Go telemetry could've been useful

[@fasterthanlime]

Erratum #1: the hours for Ramadan in France aren't "9pm to 2am roughly", they're "from sundown to sun-up" (or rather, the fast is "from sun-up to sundown"), so it really depends on the season. Erratum #2: I just remembered this video is on KZbin, so I was just kidding when talking about how Google works haha.. Erratum #3: Errata is the plural of erratum, as pointed out below!

[@andrewdunbar828]

I was hitchhiking in Turkey in Ramadan once and everybody knows the exact time of sunrise and sunset for every day. I think they even have phone apps for it. Everybody is starving and arrange to meet up and have a big meal together the minute sunset arrives.

[@tsukuyomin]

Erratum #3: Errata is the plural of erratum ;)

[@0x07e4]

France's smart energy meters (aka Linky) send metrics over power-line communication to local hubs which forward them over mobile networks to datacenters. The energy meters don´t produce any radio waves. The whole radio wave controversy is even more dumb when you know that. But as you pointed out, they managed to shift the focus from data collection and its consequences to radiophobia.

[@fasterthanlime]

​@@0x07e4 ohhh that's how they use power-line. I saw them mentioned but was confused. I agree, the controversy feels even stupider now.

[@Kabodanki]

And after sun down, we eat as much as we can, litteraly stuffing our face with sugary and oily stuff.

[@trejkaz]

Nothing like getting an ad for sleep deprivation when you're already a heavy user of the product.

[@fasterthanlime]

100% accuracy ad targeting, every time

[@ehfoss]

All that data and KZbin just gives me ads for things I already have.

[@junelawson6708]

I guess the thing with telemetry is it requires trust, and nobody would or should trust Google. Opt-out telemetry would be helpful, and there isn't any reason to think Google could misuse it as described, but ultimately, it's hard to get around the lack of trust.

[@KaneYork]

Google is excellent at doing boiled frog maneuvers. They've practiced.

[@versacebroccoli7238]

Do you think Amish people are allowed to study computer science? I mean the electricity and computer are kind of superfluous anyways. Like can an Amish person learn Lambda Calculus or is that against the rules?

[@somebody_on_the_internetz]

But is math bad?

[@fasterthanlime]

I'm debating whether this counts as a comment that would force me to quit KZbin and join the Amish. (But I'm also wondering about your actual question now)

[@funkintonbeardo]

I've read that Amish are starting to use phones, computers and the internet. I think the justification is that they are not inherently bad and ok if they are used for good things like buying new Amish bibles and not for checking out hot Instagram models So that depends on morality of lambda calculus

[@AlexanderPayneMyrrlyn]

@@fasterthanlime i live in amish country and had a mennonite (amish but allowed to go on the computer) department chair in uni. they are perfectly fine using modern technology at work; they have offices and powered carpentry tools and websites and can be paid by credit cards. the objection is to technologies that sunder the family and promote individualist isolation, which computers, forced-air hvac, and automobiles tend to do tldr they can study how the computer works they just have to leave it in the office

[@versacebroccoli7238]

@@AlexanderPayneMyrrlyn Thanks! I was genuinely curious.

[@gr3ut]

“Talk to your doctor today to see if sleep deprivation is right for you” made my day. Keep up the good work and take care!

[@fasterthanlime]

You too

[@casperes0912]

This video must make a lot of money with all those sponsors.

[@fasterthanlime]

It's like they say: money never sleeps!

[@andrewdunbar828]

Useful for lobbying

[@AnupDhakalSharma]

Interesting points. Some people raised issues about it being auto opt-in for future revisions of the policy as well, which meant you would have to not only trust the current team and their motives, but also any future team members or their boss and their motives.

[@SaHaRaSquad]

Exactly. This is the same as with surveillance done by the government - it may even have good intentions but you don't know who's going to be in power in 5 years and what kind of law changes they're going to make to "protect the children" or "fight terrorism". Power gets abused, period.

[@eMbry00s]

An evil future team could just as well silently change it to opt-in. There's a small difference in how that would be perceived compared to changing the terms of the policy, but in practice I think the effect would be the same: not much would happen.

[@evandrofilipe1526]

​@@eMbry00s this wouldn't be a problem for Foss however

[@owlmostdead9492]

You missed that they explicitly can change what they collect at the whim of their will, Google is THE example that proves the slippery slope is NOT a fallacy

[@NathanHedglin]

Exactly this! If they're in charge of what changes and when, that's a huge risk.

[@dieSpinnt]

Every good intention, such as making life easier by collecting data or improving the quality of life, is soon perverted to the opposite. Where data has been collected and is now available, no matter what kind, covetousness also arises. The "black coat" (secret service agent), thief or do-gooder is already waiting around the next corner. You don't even need a computer to do this. From August 17, 1935, for example, Jewish associations had to hand over their lists of members in triplicate to the state police stations controlled by the National Socialists. This was done under threat of punishment, but often also voluntarily and with a willingness to help. According to the motto: The Germans are precise and hard-working. Everything has to be in order. This is completely normal and what should happen? ... What horrific and unimaginable crimes would happen, we all know these days. (Edit: BASED ON THAT VERY DATA!) Every democracy has the protection of privacy and the free development of the personality of its citizens written down and anchored in their constitutions as a valuable asset. So in addition to the immensely high risk of misuse of any kind, this data collection frenzy is also undemocratic. One can't say it often enough: Once the data has been collected and then circulated, it is irretrievably too late. "Data Frugality" [1] is the goal! Resist the beginnings. [1] or data economy? ... meant is the German term "Datensparsamkeit" where this designation goes back to the Chaos Computer Club in the public discussion on data protection and various spy affairs and data leaks over the decades. Anyways, thanks for your effort in making those videos for us. I admire your courage that you don't shy away from researching the insane hamsters yourself! (The logo of Go shows the founder and chief developer ... a crazy genetically modified hamster working at google, or did I get that wrong?). Thanks for the video, Amos!:)

[@pluieuwu]

interesting take on the subject of telemetry! personally i'm still biased against opt-out telemetry - been bitten too much by hidden switches in obscure corners of an app to turn the telemetry off, which sends literally everything about your system by default - but regarding the Go telemetry situation, can't they just give you an Y/N choice on installation? sure, lots of people will still choose N but at least the venn diagram won't be a circle ;3

[@oglothenerd]

I am creating a programming language that combines Rust's safety and Zig's insane power, and one thing I want to do is make the package manager and compiler never need internet access to function. (The only time you would need internet is to add dependencies, and even then, they would be cached so the package registry could be nuked with no consequences.)

[@shirshak6738]

the problem with Google is they start with novel idea and later ruins it. So, opt in telemetry is always better. That can ask if user want to send telemetry data during the installation of Go?

[@PwnySlaystation01]

The problem with this idea that "telemetry would solve this" is that I can't think of any areas where this has obviously happened. Windows, for example, has steadily increased its mandatory telemetry and data collection (and still is). Yet we can't go three months without major problems/incompatibilities introduced by patches released by Microsoft, that had been collecting MASSIVE amounts of telemetry showing these issues through the insider program for months. Does anyone have an example of software that, once telemetry began being collected, the software improved in a significant way? This is a serious question. In short, the theory that telemetry can improve software/user experience is pretty clear. It should. But I don't see much evidence that this is true. All the downside still exists though and gets worse all the time.

[@invitapriore]

Well, in exchange for garnering extreme distrust when presenting a proposal like this, Go benefits from lavish investment from the organization that has done a lot to earn that mistrust. That sounds like a pretty good deal to me, and it seems pretty silly for anyone to be upset at not also getting the cherry along with their enormous sundae. It's also not a credit to your argument here that the statement "big companies don't really like breaking the law" is flatly ridiculous. Big companies happily break the law all the time, as this particular big company has with its participation in e.g. wage fixing and retaliative firings against union organizers. They made their bed here, and everyone involved should accept that graciously instead of wringing their hands about the logical inconsistencies being marshaled by the anti-telemetry brigade. That's not the point, and it ought not to be.

[@inertia_dagger]

I think what they could do is add a message banner in the terminal, asking users to enable telemetry to allow for bug fixing, or suppressing that message in a config

[@hacatu]

Yeah, gdb does that with downloading debuginfo; for example, on arch linux it will say "This GDB supports auto-downloading debuginfo from the following URLs: ... debuginfod ... archlinux" and explains how you can permanently specify a choice in gdbinit

People don't enable it. That's work people who don't care won't do, and the projects are worse off. People who care however will expend the one edit it takes to disable it.

[@hamu_sando]

Go features can be deprecated and removed regardless of whether people are using them; telemetry isn't needed for that either, people will complain when something breaks or just fix it and move along. If it was going ahead then I would prefer if they collect the info locally then present it as a list of tick boxes with the description and data so you can see clearly what would be sent and could untick anything you don't want to send. It should store preferences instead of asking each time and when new metrics are added it should either request that the user selects a preference or defaults to 'do not include'. If developers want control just give them as much control as they want and guarantee it will remain that way.

[@theondono]

For me at least, hearing this proposal from Google sounds like hearing a kid beg his parent for a new gaming computer for doing "homework". Yes, a new computer would probably help with homework, but I don't trust for a second that this is why you're asking. Even if the current devs are angels who care deeply about privacy, I don't trust the internal structure of Google to enforce this in the future. Monitoring if the go compiler is sending anything at all is trivially easy, monitoring if it's sending *just the right things* requires a Phd in data collection.

[@ateijelo]

I wanna address two things: - With Telemetry, Google would know that no one is using a feature and they'll be able to remove it: yes, that's true; but I don't want my toolchain phoning home; not weekly, not once a year, not ever; Google can still deprecate features, have the tools throw warnings if you use those deprecated futures and at some point just remove them. - If you don't trust Google, how do you trust the whole toolchain you just downloaded from their servers: because it's open source, and while *I* don't audit the code, I know the code *can* be audited, and therefore people do it. I know, because it's open source, that, if Google tries to sneak in some nefarious piece of code, *people will notice*. And I know that, if I was so inclined, I can build the toolchain from source. And some distros do that as policy, e.g. Debian, Gentoo, etc. E.g., if I install the Golang tools from Debian, I have some confidence on their safety. All these factors combined is what makes me trust the tools.

[@Marcus001]

This is a very interesting take that I’ve never considered. I feel like some people are going to read the title, auto-dislike, and leave without watching, but the points you made are hard truths not everyone wants to accept.

[@guest7329]

Is Go has bot that compiles all go packages on github against new release of compiler? It's help a lot more than 2kb/year of telemetry data

[@sillybuttons925]

The sponsorship segment

[@TehGettinq]

general issue with telemetry and data collection is you need to trust that the data collector will be clever enough to secure the data.. it's not always just a matter of a perception of their motives (which is another topic, implying we can infer their motives is dubious at best). nukes good for energy when in good hands, bad when mojo jojo gets his hands on it..

[@paulsanchez5030]

Everyone agrees telemetry could be good, but no one, absolutely no one will ever trust google on anything like this.

[@fasterthanlime]

Here's what I don't get: if you don't trust them when they come out and say it, how can you trust a whole toolchain you download right from their servers? I just think it's a weird place to put your foot down.

[@funkintonbeardo]

​@@fasterthanlime the thing I love about Rust is that it's not Go

[@DarrienGlasser]

Until there are other proper compilers, folks may need to download it for work, school, or even just a PR on an open source project. I’m no Google fan, and yet I have Chromium on my laptop for those rare instances where a site has decided not to support both Firefox and Safari.

[@casperes0912]

I don't know. People also riot about all open source projects including telemetry for benign reasons. Some folks are just absolutist in dislike of telemetry

[@bugs-bunny-k6g]

@@fasterthanlime because other developers who read the code will point it out?

[@davidboeger6766]

The thing is, you touched on the problem yourself when you said that people should better understand the world they live in. Ideally, yes, but it is simply impractical to expect every human on Earth to fully understand everything in their lives that has been built up over the course of human history by billions of other people. I'm an educated software engineer and there's just no way I'm going to learn rocket physics to understand how satellites are launched before I use the GPS features on my phone. The ethics get even more complicated when you consider other factors like socioeconomics, political persecution, espionage, etc. Sometimes it makes sense for society to push certain defaults, like vaccines for kids, in order to save millions of lives. Note that there are still plenty of anti-vaxxers, and that at the very least, government programs are in theory less motivated by profit than corporate ones. Also, for what it's worth, why is the responsibility for making Golang better on users and their data? If Google wants to make a better language, they can make one. There are several more beloved languages that have been created without opt-out telemetry. I get that for loyal Golang users, they might stand to benefit personally from other people being forced into telemetry, but that's a really weak justification for data collection if you ask me. I don't just get to collect private data (I'm considering private data a superset of PID) on all of my neighbors just because I might enjoy my day a little more. I would much rather err on the side of Google not collecting more data. Indeed, I believe your point about Google already having all of the data further justifies this push-back, rather than supporting further erosion of privacy. Edit: I thought I would add that while you make an excellent point about progress. I think it's worth considering the progress in question. If Google was developing life-saving technology, for example, that'd be one thing. The main "progress" they would achieve from their telemetry is catching up to other languages so as to make their ad-serving systems more robust. It strikes me as odd to suggest that we need more capitalism literacy, but then to ignore that capitalism has a strong financial incentive to paint all profit-generating investments with the broad brush of "progress". The simple fact is that pushing back on Golang telemetry does nothing to hurt real progress that improves our lives, and if Golang as it exists today is hampering said progress, there are plenty of great alternative languages to choose from. Losing telemetry is Google's loss more than society's, and I think most people are fine with that.

[@kiuxex4875]

can't wait to see ads in my compiler

[@spoofilybeloved6729]

man, your videos are truly and thoroughly fantastic. nothing like this exists in the tech space. please keep it up.

[@kelownatechkid]

Very interesting, thank you for continuing to make these videos! IMO, telemetry should be opt-in but easy and clear to do so. Ceph is probably my favourite FOSS example for this - it asks you when you first sign into the dashboard if you want to contribute and it only takes a second to understand whether you want to do it or not.

[@nicholaswood3250]

I was never against Google wanting to collect this data, or even the proposal itself. I think the way they wanted to implement the feature was very well thought out, and I actually hope they go back to the drawing board and come back with a revised proposal. I am absolutely not anti-telemetry, or even anti-this-proposal. The only problem I have with said proposal is entirely that the feature is turned on and built-in by default rather than a feature flag that volunteers can turn on to help the language designers. This causes a lot of problems, not only for hobbyists (who are probably not really impacted by this telemetry to begin with), but also for companies that are evaluating Go, or using it in their tech stacks. I think unless the Go authors are *extremely* careful about how they develop this sort of telemetry, it'd be exceedingly interesting for Go to go from being a viable option for writing software to a Google-world-only language when engineering managers start getting hackles raised about possibly leaking intellectual property due to the compiler phoning home about the wrong sort of thing. Again -- not something the proposal allows for -- but the appearance of impropriety may develop. I insist, and continue to insist that language designers should be fully capable of improving their language without always-on dragnet telemetry policies. It's not right to just turn this sort of thing on without asking permission, and doing so because you think your users are too X/Y/Z pejorative to want to help you out should give people second thoughts about being part of the community in the first place.

[@nicholaswood3250]

Adding a follow-up because the edit button isn't working. I am also hoping that I'm not coming off dismissive of fasterthanlime's argument. I don't even disagree with him -- some people act like Google is trying to record their keystrokes, or something, and that's definitely not what is happening. I just think I'd point out that I'm very much for allowing Go's authors to have better ways to instrument code, and make their work easier. I just think the whole thing got off on the wrong foot, and I think the only way to build enough trust that a feature like this works at all, and still leads to a healthy community in industry and elsewhere, where people don't view Go as another way Google is trying to gain leverage over their competition (again, probably unfair, but that's how people *will* see things) is to make it opt-in and behind a feature flag by default.

[@alexchichigin]

Simply want to express my love and support. You're a terrific human being in the best way possible. Please, take care and get more sleep. The humanity really needs you to stick around as long as possible, even if it doesn't realize it. :)

[@michaelutech4786]

Good & bad tech + data collection: I think there is a simple criterion that decides about whether data collection is good or bad in general and that is who owns the data, before and after it's collected. I guess we can agree that uncollected data belongs to the subject of the data. That's what privacy means, you own data not collected about you. If someone collects data about you, they gain some value (the knowledge about you) and you lose some value, the privacy. Such a transfer of values is usually or at least should be subject of supply and demand or negotiations. Every other kind of such transfers has a smell of theft. If Google would collect data without taking ownership over that data, f.e. by publishing the collected data, then there is still a transfer of values, but then Google produces at least a counter value for the information they obtained. Everybody else could access the data. That would feel like a fair exchange. If Google just collects the data, maybe without asking, probably without informed consent and maybe even without giving users a choice, then it is clearly bad, no matter what Googles original intentions are, no matter whether they succeed to preserve anonymity and are careful about how they go about it. What is bad about it is the disparity of power or the factual monopolies tech companies, especially the big ones, have. And here it does not even matter whether these companies are benevolent or malicious, the sole fact that power is distributed so unevenly makes any development that increases the disparity bad, just as all monopolies are bad in an environment that is designed to regulate itself by competition. I remember the CEO of Sun saying that there is no privacy, get over it, before Sun became Oracle and long before Google abandoned "don't be evil". My immediate reaction was to think: "Who are you to tell me that? It's my privacy, not yours!". Of course he was right in the sense that Tech is stronger than me. They can just take my privacy, my money and much more. What they can't take away from me is the right to be upset about this and to call it out as the theft that it is and trivially as something bad.

[@swapode]

I have to disagree on a kind of fundamental level. Telemetry being opt-in is always the right way. So much so that it should be common law and in some places already is. If the tech industry again throws a hissy fit like it did for privacy laws regarding cookies etc being opt-in, so that now you have to click "decline" before you get to a website's content. That's on the industry. When an overwhelming part of the tech industry is built on an immoral foundation, it's not the morals that need to budge to a new reality. It's not the users' job to opt-out, it's the industry's job to convince the users that a particular data point is worth being collected. And so far the industry has done everything to say that it's not.

[@JusticeNDOU]

one time long ago i was thinking of how much electric companies can ccllect about the population, and how this data the government can use for easy service delivery combined with AI the sky is the limit the government can be highly effective in addressing the needs of the people. but of course we live in a world of freakouts.

[@cthree87]

Your timing is impressive.

[@irlshrek]

I respect the hell out of you for saying the right thing. that being said...im all in on Rust

[@fasterthanlime]

I am too! but do I resent being an automatic ally to folks who just enjoy shitting on anything Go-related. That's not what I did and I'm not automatically ok their side, so I thought I'd take the opportunity to emphasize that!

[@fabianbecker6266]

You made the first pro-telemetry argumentation that might change my mind on the matter.

[@moy2010]

Justifying giving more data to Google just because they already have a lot of data is...

[@SaHaRaSquad]

Comment about the Amish

[@__noob__coder__]

after long time I saw some youtuber having quality sponsors !

[@funkintonbeardo]

I loved the outro

[@youtubesins6099]

No idea what you talking about, but I did have a nice conversation with the Alpaca at the back!

[@ZiggyGrok]

Frankly, I'm just sick and tired of surveillance everywhere I turn (Google or not). Not to mention that the pro-opt-out camp claimed it was super easy to turn off, well documented, etc., and the flip side of those arguments are now the same arguments why it's never going to be turned on, even by people that'll be ok with it. Hmmm

People who are ok with it often just don't care. Them enabling it is effort they won't expend. In situations where I care, I'll however definitely expend the effort to turn it off. Hence, opt out.

[@ZiggyGrok]

@ except one of the arguments is "need to know about it" which cuts both ways. I don't expect `ls` to spy on me, nor my compiler toolchain. If you're really worried your opt in rate will be too low, make everyone choose one way or another.

@@ZiggyGrok Yes, a clear choice during install time (with follow ups if the collected data materially changes) makes sense. I think "spying" is an exaggeration that helps nobody, though.

[@hitchen1]

As far as I'm concerned, it's a matter of consent. It doesn't really matter if it's useful for the developer or not, or the degree to which it collects data. They should ask, or assume the answer is no.

[@zachend2750]

Rust did telemetry and put hidden folders on your root and no one said a thing

[@nWestie]

This video was absolutely *chefs kiss* It is refreshing to hear well laid out, balanced arguments about something like this; its good to be reminded from time to time that Google is not evil; cold and kinda creepy maybe, but there are also real people there doing good work. Hope you can get some sleep, I heard it's good for you :)

[@Dygear]

Actually laughed out loud at several points!

[@fasterthanlime]

Haha thanks Mark! I'm glad some of the jokes landed.

[@ChristophVonBagel]

How does rust do this? Do they just collect bug reports ? What if they did telemetry ? Would that hurt rust ?

[@fasterthanlime]

The Rust project has crater, which compiles all the crates published to crates.io iirc - but I also think it would benefit from telemetry (which it doesn't do at all right now)

[@AmeSoftware]

For my part, I consider that opt-in is the best solution, nobody wants their data to be taken without their consent. Asking is always the best, everyone can decide what to do and I'm sure most people will gladly provide their information, I don't really see the problem. Asking is not that difficult.

Telemetry is a very contentious topic in Open Source projects. It's been plaguing my professional life for years: we could do better if we knew. But users complain A LOT, not always justified. Mostly they complain from a fully instrumented app on their fully instrumented devices.

[@fasterthanlime]

> Mostly they complain from a fully instrumented app on their fully instrumented devices. Yyyyep. I mean this is borderline "we live in a society" stuff but.. it's also true.

[@FrankRehwinkel]

Very nice. And funny too! Thank you.

[@technologyondemand4538]

so what happens if you see multiple comments about the Amish? do you still consider the first one as a single comment you saw and will thus join the Amish? probably. i guess 2 does not invalidate 1. but then again, "a single comment" is not the same as a group of comments (with more than one comment about the Amish). but the first one you see is still you seeing a single comment about the Amish, unless there are multiple comments about the Amish in view simultaneously. so i guess by commenting this you are obligated to join the Amish unless there is at least one more such comment within view as you are reading this one... oops

[@fasterthanlime]

I can't fault your reasoning. As you can see there are MULTIPLE comments about the Amish, so I think I'm safe.

[@GiovanniCKC]

*inhale* But the am--

[@fasterthanlime]

Don't even think about it

[@zeratax]

reminds me a lot of the audacity situation ^^

[@cpli7783]

about the amish

[@somebody_on_the_internetz]

Rust man stands up for Go thingie bad?!

[@fasterthanlime]

Rust man is also made of people

[@somebody_on_the_internetz]

​@@fasterthanlime all kidding aside, I hope you have a good night's sleep. Good night Amos :) and thank you for great content 👍🏻

[@Hector-bj3ls]

​@@fasterthanlime How many people 😳

[@recklessroges]

@@Hector-bj3ls You need at least six to reach an inteligence that is fasterthenlime. /jk

Great video! On another note, your voice seems not to be in sync with the video. Not sure about that, but it felt weird. FYI

[@fasterthanlime]

Huh, it does feel weird. That means something went wrong very early on when editing the video 🙃

[@DavidOlofsson]

My only complaint about this video is that I can't like it twice, but if I understand the engagement metrics a comment is about as good.

[@kmichal456]

What about the Amish?

[@ac130kz]

just make the thing opt-in, lots of people will contribute for sure

[@NathanHedglin]

Exactly.

[@fasterthanlime]

It's going to be opt-in now, and thus useless - which is covered at the end of the video.

[@felixjohnson3874]

I'm going to hesitantly, VERY, hesitantly agree here, but I don't think hating telemetry is a bad response. The biggest danger of progress isn't "whoops we did this thing and accidentally found out your religion" it's, "you know, over the past 2-3 decades we've been doing this more and more and, in hindsight, we really shouldn't have". I don't care if some file somewhere said some people were muslim, just like I doubt anyone in this section would care if Christians had the same treatment; what I DO take issue with is the perpetual micro-managing of our lives enabled by large scale data collection. That same data could be used to better-than-chance identify who has an electric car, who has an electric heater, who has an electric stove, how many people are living in a house, if you mine crypto, etc. by looking at timestamped power usage records. Now mining btc is illegal because the government wants to crack down on private payment methods, and you're on the list. This wasn't a single "oh, we fucked up", it was the direct result of constant high-fidelity data aggregation which enabled a corrupt government action to be enforced with a resolution it shouldn't and wouldn't have. DO be fearful of technology, because if it ain't broke, fixing it might break it, or something else down the line. That fear however must be balanced by the gains it could offer. If it doesn't offer anything of value, don't do it, it might cause issues for no reason. Also, the amish. No-one wants your "muh capitalism bad" political schpeel

[@ankur-dhama]

This video was sponsored by the previous video.

[@YTCrazytieguy]

I bet you put the Amish bit in to bait people to leave you comments... It worked!

[@fasterthanlime]

It was either that or everyone would've "well actually'd" me with "you CAN opt out of the world, look at them!"

[@SolarLiner]

I think there's also cognitive biases at work when talking about telemetry -- nobody hears about telemetry success stories because it's not newsworthy. But we do have a treasure trove of failures in data collection. That weights in the reputation if it all.

[@alwin-ob9ld]

Is there any way to support you with bitcoin? I can’t support with traditional currencies for privacy reasons.

[@fasterthanlime]

I don't accept Bitcoin at this time, sorry.

[@nitproject5193]

There is nothing bad in telemetry if it'll be possible to easilly and completly turn off telemetry

[@Cornyfisch]

6:37 Haha why that accent of all the possible accents you could've choosen?

[@fasterthanlime]

I'm subtweeting a friend in particular.

[@evandrofilipe1526]

I think it's somethng to do with you saying why do people need to be protected when they are not really protected. They are just not exposed. Also, I wouldn't trust google with any data. I always assume something shady is going on.

[@vladlu6362]

Can people stop confusing opt-out and opt-in telemetry!? Opt-in: you CHOOSE to get INTO the program. This is good telemetry... Opt-out: you CHOOSE to get OUT OF the program. This is bad telemetry. Opt means choose, people.

[@fasterthanlime]

I think you're confused: opt-out is when you piss off everyone, and opt-in is when your data is useless. (But seriously, this is covered in the video and idk who you think is confused about it)

[@NathanHedglin]

​@@fasterthanlime Google doesn't run enough Golang themselves to get enough data?

[@fasterthanlime]

​@@NathanHedglin oh they definitely know about /their/ use of Go, but it's being used in a lot of other places / other ways, which the Go team also cares about.

[@vladlu6362]

@@fasterthanlime it was mostly a thing for the remainder of commenters of your video than you :). See lots of these tiny mistakes on the comments, like "auto opt in", whatever this means. You did get it right, just a random explanation from a random commenter to make sure people don't get confused over both the content of the video and future content about telemetry.

[@Cadey]

but the amish xddddddddddddddddd

[@fasterthanlime]

Cadey plz

[@misium]

The "you already sign all your data to Google" argument is particularly unnerving. What a straw man.Rhis doesn't belong in an intelligent dispute. You can flip it over: if google had all my data already, why do they need more? Anyways, however well-thought out this implementation claimed to be, the principle is flawed and ripe for abuse. You said it yourself: reading through and understanding its advantages takes a lot of effort. It would take a lot of effort to discover faults in a more mischievous scheme.

[@KernelUwU]

This video was sponsored by

[@tauon_]

me! :3 (not really)

[@inertia_dagger]

me (yes, really)

[@CompletelyCovered3]

At least consider becoming Mennonite instead. Odds are good that we can keep this thing going after your conversion. 👍

[@fasterthanlime]

Fun fact, my grandma attended a Mennonite church!

[@dorktales254]

Who even are the amish?

[@tauon_]

they pretend it is the 1800s (no xbox!!!)

[@fasterthanlime]

PSP okay?

[@dorktales254]

@@tauon_ no xbox!!! 😨

[@recklessroges]

I like go-lang telemetry because it helps more people opt out of go and into Rust ;-) lol

[@arandompersonontheinternet7591]

Amos, I am going to HEAVILY disagree with you. Why would it be turned on by default? It makes sense that IF you want to help GO then sure, enable telemetry and send the data. If you don't want to and just are just using the language to write your own little projects then why bother? Their github discussion intro was full of over simplistic ideas like: "Specifically, only about 16,000 reports are needed for 1% accuracy at a 99% confidence level" oh really? How convenient! Where is this statistic coming from? Manipulating this and saying "Oh we miscalculated something and now it's 100,000 or more reports for 75% confidence level" can be a perfect opportunity to change the requirements. How can the go team be sure that the data being send isn't some garbage data? Many people will opt-in to use Telemetry the same way IntelliJ asks you if you want to contribute back with the error reports, a much better way to do it than VSCode not telling you that your data is in fact collected. I still use VSCode because it's good but I want to have a clear choice as an engineer to make an informed decision

[@LeonardoVieira-pw3pw]

let gooo

[@mme725]

Holy shit, you've gotta be the richest KZbinr ever. I don't think Mr. Beast has anything on your sponsorships.

[@suncrafterspielt9479]

Join the amish xd

[@chonkusdonkus]

I had to unfollow Drew DeVault because he was constantly raving about it after the news broke. Followed him for news on Hare, not tin-foil hattage.

[@fasterthanlime]

Years ago I've made the decision to pretend Mr. Devault doesn't exist. Self-care is important!

[@arkeynserhayn8370]

A person who have opinions against mine is tin-foil hat wearer. Got it.

[@glennmiller394]

Where do you live? I want to buy the house/apt/townhouse/condo next door to you.

[@styraxgum]

The Amish

[@Mpdarkguy]

The community post about sponsors was just a meme?! This comment was sponsored by capitalism

[@fasterthanlime]

The survey I did was serious, but after negotiating for a bit, the amounts being offered are too low for me to justify permanently adding a segment to one of my videos. I'll reconsider later if my numbers start looking different!

[@Mpdarkguy]

@@fasterthanlime nice, I'm glad you're in a position to afford doing things your way

[@hypergraphic]

This whole thing seems a lot like a management vs employees mismatch. It sounds like the maintainers were thinking that telemetry will make their job easier, but for the average go developer, who was actually thinking that telemetry would make their job easier?

[@0xhiro]

Amish

[@recklessroges]

Mennonites ;-)

[@blarghblargh]

I don't know you, and I can't pretend I do, but I also don't want anything bad to happen to you. this is a couple videos where your charmingly salty takes have turned to what looks from the outside like a grim existential crisis. please take care of yourself. please give a rest to whatever unhealthy coping mechanisms you are leaning on, at least for a few weeks. please lean on your support network for a while.

[@roberthambrook150]

I will always object to any mega company making one cent off me. Pay me for my data. Problem solved.

[@0xTas]

Tools like Go are entirely free-of-charge to use, even if you make money with the Go code you write. They consider that payment enough, and they aren't necessarily wrong in that regard. Their payment to you is the fact that you can use the convenience of their platforms free-of-charge to begin with. Then it's up to you to decide if the data they will take matches their "payment" in value.

[@conando025]

Oh god I'm so sorry to tell you that. I love your videos but man was I distracted by your looks. And don't think it's intentional or that you need to change but sadly yours looks reminded me a bit to much of a German leader that shall not be named. I don't know what this comment is supposed to achieve because that's a me problem and i fixed it by treating this like a podcast

[@fasterthanlime]

Fun fact: Hitler's style of mustache is called a "toothbrush mustache", and it's about the width of one's nose. I have a pretty big nose, but I promise it's not as wide as my mustache.

[@conando025]

@@fasterthanlime yeah if it was just the mustache then it wouldn't be an issue but that combination of black hair to one site with that mustache reminded me just a bit to much of all those history lessons about the past of my country so that i was tok distracted from engaging with your excellent video

[@EzequielRegaldo]

Just no, telemetry inside programming language isnt good, uninstalling !

[@AlexanderPayneMyrrlyn]

people who write compilers do in fact derive value from seeing how those compilers operate

[@NathanHedglin]

​@@AlexanderPayneMyrrlyn so the compiler is a SaaS now?

[@0xTas]

@@NathanHedglin If you expect them to continuously *service* and improve that compiler, then yeah, why the hell not?

[@pgtrots]

If only you put as much thought into your facial hair as you do your content.

[@keokawasaki7833]

Wtf lmao

[@fasterthanlime]

I'm afraid in this case, it's not the thought that counts.

[@casperes0912]

That's very rude

[@bytefu]

​@@fasterthanlime I guess his theory is that if you concentrate all your mental powers into your facial hair, your face will be straining so much that you'll be literally squeezing hairs out of your face.

[@SagarHingalAI]

agreed, although I would like to know why you hate Go so much, I am go developer so curious to know, do you have a video on that? @fasterthanlime

[@philippecholet9484]

Amos, je te souhaite de bonnes nuits de sommeil réparateur. @Google: no I am not learning Go, but I like his thoughts on things.

[@fasterthanlime]

Merci !

[@erictrinque6513]

Amish