I was surprised how hard it was to stop the Python transformers library from phoning home to Hugging Face. I set HF_HUB_DISABLE_TELEMETRY=1, and when I called Wav2Vec2CTCTokenizer.from_pretrained I explicitly passed local_files_only=True, but still I got got a warning about not having a valid HF_TOKEN. It wasn't until I stumbled upon HF_HUB_OFFLINE=1 that I'm somewhat confident that I'm not making outgoing connection…
Does something like Little Snitch catch these to help find the things doing hidden shenanigans?
Do_not_track
171–179 of 179 posts
Re: Do_not_track
#172I personally do not use this. The reason is quite simple: I do not want to give out ANY information to external sites. Meaning, they could want to group me into "wants to be tracked" and "does not want to be tracked". I expect a general content blocker, which ublock origin is, to protect me from any malicious external actor, including horrible UI, such as nowadays google search has. I mean, just make a regular google…
Totally agree! I've built a product doing exactly this in my previous job. I'm building something new & similar now, but much better :-)
Re: Do_not_track
#173My projects could genuinely benefit from telemetry as I have no idea about usage patterns and my community (mainly artists) is not famous for maintaining a close dialogue with software developers. I haven't bothered because a) opt-out risks a backlash and b) opt-in affects the data so much it becomes useless (much smaller sample and probably self-selecting a certain type of user) Skimming the comments here, it seems…
Telemetry only tells you what users do, not why and doesn't explain their mental models. Try asking directly: open a discussion board (for example Github's Discussions) and encourage them to post about aspects of the software they found puzzling/annoying/inefficient. Take 15 minutes a week to go through the posts to see if anything attracts your attention.
People only bother when something has made them really angry about something and need to vent.
This is why default analytics is the correct option. It gets the average people who don't care about forums and usually won't even bother to change many of the settings. The crowd who doesn't open HN first thing in the morning.
Re: Do_not_track
#174Earlier quoted context omitted.
Anonymous telemetry is allowed – and I don’t have a problem with that.
Unfortunately there's no such thing as anonymous telemetry. There are multiple techniques to re-identify scrubbed data, and some [seemingly innocuous] data is inherently identifying. https://techcrunch.com/2019/07/24/researchers-spotlight-the-... | https://www.eff.org/deeplinks/2023/11/debunking-myth-anonymo...
Not every company will do it well. Simpleanalytics.com seems to be one of the better ones.
But it’s still way better than the alternatives which don’t even try to be anonymous.
Re: Do_not_track
#175My projects could genuinely benefit from telemetry as I have no idea about usage patterns and my community (mainly artists) is not famous for maintaining a close dialogue with software developers. I haven't bothered because a) opt-out risks a backlash and b) opt-in affects the data so much it becomes useless (much smaller sample and probably self-selecting a certain type of user) Skimming the comments here, it seems…
Opt-in data is "useless" That's one I have not heard before Useless for what Targeting a certain "type of user" perhaps "I get the distrust of large corporations and other obvious bad actors - but the blanket cycnicism for all telemetry here is kind surprising" There is effectively no way for a user to determine whether an actor is "bad" or "good" and that definition may vary depending on the user The user cannot ver…
No offense, but if that's the case, you are very new to the discussion. It's been pretty well-documented that opt-out provides orders of magnitude more useful reports than opt-in.
For the best example: Factorio, a game with an almost-exclusively-technical playerbase and extremely well-regarded and community-friendly dev team, which already had a ton of people writing good bug reports on the forums, [fixed 12 crash-causing bugs](https://factorio.com/blog/post/fff-231) within two days after making crash reports automatic and opt-in.
And if it has that much impact for Factorio, you can imagine how much bigger the impact is for non-technical software.
Re: Do_not_track
#176My projects could genuinely benefit from telemetry as I have no idea about usage patterns and my community (mainly artists) is not famous for maintaining a close dialogue with software developers. I haven't bothered because a) opt-out risks a backlash and b) opt-in affects the data so much it becomes useless (much smaller sample and probably self-selecting a certain type of user) Skimming the comments here, it seems…
I’m sympathetic to both the default distrust and to devs like you who want telemetry to improve their software and won’t use the data for anything else, but it is because of bad actors and enough dark ad patterns that we just can’t trust companies to play nice, and it’s too difficult to expect people to scrutinize each and every app or site individually. So I get why the default assumption is nefarious behavior. But…
That's what I do in my apps. And it turns out, that actually increased the quality of the bug reports I got, because users were more invested and willing to cooperate.
Re: Do_not_track
#177Re: Do_not_track
#178My projects could genuinely benefit from telemetry as I have no idea about usage patterns and my community (mainly artists) is not famous for maintaining a close dialogue with software developers. I haven't bothered because a) opt-out risks a backlash and b) opt-in affects the data so much it becomes useless (much smaller sample and probably self-selecting a certain type of user) Skimming the comments here, it seems…
Opt-in data is "useless" That's one I have not heard before Useless for what Targeting a certain "type of user" perhaps "I get the distrust of large corporations and other obvious bad actors - but the blanket cycnicism for all telemetry here is kind surprising" There is effectively no way for a user to determine whether an actor is "bad" or "good" and that definition may vary depending on the user The user cannot ver…
https://factorio.com/blog/post/fff-231
Here, the developer is trying to infer user reasoning and intent
Curiously, he omits the possibility that users would prefer not to send the data
When in fact this is exactly what users indicate they prefer
He pretends that "Don't send" is ambiguous, for example, that despite clicking "Don't send" users actually don't care if data is sent
But there is nothing here that indicates users wanted to send data or that they do not care
Software developers can obviously do whatever they want and they can act against the interests of users
This includes ignoring or explaining away the preferences of their users ("Don't send") and engaging in speculation about user reasoning and intent
The developer here seems dismissive of users' reasons for clicking "Don't send", even though he does not know the reasons and can only speculate. At the same time he expects readers to take his reasons for collecting crash logs as justified. Then he unilaterally decides to remove user choice (the "Don't send" button) and substitute his own choice (send data) as a default
Perhaps lack of developer pre-release testing and quality control is relevant to this discussion. Alas, the problem is framed as one of data collection and user consent where the "solution" is making data collection surreptitious and making "consent" uninformed, implied
Re: Do_not_track
#179The issue is that it is not enforced. My version of My IP will tell you if 'Do Not track' and 'Global Privacy Control' are set by your browser but it is up to the website to honour your requests. Check if your browser is sending them by visiting: https://fshot.org/utils/myip.php
That's great, but isn't DNT deprecated?