Comments URL: https://news.ycombinator.com/item?id=47117096

Points: 3

# Comments: 0

BTW, they also alter paywalls and other elements, because otherwise, many websites won't show the main content these days.

It kind of seems like "altered" is the new "hacker" today?

They're basically recommending changing verifiable references that can easily be cross-checked and verified, to "printed on paper" sources that could likely never be verified by any other Wikipedian, and can easily be used to provide a falsification and bias that could go unnoticed for extended periods of time.

Honestly, that's all you need to know about Wikipedia.

The "altered" allegation is also disingenuous. The reason archive.org never works, is precisely because it doesn't alter the pages enough. There's no evidence that archive.today has altered any actual main content they've archived; altering the hidden fields, usernames and paywalls, as well as random presentation elements to make the page look properly, doesn't really count as "altered" in my book, yet that's precisely what the allegation amounts to.

They bypass the rendering issues by "altering" the webpages. It's not uncommon to archive a page, and see nothing because of the paywalls; but then later on, the same page is silently fixed. They have a Tumblr where you can ask them questions; at one point, it's been quite common for everyone to ask them to fix random specific pages, which they did promptly.

Honestly, you cannot archive a modern page, unless you alter it. Yet they're now being attacked under the pretence of "altering" webpages, but that's never been a secret, and it's technologically impossible to archive without altering.

Who exactly has the environment where you can add, let alone promptly repair/replace, USB key sticks, on your server? Or run PXE when you have just a single server? How exactly do you do that in Hetzner or OVH? Let alone any other service where you get just a single dedicated server or two.

So, we're big enough to have our own quarter-rack in a collocation facility, let's do PXE. Now you have to have a whole separate infrastructure server, just for your other servers to be able to boot properly? (And how exactly does that server itself boot?) Plus, have an extra infra server for redundancy?

Sorry, but this is the reason noone would use SmartOS. You can't build a fortress on such a shaky foundation.

It's simply out of touch with the target market. At least with FreeBSD or OpenBSD, you known it'll just work™ on any single server, as long as serial console access is available, which is standard-enough. Going against the mainstream of Linux is already hard-enough, there's no reason to make it any harder.

SmartOS sounds like a lot of work, for negligible or even negative benefit.

There's zero good reasons why any machine with 450GB+ of zfs-backed redundant storage, needs to rely on USB keys or networking, in order to function properly. There's a reason Samsung's Joyent entirely abandoned and divested of SmartOS, because this sort of over-engineered mentality, simply doesn't compute. It prevents all sorts of usecases, and even with a growth mindset, still prevents the organic growth from a couple of servers to a rack and more.

How exactly does it make any sense to use ECC memory and ZFS RAID for error correction and redundancy, but then rely on the modern floppy disk for the OS itself?

In other words, if all you want to do is run a personal mailserver, or even a corporate one, you'll probably not have to deal with this supposed IP reputation issue, unless the IP addresses you use, have already been added to the blacklists even before you start at it.

That's exactly my point, that the reputation need is overstated by all those services that claim to solve a known problem that everyone has heard of, but noone has actually experienced, because, guess what, it might not actually exist.

I've seen plenty of cases where the emails sent out through Sendgrid et al, end up in the Spam folder, or these "professional" services don't even attempt to retry, thus, never getting through the greylisting, or other bugs which cause deliverability issues, which would never happen if you were to run your own real mail-server on your own hardware yourself.

Sadly, it's been announced yesterday that the nginx.org mailing lists are being shutdown by end of month (Sept 2025).

P.S. Probably one more reason to look into into the freenginx fork of nginx — https://news.ycombinator.com/item?id=39373327 — their mailing lists are at http://freenginx.org/en/support.html.

Comments URL: https://news.ycombinator.com/item?id=45390121

Points: 248

# Comments: 197

Comments URL: https://news.ycombinator.com/item?id=45364532

Points: 5

# Comments: 0

But I feel like ALL browser vendors are not doing enough to combat this bloat. There have to be resource limits, warning messages/icons, and stop-gap measures to avoid pointless JavaScript wasting our electricity; but NONE of the browsers do this to an extent I'd wish they'd do; in fact, Chrome has actually been ahead of Firefox and Safari in reigning these sites, probably because it has to run in production on 4GB ChromeOS machines costing $99, whereas all the Firefox and Safari devs are probably using 48GB machines costing $2399 as their benchmarks. So, the reality, is that, ironically, Chrome is again the leader even in this area. Because Chrome on a $99 4GB ChromeOS machine feels snappier than Firefox on a $999 MacBook, given enough open tabs.

Your point about feature bloat sounds good in principle, but is not practical in reality. In reality, if things don't work in Safari, you're simply asked to install an app from the App Store. Or if you have to configure a keyboard on a Mac, you have to use a Windows machine with the native keyboard configuration tool, instead of VIA in Chrome WebHID or WebUSB. Why in your opinion are these alternatives not worse than having these sorts of things as web standards as written by Chrome?

How exactly is Chrome the same as Edge or Brave or Vivaldi or Yandex Browser or Opera?

Why are there no browsers on iOS besides Safari, and how is that not a monopoly?

The "Internet Explorer" issue culminated with Microsoft attaining a market share that allowed them to stop all innovation and investment into the product, where the browser became substantially lagging behind the competition, as well as lagging substantially in standards compliance. Something that's currently an issue with Safari, not Chrome. (Please enlighten me if that's not the case — which exact standards does Chrome NOT support today? Else, how is supporting EXTRA experimental standards a bad thing?) Chrome and Blink, on the other hand, became market leaders not because they couldn't be uninstalled, but because of superior engineering; Blink is the only browser engine today where you can configure your gaming keyboard, for example. How's that NOT innovation?

Why do you have to keep redefining words according to some laws some politicians wrote, or misplaced analogies that turn things upside down, in order to sustain your points? The only Internet Explorer of today is Safari — severely lagging behind in most modern features, without any ability to be uninstalled or replaced on the iPhones and iPads. Again, I'm actually typing this in Firefox on desktop. As I said, I don't use Chrome, it's not even installed on my machines; because it doesn't have a monopoly in any way, on any device besides ChromeOS. (If you're curious on why I don't use Chrome or Blink on any desktop, it's because I cannot stand blurry text, and there's no way to disable blurry text in Safari, WebKit, Chrome or Blink, which have mandatory antialiasing, making all text super blurry and ugly; that's the actual monoculture we should be talking about.)

Whilst at it, can you kindly explain how Blink is even a monopoly if it's actually separately distributed by 6+ distinct and unrelated/competing vendors, namely, Google, Microsoft, Brave, Vivaldi, Yandex, Opera, etc? Out of these 6 vendors, a total of at least 3 are running an entirely independent search engine, so, these aren't just "fronts", but real competitors.

Whilst at it, can you kindly explain why is it better than I have to use a Windows machine to configure my keyboard or mouse, or the Bluetooth headset, instead of using a web browser on any device with any OS? Or why do I have to download extra apps to get video conference access instead of using a Blink-based web browser from one of like half a dozen vendors?

First of all, it's simply a trial, a whole year from now, in Sept 2026, and only in 4 mid-sized countries around the world.

Second, they'll only be verifying developers, not users. They won't be reviewing the apps any more than they already do today. They already do scan all third-party apps, which is partly why people are upset about the needless doxxing of the devs.

Also, as far as I understand it, an app store like F-Droid already does app signing on behalf of other developers, to ensure funny stuff couldn't simply slip through undetected, so, as such, F-Droid probably already "owns" the Aurora Store and all the other apps you can download through the F-Droid app store, so, it would be my expectation that even in those 4 countries in Sept 2026 during the trial, you could still sideload the same apps the same way I do today.

In turn, the apps installed by Aurora Store are signed by the Play Store; this ensures that the private data cannot be hijacked through modified updates of the app, since the developer profile won't match. So, there's no concern there, either, since everything is signed.

Basically, it's not a good precedent, but at the same time, nothing will really change at least for my own workflow (as a non-publisher), where I don't install anything outside of F-Droid or Aurora Store anyways.

Keep in mind, it's still just a trial. And even if it goes worldwide in 2 years in 2027 (which is still a big if), it's still FAR more consumer-friendly than anything Apple has ever allowed on iOS in any jurisdiction I'm aware of.

The latest trend in OS design are an immutable system partition, so, obviously, you cannot modify the underlying system image, neither on macOS, nor iOS, nor Android, but what evidence do you have that doing an overlay disable isn't enough?

I've been using Android for years, and have not seen funny business after I disable Chrome. You can use Brave or Vivaldi or Yandex Browser or Opera in place of Chrome at all times. Or Firefox in many cases. I routinely have fully functioning test devices with stock Android without any Google Accounts or any Chrome. Everything just works the way it should. Including the banking apps installed through Aurora Store through F-Droid, as well as the streaming apps like Amazon Prime Video etc. Again, all of this works without a Google Account in any way on my side as an end-user, and it's expected to continue working even in 2027 even if the trial they've announced goes through worldwide. It works on any Pixel device, it works on any Motorola device, it even works on Samsung, too.

I'd much rather have to switch to Brave or Vivaldi for a video phone call, or keyboard configuration, or NFC, than install half a dozen of outdated third-party XXX-only apps with full permissions and questionable security practices or distribution methods.

The better question to ask here, is, why would you NOT want to have a CHOICE to have these things in a secure browser by SEVERAL distinct major vendors like Google, Microsoft, Brave and Vivaldi, and Yandex, and Opera, and others?

Again, I don't even use Chrome. I replace it even on Android. So, I am not concerned with Google taking me over, because they clearly aren't.

But how am I more secure when I have to install lots of dodgy apps to get the most basic things like video conferencing working?

I'm typing this in Firefox on a Mac; I usually uninstall Chrome even from Android, usually after first using it to install F-Droid, then Aurora Store; then it's disabled promptly. Why should I not be allowed to disable Safari?

Apple's iOS monopolies are a far bigger issue than Chrome. These issues you guys talk about, don't really exist for me. I use YouTube regularly without Chrome, I use Android without a Google Account, I use all the banking apps without Play Store. None of this is possible on iOS. On iOS, you cannot preserve your privacy at all, because everything depends on having an Apple Account, and being monitored by Apple. Hence, I don't take iOS or iPhone as a serious contender for a daily driver for me.

Personally, I uninstall Chrome whenever I see it (on Android, I first use Chrome to download the F-Droid app store, then I uninstall/disable Chrome promptly).

But if you do agree with the author that Safari is so poor that noone would keep it, if other browsers were available on iOS… Isn't that a pretty weak argument against browser choice? You're basically agreeing with the author that Safari is so poor noone would use it if it wasn't mandatory? How's that a good thing we're all forced to keep using such an uninspiring piece of software?