Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

What is triald and why is it taking so much disk space?

eclecticlight.co

Post reply on HN
Print view View original

91–100 of 295 posts

Re: What is triald and why is it taking so much disk space?

#91

Post by npteljes »

AussieWog93 wrote: post #11
sdfjkl wrote: post #6

Open source is where you don't have to apply guesswork to find out what a suspicious looking piece of software is doing on your computer.

Nope, instead you have to apply guesswork to figure out why an apt-get call accidentally uninstalled your desktop environment. :P

I'll take that every day. Because it's not like proprietary doesn't fuck itself up. For an anecdote, I had a laptop with Win8 preinstalled, and Update just wasn't working one day. The solution? Scouring the internet to find a Microsoft executable that fixed update. How? We'll never know. Because we're not meant to.

Re: What is triald and why is it taking so much disk space?

#92

Post by morganvachon »

Brian_K_White wrote: post #55
can16358p wrote: post #44

Earlier quoted context omitted.

Why? (Assumimg Vscode respects the setting)

I don't even understand the question. If you found that your next door neighbor was sneaking into your house while you're away, and he said "oh I have an option not to do that. Just ask me not to any more." Would that be good enough? You say "ok I would like to invoke that option" and that's it, you happily go away on a month long vacation without installing any cameras or anything because they guy said he won't do t…

I agree with your position but that's a bad analogy (as most analogies are). With your computer you can choose to uninstall and no longer use VSCode. You can't choose your neighbor and you can't force him to move away to prevent his bad actions. Also, it would be more like your neighbor is peeping in your windows or planting a monitoring bug to record your activities, rather than breaking in.

Taking that further, you can ditch your Mac and switch to Linux or BSD if you're concerned about the article subject. In other words, you can sell your house and move if you don't like the direction the neighborhood is going with all the peeping toms about, but there's extra work involved in that solution so you have to decide if you can afford such a change for more privacy. So, you can sell your Mac and move your workflow to a less invasive OS on hardware you choose, but depending on how deep your workflow is intertwined with and relies upon the macOS ecosystem, that can be a ton of work as well.

Re: What is triald and why is it taking so much disk space?

#93

Post by koprulusector »

>Over the last year or so Mac users have run into problems that appear related to a background service named triald. Some report it stealing huge amounts of CPU, others associate it with various glitches, and a few have noticed gigabytes of disk space apparently being taken up by its folder at ~/Library/Trial, and in their Time Machine backups. This article explains as much as I know about this mysterious new service, and ponders what’s going on.

- I find it so odd, Apple users fascination of their own operating system, this sort of voyeur or wildlife documentarian attitude. It’s your machine, but the inner workings are this ineffable mystery.

>I’m not aware of any control over or opt-out from Trial, although it could be covered in Apple’s general request to share data including panic logs, if you can remember where to control that.

>I first noticed Trial when studying Visual Look Up in the log. At the start of each Look Up, the subsystem com.apple.trial enters the following into the log:

>Initializing TRIClient. Trial version: TrialXP-292.18

>_PASEntitlement: Entitlement "com.apple.private.security.storage.triald" is not present.

>Found entitlement: "com.apple.trial.client" -->

> 0x12d9547e0 (no container): using Trial root dir >This suggests that Trial is recording information about Visual Look Up, although nothing more explicit appears in subsequent log entries.

- Apple can double speak all they want about digital privacy, but their actions tell the truth. As a paying Apple customer, you’re not deemed worthy enough to even understand how your own machines work, less even to “tamper” with their working, and who knows what data is actually being shared?

- Personally, I say no thank you. I’ve been on Debian or a variant for almost 10 years now, and haven’t looked back. It’s only gotten easier, as software generally moves more and more to SaaS web apps and the intrinsic cross-platform nature of the browser.

Re: What is triald and why is it taking so much disk space?

#94

Post by dangerface »

mschuster91 wrote: post #8

Jesus I hate being A/B tested without my consent. Particularly, why is Apple of all places doing this, given their otherwise strong track record of standing up for their users in regard to tracking and other bullshit by major players?!

> given their otherwise strong track record of standing up for their users

Citation needed, Apple have an advertising identifier built into the os to help advertisers track you, they share analytics about how you use their os by default, they analyse the data you store with them in iCloud.

I know Google, Microsoft, Amazon they all do these anti-user things thats not my point, my point is that Apple say they have a strong track record of standing up for their users and their users believe them regardless of their track record.

Re: What is triald and why is it taking so much disk space?

#95

Post by mhitza »

londons_explore wrote: post #88
secondcoming wrote: post #83

Earlier quoted context omitted.

> The privacy implications of this are dubious. Really? Apple can essentially be asked by Law Enforcement for answers to questions like "Give me all the IMEIs of people in location X, radius 10 miles, that appear to have pictures of cats"

The model fine tunings can't immediately be used to say "this person has pictures of cats". They effectively are just a big diff of the ML weights. It's a big set of floating point numbers which is hard to get any meaning out of. But experts believe that techniques might be developed in the future to figure out based on the numbers what data must have been in the training set (and hence your photos). Nobodies managed…

No experience with NN, just basic understanding of these things.

If you train a model A) cats, one for B) dogs and one for C) dogs and cats. Is there any research which asks if you can do a network subtraction C-B = D, and get a sensible network D than can loosely do what A does?

Re: What is triald and why is it taking so much disk space?

#96

Post by morganvachon »

jeroenhd wrote: post #24
okamiueru wrote: post #14

Earlier quoted context omitted.

Purging snap has so far been a 5 minute do-and-forget operation, all the way to 21.10.

Sadly, Ubuntu is moving Firefox to snap in the upcoming 22.04 release. I think I'll be switching to either Manjaro or plain Arch before 21.10 becomes unsupported.

Unless you rely on systemd, try Void Linux. It has a sane package manager and a very active, friendly community. It requires a little legwork to get to the same polish as Ubuntu and other consumer-oriented distros, but it is rock solid and reliable.

Otherwise be careful with Manjaro; it is a noble project with great goals and some great implementations (and a stellar community!), but they tend to go a bit overboard on customization for customization's sake which has introduced breakage in the past. Nothing that isn't fixed within a day in my experience, but stable it most certainly is not.

There's also Mint for a more traditional and conservative take on Ubuntu's method, as well as MX Linux which shows off how good Debian can be in the right hands.

Or go crazy like me and start using OpenBSD as your main workstation provided your workflow can adapt to it; I'm on 7.0 on this machine with Firefox 95 and literally everything I need is here.

Re: What is triald and why is it taking so much disk space?

#97

Post by pjmlp »

SanderSantema wrote: post #69
pjmlp wrote: post #20

Earlier quoted context omitted.

Indeed, only to learn about all the programming languages, libraries and communication protocols used in the system, and then they might understand what is happening.

Whereas on closed source you’d have to do all of that and probably a good deal of reverse engineering on top of that. That’s unfortunately not something which it necessarily makes any easier, but rather a lot harder.

No post body was provided.

Re: What is triald and why is it taking so much disk space?

#98

Post by londons_explore »

mhitza wrote: post #95
londons_explore wrote: post #88

Earlier quoted context omitted.

The model fine tunings can't immediately be used to say "this person has pictures of cats". They effectively are just a big diff of the ML weights. It's a big set of floating point numbers which is hard to get any meaning out of. But experts believe that techniques might be developed in the future to figure out based on the numbers what data must have been in the training set (and hence your photos). Nobodies managed…

No experience with NN, just basic understanding of these things. If you train a model A) cats, one for B) dogs and one for C) dogs and cats. Is there any research which asks if you can do a network subtraction C-B = D, and get a sensible network D than can loosely do what A does?

Try it... but I predict you'll get a model that doesn't work at all...

Re: What is triald and why is it taking so much disk space?

#99

Post by greggsy »

sneak wrote: post #86
jtbayly wrote: post #82

Earlier quoted context omitted.

Do you have any evidence this is what’s happening? Why should I trust your declarations?

https://web.archive.org/web/20210903214920/https://www.apple... (linked from the above-linked wikipedia article as citation for the clientside scanning feature documentation on wikipedia) https://web.archive.org/web/20210903214920/https://www.apple... Apple said they were doing it. Apple never subsequently said they were not doing it. (They did say some things that implied that they MIGHT not do it, designed to make…

This has been discussed ad nauseum. The client side code isn’t looking at your files, it’s looking for a signature.

If you don’t have the signature, it won’t do anything.

I’m a privacy advocate, and this approach actually increased my trust as it was a more palatable option than allowing unfettered server side access in response to LEO requests (which happens in extreme cases, but not on a routine basis.)

Re: What is triald and why is it taking so much disk space?

#100

Post by bunchspoiler »

mschuster91 wrote: post #8

Jesus I hate being A/B tested without my consent. Particularly, why is Apple of all places doing this, given their otherwise strong track record of standing up for their users in regard to tracking and other bullshit by major players?!

While the Trial framework is capable of doing A/B testing (and historically started out for that purpose), it is fundamentally an asset delivery mechanism, and is mostly used for that (i.e. I suspect that most of the domains using Trial currently have no "B" to pair with their "A").

And some of the differentiation that IS being done is not A/B testing in the usual sense of the word. i.e. if one particular hardware model needs a different version of an asset, those models can be assigned to a separate set, but that does not really make it an A/B test.

Post reply on HN