controls and other things I don't care about.

ThomH

Pedant's point

The 2015 version of the Apple Watch ceased to be available in 2016; the 'Series 1' is a fairly quiet refresh introduced when the second-generation model came out. The Series 1 gains an extra processor core over the 2015 original: it goes from single core to dual core.

No change in the requirement for every-day charging though. So not for me.

Arm Inside: Is Apple ready for the next big switch?

ThomH

Re: bootcamp? @Thomas Wolf

The main difference I've seen is that VirtualBox supports only either software rendering or passthrough of antiquated version of OpenGL. It's pretty easy to demonstrate the difference: on my MacBook if I enable accelerated rendering then no browser supports WebGL. If I disable it then they run WebGL in software.

VMWare supports passthrough of relatively modern versions of OpenGL. So WebGL is accelerated. As is the desktop compositor I happen to use, which makes a massive difference for ordinary productivity if, like me, that means X11.

ThomH

Re: Reminder of Acorn Advert

But don't they live on only because ARM was spun off in order to allow adaptation for the world beyond Acorn... as a joint venture with Apple?

ThomH

Re: @Dan 55

I think bitcode is merely alignment plus calling convention plus data size dependent. You could use it to port to a different instruction set if all of those things were the same.

That being said, if Apple didn't care about backwards compatibility then why did it expend so much effort on the 68k emulator and on Rosetta? The company even skipped the very first PowerPCs because the instruction cache wasn't quite large enough to fit the 68k emulator no matter what they did, so didn't produce acceptable performance with 68k applications.

Back to the Fuchsia: The next 10 years of Android

ThomH

@Shadmeister Re: anarchic, fragmented, insecure, ...

You should consider a Boogie Board Sync. Boogie Boards in general are cheap paper-sized eraseable boards that use an electronic means for self-erasing. The Boogie Board Sync is a 9.7" eraseable surface that also saves whatever you write as PDF. It has no screen — you're still making analogue marks on an analogue surface that can be manipulated only by the external use of pressure, there are no pixels, no refresh rates, no way to review a saved PDF directly on the document. But everything's preserved for you to sync to a computer later.

I don't think they're officially distributed in the UK though. A quick Amazon check shows availability for $100 in the US, one marketplace offer in the UK for £146.52. So close to twice as expensive. I chanced upon them in Hong Kong airport but no longer recall the price there.

Some 'security people are f*cking morons' says Linus Torvalds

ThomH

Re: Exercise stack to avoid everything living in registers

But what are the good kernel-language alternatives? I love C++, I'm just not sure about the other twenty or thirty languages I see people using that apparently also are C++.

US trade cops agree to investigate Apple's 'embrace and extend'

ThomH

Re: When you swim with sharks...

For an example of semi-modern Apple doing exactly this look up Sherlock 3 versus Watson; Watson was a commercial third-party internet search tool somewhere around 2002 that brought a large variety of different vendors' search functionalities together in a light, clean box without any of the web's presentational baggage. Apple gave it a design award and then shipped the suspiciously-similar Sherlock 3 in the next version of OS X — Sherlocks 1 and 2 had been purely local search; they're Spotlight predecessors.

There was no court case partly because the authors of Watson seem to be quite mature about this sort of thing, and probably because the Sun swooped in and bought the original anyway. Then did nothing with it, naturally, but I'm sure the cheque helped.

Silverlight extinguished while Angular wins fans among developers

ThomH

Re: Or to look at things a different way...

I think you're overestimating the larger part of developers. Why bother reading the documentation when you can just throw some approximate words into a search box and find some code to copy and paste?

Crowdfunded ZX Spectrum revival just days from shipment

ThomH

Re: As a BBC owner...

Only two years late...

The Z80's most primitive operations take four clock cycles to complete. The first two cycles fetch the opcode. The second two are the refresh cycle — the Z80 can refresh dynamic memory for you, and does it then.

So when the ZX80 and ZX81 are walking through display memory, the video output gets two memory accesses for each instruction, not one. It grabs the opcode during the first two cycles and (normally*) supplies a NOP to the processor. Then during the refresh cycle it puts the captured value back onto the bus along with an internal 3-bit counter to fetch the actual pixel byte. Then it outputs that.

There is only a one-byte latch, to preserve the character from the opcode fetch and reproduce it as part of the address during the refresh cycle. No 32-byte buffer. And the top part of the refresh address is unchanged during the refresh cycle. So that's how you can change the character set lookup location.

Both the ZX80 and ZX81 use static RAM. So they don't need a working refresh cycle.

* unless bit six is set. That ensures HALT gets through. It also explains why the ZX80 and ZX81 have only 64 characters in their character set.

Where hackers haven't directly influenced polls, they've undermined our faith in democracy

ThomH

Re: The biggest UK hackers of the lot then? @Dr Stephen Jones

I'm not sure that "the referendum got the result I wanted" justifies failure to consider whether a hostile foreign propaganda source that has been found to be active in the US might also be attempting to influence the UK*.

(* though I'm more inclined to point to the costs associated with being in an economic union looking outsized after ten years of economic turmoil that made the benefits look slender; probably worth a quick peek under the covers though, wouldn't you say?)

Fitbit health alert: You appear to be bleeding

ThomH

I'm not sure about the two-horse race

With HealthKit and Google Health anybody can make a device that contributes data to and consumes it from the same store as everybody else*, and a lot of companies are doing that in modest but profitable ways to different niches — those are the Polars, the Garmins, and the Xiaomis. There's no need to have a blockbuster device, and if you look at Fitbit's result then it looks like it's not even necessarily workable to try.

* other than Fitbit, which likes to own your data for itself.

Bored 'drivers' pushed Google Waymo into ditching autopilot tech

ThomH

Re: Looking like mixed tech isn't an option ...

Mixed tech definitely has a place, but in much the way it is actually being implemented in production models: to warn the driver if what they're doing is dangerous, and possibly even to override the driver if what they're doing is egregious. Lane assistance is probably already doing the world a statistically-significant benefit and extending that line of logic to, say, a hypothetical car that would seek automatically to pull over if the driver became unresponsive feels like a good thing.

Offering to take over driving but only in spates and with the handover points being unforeseeable and only immediately before a crisis does indeed feel extremely foolish.

First iPhone X fondlers struggle to admit that Face ID sort of sucks

ThomH

Re: Do you know what works better than Face ID and Touch ID?

Yeah, let's not pretend that Touch ID is super reliable. On my previous 5s and 6, and on my current 6s*, it's probably one time in five that Touch ID fails and just becomes the much more long-winded way that I get to entering my PIN. I'll bet Face ID works just as well as Touch ID ever did.

* guess when I moved out of iOS development.

Google's phone woes: The Pixel and the damage done

ThomH

Re: Unfortunately, the high-end smartphone market is Apple's to lose.

In all fairness, Palm did a really good job at responding to the iPhone quickly by launching the Pre in 2009, a better device in a whole bunch of ways that showed a forward-looking awareness of the direction of the market. They just didn't really have the resources or the clout to get it off the ground, so were severely region and carrier-limited, and shortly out of money.

But I definitely wouldn't charge them with being asleep at the wheel. For my money, shoddiest attempt at survival has to go to Symbian and its absurd degree of modality, especially re: the hacking back on of a soft keyboard after the UI had long ago lost support for it. That was an entire extra context-free screen every time you wanted to type text anywhere. If Symbian had decided to proceed with UIQ (as per Sony's P900) regardless of most manufacturers' indifference, I think they would have been in a much healthier position, but once Nokia became the near-exclusive paymasters I guess that would have been politically difficult.

What just trousered a $4.5bn profit, has glum desktop chip sales, and rhymes with go to hell?

ThomH

Re: I thought it was AOL

I thought maybe Honeywell had really turned things around.

How to make your HTML apps suck less, actually make some money

ThomH

@myhandler

There's still no native supplied SVG renderer outside of a web view, which makes developers more conservative. Apple's current advocated solution is to produce your assets as PDFs, after which Xcode renders them to the appropriate size classes at build time.

For iOS 11 onwards, there are two size classes, as there are no @1x 64-bit devices.

It's unclear exactly to what this article refers, but Apple's App Store uses a process called 'app thinning' to deliver to each device only the assets that device will use. So a 56mb upload is rarely a 56mb download. Or, if they genuinely mean a 56mb download, then that should vary.

I would guess that they're holding off on SVG support because it's a many-tentacled thing, requiring both SMIL and JavaScript support for a full implementation. But it's a real hassle that they don't support, say, a static subset or some other vector format, especially as they provide COLLADA support for SceneKit.

'Screaming' man fined $149 for singing 'Everybody Dance Now'

ThomH

Re: Forget it Jake, it's Montreal

Forget it Jake, it's Funkytown.

Neglected Pure Connect speaker app silenced in iOS 11's war on 32-bit

ThomH

Do we have grounds to believe the 8 and X can even run 32-bit code?

If Apple's choice, given that it designs both the processor and the OS, was to spend the transistors on something else then might that not actually be more forgiveable than just dumping 32-bit support for maintenance cost reasons? I'm not sure the comparison to Microsoft necessarily holds.

That being said, I also think the reference to 2015 by Pure is disingenuous. Apple's first 64-bit handset came out in 2013 — twice as long ago. If you weren't supplying 64-bit builds even by 2015 then you were supplying a poor product even then.

I can understand an interregnum if there were developer issues, but this is Apple. If you spend four years ignoring a transition it is trying to make then you're being very foolish indeed. I think Apple deserves a lot of blow-back for the unsupported apps that are now dead, but a company trying to keep a 32-bit app as a going concern has only itself to blame.

Look! Over there! Intel's cooked a 17-qubit chip quantum package

ThomH

Re: around the size of a US quarter

It's about 1,131th as far as Intel had to travel to "steal the base" mentioned in the first sentence.

Apple's iOS password prompts prime punters for phishing: Too easy now for apps to swipe secrets, dev warns

ThomH

Or...

The pop-up says "your credentials are required", tapping takes you from the app to something somewhere in e.g. the system settings, entering your password there takes you back to the app in question. Very hard to fake effectively, besides anything else because the status bar belongs to each application individually, so is part of the transition animation, but individual apps can't animate it individually, and impossible to fake flawlessly as the user will be able to tell if they press the home button.

It's pretty much only in-app purchases that I can think of where you might be asked to enter your password within a third-party application, so the extra friction wouldn't be a constant hassle.

Dumb bug of the week: Apple's macOS reveals your encrypted drive's password in the hint box

ThomH

Dumb bug of only the week is being fairly generous

An alternative suggestion as to the source however: when one uses Apple's interface builder, one task is to connect outlets to graphical elements, e.g. you know there's an NSTextField that the user will type a password into so you declare an NSTextField property and then you switch to the interface builder where you have laid out the dialogue and you wire the property to the control — literally drag a connection from the one to the other. Then in code you access the text field's contents via the outlet.

A drag and drop error that connected both the 'hint' and 'password' outlets to the password text field would then result in the password being recorded as both, even though the code says 'self.password' for one and 'self.hint' for the other. And the wiring is all within the undocumented XML format used for interface layouts, so good luck getting a meaningful code review on that.

Given the whole purpose and importance of a password hint, it's mind boggling that nobody tested the feature.

Google touts Babel Fish-esque in-ear real-time translators. And the usual computer stuff

ThomH

Re: Well Done, Google, From The Large Unilingual Traveller Road Warriors Gang @big_D

When did you try this test?

Today Google gives: Öffnen Sie nicht den Fall, keine benutzerfreundlichen Teile im Inneren, which at least keeps the proper negatives, even though the parts switch from being user-serviceable to user-friendly.

So: same conclusion — don't trust — but it's clear that they continue to work on it.

Commodore 64 makes a half-sized comeback

ThomH

Re: Look in the loft @John Sanders

Batman Forever is both (i) the obligatory CPC demo of modern times; and (ii) the only good thing called 'Batman Forever'.

ThomH

Re: Look in the loft

The CPC is a well-designed machine — it's trivial to set yourself up with a full address range of RAM, pick any portion of that to be the size you want of video output, you've got the equal-highest resolution graphics of the era, one of the faster CPUs, a better-than-usual palette, great support for the disk drive machine (even if the disks are weird), and the BASIC is well-structured and provides good hardware support.

They don't all support interlacing, as Amstrad switched CRTC supplier a few times, but if you get one that does then you can output 1bpp graphics at a resolution slightly higher than DVD. If you bought the model that comes with a monitor, you can probably even see them all. Though you'll probably want the 128kb machine for that, as I count almost a full 64kb spent on such a frame buffer.

You only suffer because developers tended to treat it as that thing you hastily port your Spectrum game to in a couple of weeks somewhere near to the end of development ― as long as it sort of works, that'll do. So there's a vast library, but you need to be a little selective.

ThomH

Re: Seems.... @LDS

Then whatever you do, don't try to look at the catalogue after you've started writing your program. Though retyping it might be faster than saving on a 1541 anyway.

The axeman strikes again: Microsoft has real commitment issues

ThomH

The difference is that whenever Google cancels a service, it offers three to five replacements. Hit me up on Allo if you want more details.

Apple Mac fans told: Something smells EFI in your firmware

ThomH

Re: @boltar

Apple has all-but deprecated Objective-C. Objective-C's only substantial surviving use in new development is that via Objective-C++ it bridges to C++ slightly more easily than does Swift, though an alternative route through C may be more desirable — one can now annotate appropriately to produce an object interface in Swift as desired; it has always been able to make plain C calls because, as mention that you are aware, Apple's Core frameworks are generally plain C.

As a daily X user it is hard not fully to be aware of the distinction, I just pointing out the juxtaposition of a claim that inconsistency is problematic and a claim that X is essential. My work machine is Ubuntu MATE but I spend a lot of time NX'd across to a RedHat machine running a distinct version of GNOME. Then Eclipse on the NX'd machine for most of the actual work. So in net I deal with a completely incoherent UI. The Mac I used in my previous job at one of the 50,000+ head Silicon Valley companies was infinitely more consistent. But you'd be an idiot to use a Mac as a back-end server, and I've experimentally switched which side I develop for, so here I am.

I've actually been a Mac user for over a decade. I think inconsistency probably peaked somewhere around 10.4 or 10.5, when you'd frequently see at least three types of window chrome just on Apple's own apps (brushed metal being the oddest detour, but unified versus non-unified toolbars ran for a while, and drawer interfaces took a while to die off). I really don't see that an open minded user would have any cause for confusion.

ThomH

@boltar

Trust me, the people who are actually paid to keep 10,000 employees in laptops have little emotional attachment to each machine. They just like to be able to stockpile them, swap them at a moment's notice, or even authorise a travelling employee to replace their own with a quick trip to the High Street.

They're probably also aware that Objective-C has never been a systems language and isn't even the thing people use for UI development now. It is frustrating that Apple has its own language for UI development perhaps, but given that the alternatives are C# (Windows), Java/Kotlin (Android), C++ with a custom preprocessor (Qt), C (GTK), there's no trend being bucked.

Apple's system languages of choice are C and C++. You can observe this from the open source kernel, or anything else Apple has ever open sourced. Such as their kernel. Or the Clang compiler.

I have never before heard somebody simultaneously try to argue that randomised menu scattering is a bad thing and that X Windows is a good thing.

ThomH

A Mac is: the UNIX laptop that's durable, supported for a prolonged period, readily available in bulk everywhere in the world, and safe to base a fleet on because it or the next model will still be available next year. As a result they're ubiquitous across Silicon Valley.

iOS apps can read metadata revealing users' location histories

ThomH

Re: Accessing as files @DougS

It's a metadata interface, with the metadata currently including location*. The actual file might not even be on the device — it might be in the user's cross-device photo stream, in which case metadata could be the full extent of the device's knowledge. You have to issue an asynchronous request that may involve a network access if you want to work with image contents.

* and dimensions, duration if a video, creation and modification dates, and whether it's a favourite.

Mac High Sierra hijinks continue: Nasty apps can pull your passwords

ThomH

Re: only "signed apps" mentality?

The exploit isn't public so this is speculation at best but it sounds to me like the signed/unsigned distinction is a bit of a red herring here. As per the article "[n]ormally, apps, even signed trusted ones, trigger a prompt to appear on screen when touching the operating system's Keychain database"; it sounds to me like he's found a security exploit and, separately, demonstrated that the exploit is present whether your app is signed or unsigned.

I take that to mean: Apple made an implementation mistake somewhere, which is orthogonal to signing. I don't think signing is meant to be Apple's solution to guarding the Keychain.

Also although special rules apply for kernel extensions, I think anybody who pays $99 gets a signing certificate, no questions asked. Signed apps are more heavily sandboxed (e.g. no access to a directory unless the user has used the OS-provided file dialogue to open a file from there) but nevertheless that'd be the worst web of trust ever. Apple seems more interested in having permission they can revoke than in vetting those who want it in the first place — and if it makes some money too, fantastic.

Web devs griping about iPhone X notch: You're rendering it wrong

ThomH

Re: When you turn it sideways

"Horton insisted on Friday that today's site layouts will look fine on the iPhone X from the outset because content is automatically placed in a safe area where it won't be obscured by the sensor housing."

This is what the phone already does. By default web sites don't get the area covered by the notch. If you use the WebKit extension you can specify that you want the entire screen area to count and can find out which part of the display is actually display.

So I don't really understand the controversy.

So. Should I upgrade to macOS High Sierra?

ThomH

Re: New file systems are a rare treat

They didn't miss it, since they've offered an argument against it. It's just not a very convincing argument. Per Apple, since SSDs contain ECC error correction they were unable to create a statistically significant number of errors over the service lifetime of any of their machines.

Except that, you know, they then thought checksumming was worth it for file metadata. And apparently are unaware that external drives exist?

But 'missing it' isn't quite right.

iPhone 8: Apple has CPU cycles to burn

ThomH

Re: Wasn't this fantastic GPU allegedly made...

I think the concept of "there's no smoke without fire" speaks volumes here. Teach the controversy!

Apple has notified Imagination that it expects no longer to be using Imagination IP next year or the year after, and will cease to pay licensing fees then. Imagination says it doesn't think Apple could achieve that.

Imagination has not alleged employee stealing. Imagination has not alleged that Apple is failing to pay for its IP right now.

ThomH

It's probably valuable to the Instagram crowd that they can update the heavily fictionalised versions of their lives more quickly, and mobile gamers can now enjoy nicer graphics between their micropayments. Indeed for them, the phone is the cheapest part.

ThomH

Re: Wasn't this fantastic GPU allegedly made...

"A lot of people are saying X. Now I've asked some people — smart people, the best people — and, you know what these people are like. They don't speak to me, I speak to them. And, I don't know, but a lot of people are saying X!"

Behold iOS 11, an entirely new computer platform from Apple

ThomH

Re: Typos a Gogo!

"I come from the past and can see that much has changed. Tell me, what are these strange metallic boxes that now dart down every street?"

"Why, those are cars! Did they not have those where you come from?"

"Sure, we had cars. But these are nothing like the cars of 30 DAYS AGO!"

ThomH

Re: Is this a step backwards?

No, it's fine, because it's not a "proper files app" in the sense of the Finder. It primarily allows you to navigate your iCloud, Dropbox, etc storage and maintains local caches of that, providing applications with pipes into and out of that conceptually remote storage. It doesn't expose the native file system. You can't investigate /Applications or /Library or /usr/bin anything else. iOS-level sandboxing still applies.

Sacre bleu! Apple's high price, marginal gain iPhone strategy leaves it stuck in the mud

ThomH

Re: £1,149

Per Apple, FaceID doesn't work if your eyes are closed. Or you can hit the power button a few times to disable FaceID until your security code is entered as you see the police approaching.

Or you can just rely on it not really working very well.

Apple: Our stores are your 'town square' and a $1,000 iPhone is your 'future'

ThomH

Re: These "new" iPhones

No, you're wrong. The demoer used a from-offscreen-bottom swipe to dismiss the current application rather than an on-screen home button.

So to answer your entirely-proper question: Palm.

Five ways Apple can fix the iPhone, but won't

ThomH

Re: Sound @SpeakNoEvil

I'm not persuaded I agree on DACs: I don't actually agree with the decision to remove the headphone socket but given that it's gone, the only analogue output a modern iPhone has is its built-in speaker. You're not going to get any benefit out of putting in a good DAC for that.

Incidentally, I checked: third-party DACs are available. You don't have to use the $9 one Apple sells. For those with a real love of cables, USB DACs like the Dragonfly are usable via the lightning to USB adaptor.

Mega VR roundup: Lots happening in the virtual and real worlds

ThomH

I just switched jobs from one that provided a Vive and an appropriate computer in my home to one that does not. So mine has been returned. But it was just just occupying space so no big loss — the novelty wore off pretty quickly and the things it adds through freedom of movement (within bounds) and a more immersive display just run against the things it takes away, such as the ability to use software for more than twenty minutes without nausea, or to do anything that requires you be less immersed, like have a conversation. We hadn't used it in months.

I think it'll end up like 3d cinema though. It's not an expensive thing to add, and a certain audience will continue to want it. You just might have to pay an extra £10.

Apple pulls massive HomeKit chip U-turn to keep up with Amazon Echo and Google Home

ThomH

It's HomeKit news, so: if a tree falls in a forest, etc. Home automation is the biggest consumer hit since 3d television.

systemd'oh! DNS lib underscore bug bites everyone's favorite init tool, blanks Netflix

ThomH

Re: I wonder what the excuse will be this time.

I searched and discovered that per its authors: "[i]n contrast to the glibc internal resolver systemd-resolved is aware of multi-homed system, and keeps DNS server and caches separate and per-interface". So the justification for a new resolver was machines with multiple active interfaces, that possibly go to different networks.

The justification for bundling that new resolver with the init system? No idea. For distributions switching to it despite it not functioning very well? Clueless.