Hacker Newsnew | past | comments | ask | show | jobs | submit | subscribed's commentslogin

Well, yeah, it's pretty good, but has a very low free tier limit. It's barely enough to cover my IOT subnet and subnet for stuff like AVR or TVs.

So if I'm supposed to pay for ad blocking, I might as well read the article and use one of the other methods[1] instead :)

[1] I was trying Rethink several times, but WG Tunnel with ad-- and tracker-blocking service works best for me.


I used NextDNS for a long time, but they offer absolutely zero support for their customers.

I ended up just switching to Control D's free tier using the Hagezi blocklist. You can try Normal, Pro & Pro+ for typical usage (I stick to pro normally). That + Brave and I'm pretty much ad free.

https://docs.controld.com/docs/free-dns


> ended up just switching to Control D's free tier using the Hagezi blocklist

HaGeZi recently launched their own public DNS-over-HTTPS / DNS-over-TLS resolvers: https://github.com/hagezi/dns-servers.

  # privacy
  root.hagezi.org
  # security
  ctif.hagezi.org

Very cool. I hope they get some NA server locations as ControlD has them beat there. Hagezi's DNS only use their Pro list + TIF, but if you wanted to use Normal or Light ControlD would still be a good option. I keep their TIF list in adblock.

https://controld.com/network


I used NextDNS for years and moved over to ControlD this year due to NextDNS literally not innovating at all and having no customer suppport. I've been much happier.

I did the same and just added zigbee repeater outside the fridge.

Works great.


It's the other way round.

If the device is sufficiently secure, it will be considered for the official support from GrapheneOS.

Of course, GOS can be forked by anyone, including Chinese OEM with insecure hardware.


Well, they designed some handsets (one is announced, maybe there's another one or two) to meet Pixels / GrapheneOS level of hardware security. That alone is BIG.

Then there's expectation GOS might be sold pre-installed as well.


Simplelogin doesn't do SRS which is pretty bad for this time and age.

That makes emails forwarded from, say, PayPal or ebay marked as scam/phishing in Gmail, AND makes it impossible to report actual spam that comes through.

With a well set up SRS all these emails are recognised correctly. SimpleLogin does some naïve forwarding which messes it up.

I'm migrating out of them exactly for this reason - if I could do it on my hobby server for my personal domains, THEY should be able to do it as well.


Self hosting is a massive PITA and time burner.

I self hosted my email domains for 15+ years. Finally got fed up with the constant trickle of the maintenance and got a service for everything.

Not quite happy with that as they don't even do SRS (implying they do), but i guess I'll just migrate elsewhere.

The upside is I've got more time for myself and have less CVEs to worry about.

Self hosted apps migration in progress; I guess I'll leave media, book and file server only, off the internet, and that's it.


Yeah, and Dropbox is just rsync and scp :)

Dropbox orginally used an rsync library

I still use USB sticks for a variety of purposes. I also still use FTP

I stopped using Windows many, many years ago. The company behind it now runs a dubiously consensual "botnet" (C&C, RCE) called "Windows Update"

Dropbox as I understand it was a company that wanted people to store their personal files on someone else's computers

This is the antithesis of "self-hosting" which what the OP is about


And windows could originally fit on five 5.25" disks.

Re Dropbox: https://news.ycombinator.com/item?id=9224


Corn for bioethanol is not feeding the nation, it's exploiting the fertile land to literally burn the output.

The argument about corn is to actually stop the waste of the fertile land for fuel. Solar panels can be installed wherever they're necessary, no one is proposing replacing the corn in situ.


Do you mean healthy foods as well? They vary from "much more expensive than" normal" food (not necessarily just highly processed/worst), to "impossible to purchase at all" due to living in the place that doesn't offer these options.

On top of that there's time cost - making healthy meals form scratch takes far more time than making them from store-bought "parts" (eg: to get boiled beans I need to soak them, them boil for hours. For this I need time, more time and a lot of energy)

I think you're not taking the average human's reality.


I think you are not aware of the average human's reality before 1980. The food deserts of today affect a small percentage of people. If you want something almost as healthy as the produce but prepared/processed you can buy it today but couldn't in 1980. You would also be spending considerably more of your income on food before 1980 no matter the quality. It's fun to blame scarcity of needed things but this has no reality. There is an abundance of preferred things.

Depends on the country.

In '80s in mine we had mostly only healthy foods. Crap sold as food today was completely unavailable; the most unhealthy would be a birthday cake or additional slice of bacon on a plate.

I think I am aware and I posit it's very hard for people on low I comes to even be able to buy a quality food.


Unprepared food was pretty much the only healthy food in shops before 1980. Poor people were more likely to need two bread winners..

The memes you are repeating do not match actual US or world statistics. They do make everyone feel better all around. I.e. even free enterprise and freedom of choice stories work worse when we are genetically tuned to choose ourselves to death and market tuned to profit from it. (A bag of potatoes is far more available globally than decades ago but a bag of doritos has a far more impressive availability increase.)


What memes?

And you didn't even engage with my main point, it's very hard for people especially on the low income to not just eat but even access healthy foods.


I did engage with your main point. Healthy prepared food did not exist in 1980 at a grocery store, the lowest income was more likely to need dual income and therefore not having more free time than now unlike richer people since 1980.

Here is the actual data that says people were spending 50% more on food despite more prep time and lower health food options:

https://www.ers.usda.gov/data-products/charts-of-note/110333

Your point is very socially satisfying but has no accuracy for why the majority of Americans are overweight or why the trend is global. (People buy expensive brand trash food but could in fact afford and find healthier food for less of their income now than in 1980.)


Why do you even call the detailed, verifiable claims about a product a "drama"?

So it's impossible now to list the bad features or shortcomings without come one calling drama?

Why do you call it drama?


One thing is to present the advantages of your system. Entirely different thing is to enter every discussion of alternative systems saying they have "atrocious" security. Also, the second link is an empty accusation, as it has no proofs.

They have listed a very long, concrete list of issues with Fairphone in the (sub)thread about Fairphone.

A very long list of severe issues. This is not presenting the advantages, it's listing of the problems.

When did that become a "drama", and not a fact-based criticism?

Re: second link - I don't have time to waste on digging through the code, I'd rather count pebbles in my garden, but I'm happy to bet £100 that it's fully facts based. I didn't even comment on the second link previously, BTW.


> I'm happy to bet £100 that it's fully facts based

Quote:

> a random identifier, generated on your device the first time it boots (a random UUID prefixed with anon), used to pick the wave.

> About this identifier:

> it is random: it is not computed from the IMEI, the serial number, the MAC address, an account or any other data of your device or of you;

> it is not linked to any account: you do not need an account to use /e/OS and the Updater does not know about it;

> it is reset by a factory reset: after one, your device gets a new identifier and is considered a new device;

> it is used to pick the wave.

https://doc.e.foundation/os/learn/staged-rollout/

No evidence that it is used or can be used for tracking. The accusation is strongly overstated. (You can keep your £100 though.)


GOS>> /e/ does have services collecting data on their users which isn't disclosed including user tracking via unique identifiers in the update client.

> a random identifier, generated on your device the first time it boots

So it's a persistent, unique identifier, is it?

> No evidence that it is used or can be used for tracking.

It's even worse than Google's Advertising ID that at least can be easily reset.

> The accusation is strongly overstated

> (You can keep your £100 though.)

I don't think how you disproved anything. You even confirmed the phone ID is immutable until factory reset.

I was talking about all four claims.

GOS>> They also spent years sending user speech data to OpenAI without informing users beyond fine print in the terms of use.

Here's the discovery by the user and confirmation from the owner of the project: https://community.e.foundation/t/voice-to-text-feature-using...

(This is sharing users' biometric data with third party without express consent, then shared further with unspecified )

GOS>> It's presented as not using Google services but has a whole bunch of Google services with privileged access enabled by default.

https://eylenburg.github.io/android_comparison.htm

GOS>> It even downloads and runs Google Play executables such as droidguard by default with privileged access far beyond the regular app sandbox.

https://doc.e.foundation/os/learn/calls-to-google-servers/

I believe it also enables Safetynet checks by default which downloads and executes proprietary code straight from Google.

So you didn't disprove the first one and you didn't touch remaining three.

I'll keep my £100 indeed.


Imagine that I would enter every conversation about GrapheneOS saying how atrocious their freedom is, including (1) reliance on numerous non-auditable, proprietary drivers and firmware; (2) full dependence on Google's direction of Android development and how fast Google shares the source code with others, as well as reliance on remote attestation [0]; (3) their disregard of GPLv3 that protects the user more [1]; (4) complete lack of flexibility concerning the user's threat model, e.g., intentional lack of support for root for whitelisted apps and lack of full user control over the running apps [2,3].

- would you call the above "a very long, concrete list of issues" or "attack and drama"?

[0] https://news.ycombinator.com/item?id=45232164

[1] https://news.ycombinator.com/item?id=49594324

[2] https://news.ycombinator.com/item?id=49543420,

[3] https://news.ycombinator.com/item?id=48767841


LOL?

I don't see how the evidence of the project unwilling to keep up with the mainstream security or a statement that their hardware doesn't come close to the top would be relevant to this:

(1) and how do you imagine releasing the OS for the hardware not owned by the project? Does your Fairphone release source code of all the firmware? Do you have source code of the BIOS, firmware and microcode of your chosen Qubes OS pc? :D

(2) LOL, Google pretty much owns AOSP, how do you imagine GOS project takes the helm? Please explain. I read these as the strategic decision of GrapheneOS project, or unwillingness to take action. Please explain.

[0] what are you even on with remote attestation? This is an optional feature for the OS provider to prove the whole chain of custody hasn't been broken. Can you explain how the GOS project providing an option for the software developers to check that the OS hasn't been tampered with is a downside.

Of course in the light of the Fairphone unable to do as much as peivide timely patches or the OS available for over 6 months now.

(3) they explained they "want GrapheneOS to have no additional restrictions beyond AOSP. AOSP uses GPLv2 but not GPLv3". I don't understand what the charge is? How does GPLv3 benefit end user more? Why aren't you complaining to Google?

(And why are we even talking about that? Are you trying to stoke some drama, LOL?)

(4) lack of support for root becomes what, a proof they're stoking drama?! what are you even on? They prove source code, build instructions and more details, you're free to build your own images with root, it's actually super simple - the only things you lose is their signature and remote attestation you hate anyway.

Why do you expect a project focused on the reasonable security of all the reasonably safe handsets to bow to you exactly?

Because frankly it looks like you're trying to cause some drama. None of your "charges" carry any weight. Quoting the team explaining why they chose GPLv2 over v3 as the "evidence" of drama?

What's your angle? What i#/your post really about? Yeah, I see you see all the above as a serious charges against the project, and you're free to throw these every time.

But.... Do you really think they're approximately the same weight?


> mainstream security

How is this mainstream security, if GrapheneOS are the only ones offering it? Cutting edge, maybe. Except Qubes OS of course.

> Do you have source code of the BIOS, firmware and microcode of your chosen Qubes OS pc? :D

My Qubes laptop runs coreboot with Heads with disabled and neutralized Intel ME. It does not offer the full user freedom, but it comes as close as possible.

> or a statement that their hardware doesn't come close to the top

I specifically mentioned that the GOS crowd comments on every single topic of other projects, which to me is just attacking them. They have different tradeoffs, just like GOS itself. This is why I chose to word it that this is akin to me attacking every GOS post with their own flaws, which are the result of the chosen (mostly fair but real) tradeoffs. I do not want to attack every GOS discussion. I want to demonstrate how ridiculous their actions are, dividing projects that in the end have the same goals: to fight with megacorps and reclaim user freedom and security.

(1) I do not. And I do not have a Fairphone. I just hate drama and fight against it. I have a Librem 5 though. All its drivers are free and all its firmware sits on external, removable devices. If (when) a certain vendor starts to misbehave and use their power against the user, we can replace the device.

(2) I do not believe that AOSP developed by Google will provide the necessary freedom in the future, at all. Google have shown again and again how they use their power against the users and against GrapheneOS. If you can't develop it independently, just don't. I use Librem 5, because I do not believe in the future of AOSP. Linux though doesn't obey a single megacorp: Its development is distributed over many organizations, greatly reducing the danger of misbehavior against users. Librem 5 is less secure than a GrapheneOS phone today, but nothing prevents the community to change that.

[0] GrapheneOS supports the option of using the phone against its users with DRM. People who value freedom are not okay with that.

> Of course in the light of the Fairphone unable to do as much as peivide timely patches or the OS available for over 6 months now.

AFAIK they follow the upstream patches, i.e., just like GOS, they are at the mercy of the corporations with their security. This is not exactly their fault, as they chose different main goals, which are important, too. Attacking them does not help anyone except the megacorps.

> Why aren't you complaining to Google?

Google do not promise privacy or control to their users. But they should definitely be fought against.

> How does GPLv3 benefit end user more?

You can read all the reasoning at gnu.org.

> (And why are we even talking about that? Are you trying to stoke some drama, LOL?)

I am trying to stop some drama by explaining why the GOS crowd should not do as they do.

(4) You completely missed my point. Read the above.

> you're free to build your own images

> Why do you expect a project focused on the reasonable security of all the reasonably safe handsets to bow to you exactly?

I did not demand that GOS does this.

> None of your "charges" carry any weight.

I did not suggest changes. I do understand the GOS choices. I demand that they, too, understand the choices of others.

> Quoting the team explaining why they chose GPLv2 over v3 as the "evidence" of drama?

The choices have nothing to do with drama. The inability of GOS to understand the choices of others and attacking these choices every single time something is posted is drama.

Thanks for the genuinely trying to understand me.


> How is this mainstream security, if GrapheneOS are the only ones offering it?

Not their fault. I should probably say "reasonable security". Settling for anything less is folly, seeing how malvertising isn't the only danger to the normal users. Broad availability of the powerful LLMs increases the amount of the vulnerabilities found and exploited in the wild.

Anything can carry an RCE now, web font, image, background audio on the cute website, attachment in the MMS or WhatsApp message.

I guess all those with "nothing to hide" (except their text messages, financial data, money on their bank accounts) are happy to have their lives sold in the open on more and more darknet marketplaces, their devices to mine some obscure crypto and serve as proxy for some nefarious actors.

You're using Qubes, I guess not because you like slower and less streamlined experience, but because you want security.

> My Qubes laptop runs coreboot with Heads with disabled and neutralized Intel ME.

Congrats, then, really.

Oh, wait, Qubes openly supports "CPU-vendor-provided blobs for silicon and memory initialization as well as other internal operations" -- which means you either strike "reliance on numerous non-auditable, proprietary drivers and firmware" from GOS (which runs it on the hardware that guarantees the separation), or raise the same for Qubes OS.

And actually, while we're at it, does your phone OS rely on the "numerous non-auditable, proprietary drivers and firmware" or not? Because if yes, why would you even complain that a very secure platform does that too?

Re: Coreboot: for me it would mean that I cannot use Qubes OS with hardware I want, because mean someone doesn't provide coreboot fw for my laptop, and even worse still, Qubes doesn't support my laptop.

I guess it's their fault then, seeing as some people throw a hissy fit that GrapheneOS doesn't want to officially support a phone they want to be supported :)

> I specifically mentioned that the GOS crowd comments on every single topic of other projects,

Will all due respect, this is just raging BS requiring extraordinary evidence.

> This is why I chose to word it that this is akin to me attacking every GOS post with their own flaws, which are the result of the chosen (mostly fair but real) tradeoffs

But you didn't list even one actual, objective flaw.

Not releasing firmware patches impacts everyone is objective, clearly negative. Not allowing to use custom keys is objective flaw. Not allowing to relock the bootloader is a flaw. Not having a secure element that processes PIN is a flaw.

Running privileged Google services isn't an objective flaw, if it's openly disclosed and benefits some users (so their handset can pass Play Integrity). No, GOS don't do it, for me it's an example what is and what isn't objective.

Supporting only one family of the phones is not an objective flaw if that's the only hardware platform that can back the security posture. And with the upcoming Motorola handsets support it's proved beyond doubt it's a flaw of the hardware manufacturers, not of the project trying to provide a secure platform for the increasingly dangerous Internet.

Maybe you remember how the freshly installed Windows XP without firewall couldn't get connected to the internet or it'd get immediately infected. Maybe you've seen logs from the Linux servers showing endless attempts to break in.

We're approaching the same for mobile devices, just worse, because most of these devices are grossly insecure, and the LLMs will be able to prepare custom exploits for all the common ones en masse.

> I do not want to attack every GOS discussion.

You do you. If you have actual flaws, by all mean, you're good.

But you didn't show a single one yet. You shared your opinion on how you believe that not supporting root is somewhat bad, or complained that GOS does something ALL the other vendors do (like reliance on AOSP, not changing licensing, or using fw blobs (at least in their case - securely))

> I want to demonstrate how ridiculous their actions are, dividing projects that in the end have the same goals: to fight with megacorps and reclaim user freedom and security.

I think it's a mistake to project your own position or understanding on all the projects.

Like, LineageOS supports a very broad range of hardware increasing security for many of these abandoned by their vendors (f*k Sony, for example, for releasing security patches for barely year-and-a-half for so-called flagship).

GOS has different goals. They result in the similar gains (i.e. ensuring security benefits privacy), but it's a result, not goal per se.

> Librem 5

So you don't have source code of the firmware for your chosen phone but you're attacking GOS for using firmware they don't have source of? I'm confused.

> (2) I do not believe that AOSP developed by Google will provide the necessary freedom in the future

Wait, wait, but you said "full dependence on Google's direction of Android development and how fast Google shares the source code with others" as presumably an evidence/example of specifically GrapheneOS flaw.

You "accused" AOSP-based operating systems team of being fully dependent on the AOSP.

And now you're saying it was about your *feelings*?

Do you attack projects stating your feelings as an "evidence" of the project's shortcomings?

Can you see how it's hard to have a fact-based, neutral discussion with that? It's broadly pointless because you either confuse facts and evidence with feelings or you equate these two.

> Librem 5 is less secure than a GrapheneOS phone today, but nothing prevents the community to change that

If the hardware is secure, sure. You can even backport a lot of GOS improvements back to Linux.

But if the hardware is not secure, then no, community cannot make it as secure as GOS phone today.

> GrapheneOS supports the option of using the phone against its users with DRM

I'm confused, first I don't know what specifically you're talking about (not saying it's untrue but throwing "DRM" randomly doesn't help much), and is it something unique to GrapheneOS that is NOT shared by other projects? Because, seriously, you're singling out the safest, most private and most secure project listing things that are probably done by everyone else but its somehow only GOS issue?

(Also, is it a safety issue or only privacy? GOS never promised a focus on privacy)

> [Fairphone] AFAIK they follow the upstream patches, i.e., just like GOS, they are at the mercy of the corporations with their security.

They're always late by many months. Months. They're free to do what GrapheneOS do, to keep requesting these patches.

But what are we talking about, they didn't even do as much as move to Android 17, which in itself is the evidence they don't follow upstream, because most patches are not backported by Google. Raising that is not "attacking" any more than listing unsafe features of a a specific car.

>> [about not supporting root] Why do you expect a project focused on the reasonable security of all the reasonably safe handsets to bow to you exactly?

> I did not demand that GOS does this.

But you specifically complained: "(4) complete lack of flexibility concerning the user's threat model, e.g., intentional lack of support for root for whitelisted apps"

They openly support users building their own images (with root. It's trivial to build the image with root enabled), and you actually complain they do not support root. They do not in their official images.

So you are now saying that by complaining that GrapheneOS do not support root in the official image, which you provided as (presumably) objective flaw of the OS, isn't in fact you demanding they support it officially and build it in, thus bowing to the request?

I'm utterly confused here.

About as much as when you send me to FSF website to read about some licence when I ask you how the project's decision to stick to the parent licence in order to not impose limitations on the users is imposing the limitations on the users.


Consider applying for YC's Winter 2027 batch! Applications are open till November 2.

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: