Hacker News new | past | comments | ask | show | jobs | submit | rigid's comments login

Signal has reproducible builds for android now? Why not f-droid then, too?

Signal's definition of "reproducible" meant for quite a while "download this binary docker image and build Signal inside of it". I don't know if that has changed since.

Signal rejects F-Droid for a different reason, though: They only want to distribute through channels where they get download statistics and control update rollouts.


Hm f-droid provides privacy friendly https://fdroid.gitlab.io/metrics/ for some time now.

I'm not sure what sort of "control" they have over the Play Store compared to f-droid, but I'd rather have a trusted 3rd party do the building transparently and verifyable.


F-Droid uses a package maintainer-esque process where the maintainers of F-Droid can intervene and prevent an update to an app from reaching users if it's deemed to be malicious or to add anti-features.

It's of particularly high need on mobile since popular apps, even those who were originally FOSS, are sold to scummy publishers who fill it with ads and subscription schemes (oft called anti-features, since removing them could be seen as a feature in and of itself), ruining the original. You can't really trust mobile app devs because the track record is downright awful. Recently that happened with the "Simple" collection of apps, where the Play Store version got filled with junk but the F-Droid maintainer froze the version and marked the apps as outdated since nobody could conceivably want the new versions.

Of course, that strokes poorly with developers who a. don't want to deal with potential third parties in their distribution chain rejecting their updates or b. are planning to add anti-features to their apps later down the line. With signal, I'm gonna guess it's mainly a; the Play Stores checks and balances are much less invasive than the sort of thing an F-Droid maintainer might check for. (As I understand it, Google Plays checks mostly are anti-exploit and keyword scans.)


> where the maintainers of F-Droid can intervene and prevent an update to an app from reaching users if it's deemed to be malicious

That sounds like a feature you want when using FOSS.

Imagine distros wouldn't have been able to intervene quickly and malicious xz would be still deployed through their channels just because the authors want to.


Oh yeah, it's an absolutely wonderful feature. F-Droid is pretty much the main app store I'd recommend to get "the basics" from if you're ever in the unfortunate position of having to manage the mobile devices of family members. Having a maintainer "on the lookout" gives so much peace of mind. Not suddenly having the gallery app turn into a data collection machine and baiting less tech-savvy people into vaguely defined subscriptions is a value that's too good not to pass up on.

FOSS isn't really the important part for me there; it's nice, but the real value is that F-Droid is pretty much the only app store that has some reckoning on how the relationship between mobile devs and mobile customers should be far more adversarial than on any other platform due to the poor track record of mobile devs and empowers users to be able to deal with that in a way that restores some degrees of trust.

It's a fucking shame there's not an equivalent on iOS where you can just say "yeah, what you find here can be trusted" and then not have that gets polluted a year down the line. Apple used to somewhat police the App Store back in the early 2010s for similar peace of mind, but that's not the case anymore.


> With signal, I'm gonna guess it's mainly a; the Play Stores checks and balances are much less invasive than the sort of thing an F-Droid maintainer might check for. (As I understand it, Google Plays checks mostly are anti-exploit and keyword scans.)

It might have been b as well – Signal did keep their server code proprietary for many months to add their custom cryptocurrency to it, and added this cryptocurrency for microtransactions into the app as well. There may be many more features like this planned, some of which F-Droid might oppose.


Their problem is that F-Droid releases are signed by F-Droid, not by Signal. This way F-Droid could potentially insert a backdoor in an update.

That's not true tho. f-droid supports (true) https://f-droid.org/en/docs/Reproducible_Builds/ for quite some time now. Those are signed by both, f-droid and the author.

I should have checked before I posted something from memory. These are the reasons they list:

https://community.signalusers.org/t/signal-android-app-on-f-...

F-Droid with reproducible builds signed by both parties seems the best of both worlds to me, now I don't understand why Signal is so stubborn about this.


> This way F-Droid could potentially insert a backdoor in an update.

Google requires app developers on play store to give goole the keys that enable google to insert backdoors in any release. I can't trust anything on the play store for this reason. There is no way to tell which apps have been backdoored by google for whatever reason (the usual reason is a NSL).


mobile users should tap the canvas for the particles to move.


or shake the phone

Is the accelerometer data available?


yep. android 13


ledger-cli is rock solid but it lacks a decent TUI.

Really all I want is an 80s TUI accounting app that "just works" as ledger does but with a menu driven interface with keyboard shortcut support.

Modern stuff like multi-user capabilities, plugins, templates, API etc. wouldn't hurt.


not sure how lightweight any of these are, but https://gitolite.com/gitolite/ just needs git and ssh deployed. And it works like a charm.


running all my homelab / private repos off of a tiny gitolite3 server, and i'm always amazed that it all runs on an alpine linux VM that uses 75mb of RAM.

for anything even just sharing with friends or whatnot i won't recommend it though, people do love their web-UI and having to explicitly give someone access can turn off people already.


For a single author, if you have a server with SSH, you don't need Gitolite at all.

The lack of anonymous public access is often a deal-breaker though.


For a single author you don't necessarily need any server at all. A cloud directory or zip files work well.

But gitolite is so easy to setup & maintain, it's not a big difference and for r/w-access management within teams, it's priceless.

I guess one could even hack anonymous access with "PermitEmptyPasswords yes" and "AuthenticationMethods none"


No Gitolite requires keys.


Hence "hack". It needs keys for administration but at a first glance, I see no reason why a git-anon user couldn't be part of gitolite's git user.


You might need to use another user if you want to set its shell to `gitolite-shell username` (no command= for password authentication) but then you'd need to chain sudo or something to have Gitolite run under its own user again... Seems very tricky.

Or maybe you can write a shell that runs a gitolite-shell command is its arguments are not already gitolite-shell?


...at least you weren't looking for any references to "mifflin" functions.


The __mifflin__ method is used to define a way for the object to waste cpu-cycles so it looks busy


I kinda miss the curiosity show.

It was a bit more science leaning but got kids to awe just the same way.


> it enabled it in the first place

it took roughly two years including social engineering.

I'd say the same approach is much easier in a big software company.


How do you mean?


I bet in the majority of cases, there's no need to pressure for merging.

In a big company it's much easier to slip it in. Code seemingly less relevant for security is often not reviewed by a lot of people. Also, often people don't really care and just sign it off without a closer look.

And when it's merged, no one will ever look at it again, other than with FOSS.


An insider could just be tasked to look for exploitable vulnerabilities in existing code and compile this information for outside entities without ever having to risk inserting a purpose-made backdoor. Considering the security state of most large codebases, there would be a bottomless well of them.


Who wants this job, that is capable of actually doing it properly?


I think you nailed it.


I've read about workplaces that were compromised with multiple people - they would hire a compromised manager, who would then install one or two developers, and shape the environment for them to prevent discovery, which would make these kind of exploits trivial.


so, Office Space?


> where no auditor ever looks

Well, software supply chains are a thing.

"where no auditor ever is paid to look" would be more correct.


> login process

RCE doesn't really follow a login process design. As soon as you got RCE you can be considered pwned.

If not now, then at the time the next locally exploitable vulnerability comes up. There are plenty.


There's a gazillion nice python SSGs but when it comes to advanced stuff like responsive imagrs, i18n or js optimization, you always have to add it yourself.

It seems most of these never became fit for use for some bigger website.


Facebook's SSG Docusaurus does those things: https://docusaurus.io/

Though you may have to use a plugin for responsive images: https://docusaurus.io/docs/api/plugins/@docusaurus/plugin-id...

I've been using it for a static docs and blog site and it seems to be one of the most mature and comprehensive static site generators available.


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

Search: