tl;dr: The browser attempts to learn your preference for each site automatically based on how you interact with videos. You can see what it's calculated by visiting chrome://media-engagement
Oh god, that reminds me of another piece of bullshit Chrome magic. They made a system when the browser shares signatures of forms/fields with a central server, which sends back "crowdsourced" autocompletion rules, and they've made it almost impossible for developers to stop it from doing that [0] even when it's flat-out wrong and users are complaining about inappropriate autocomplete suggestions being shown/recorded.
Over the last ~8 years developers have submitted hundreds of examples [1] of why it's a stupid feature that at least needs a way to opt-out, but Chromium devs have kept it mandatory, and I suspect it's because Google/Alphabet is somehow exploiting all that leaked metadata about what forms people visit.
P.S.: Messing with the issue-tracker (and bringing up unpleasant memories of desperately trying to find a workaround in their source-code) I find that the two linked issue about this anti-feature rank are both in the top-15 most-voted open issues. (Rank 14 and 9, respectively.)
Good to see, however this still hasn't started, it's far too late after AMP, and only affects "a small percentage" as quoted. It also doesn't give the special carousel placement in the results page, which is only available via AMP.
Seriously, good riddance. It added so much clutter to videos that I explicitly hid them any chance I got. I'd rather content creators just use the video description to communicate or edit their videos so that annotations would not be necessary.
So you read the entire description of a video before watching it and remember every word of it, including timestamps with comments to note corrections, even when there's 10 of them for a half hour video?
Annotations (in the sense of small snippets of text that display briefly to add info) are fine. What YouTube had in many cases was a poorly implemented MySpace page overlaid on the video. And as one of the parent posters said, the setting just kept turning itself back on no matter how many times you turned it off (on the video, or in your settings page).
"Many cases" is subjective and anecdotal. Maybe my set of youtubers i followed was particularly high quality, but almost all uses of annotations i have seen were to provide corrections to information found wrong in the video, or additional information learned after the video.
In some cases such things were applied quite heavily too.
And thus the point of my previous post is that for people doing serious work with their youtube videos providing corrections and such via the description is entirely untenable.
I'm sorry you've only seen spam, but cutting off another's tool because some people can't use it correctly is a severe case of FYGM.
I don't really have a "set of youtubers" that I follow and so most of the stuff I watch is from random sources. I think you're right that it depends hugely on who you're watching - random hobby / extreme sports videos seem to be particularly bad, but technical and engineering videos are usually fine.
I don't think they should take it away completely, just limit it so avoid the aforementioned 'myspace page' overuse.
You can add "cards", i.e. the popups in the top right that show a short text and an "i" button. Those can be links to videos, polls, or external links (if you're a big youtuber). However the text is limited, you can't set how long it shows, and you can have a maximum of 5. You can maybe use subtitles too, but you cannot set a default subtitle for your video.
So there are hacky half-working workarounds, but otherwise youtube just told everyone who cares about their videos to go and kindly get fucked.
Who said anything about editing errors? I am talking about things like exploratory engineering videos, where the creator is speculating about certain things that get apparent only after the video has been uploaded, sometimes with the help of viewers.
It is for example not uncommon in teardown, some done by leading engineers in their entire field, to find out what some proprietarily marked chip is a few days after the fact. This is not a “fuck up”.
Reediting and reuploading a video just to change a single (but possibly important) detail is not only a colossal waste of time and effort that could be better spent on other videos, as far as I know it will also present itself as a new video.
The Google Home definitely supports Chromecast now [0].
My favorite related thing to say is "Hey Google, play House of Cards on my TV", and it will turn on my TV and then start playing the next episode from Netflix. Perfect for when I'm walking over to the TV with my dinner.
Thanks for the info! Perhaps in the new year I should consider taking my Google Home back out of its box and giving it another shot. I really wanted to like it, I am an Android dev and feel comfortable in the Google ecosystem.
Out of curiosity, do you know if the the flow I describe works now? The link (and your comment) discuss putting content on the TV. The link says you can do:
> even control your media with commands like “Pause this episode”, without lifting a finger.
but that didn't work for me 3 weeks ago, does it work now?
It definitely works! I haven't tried muting explicitly, but I've been able to control play/pause and volume well. This past weekend I tested out the development of actions on Google and the development of skills is more or less similar. The documentation still needs some work but overall the experience was fairly straightforward.
Confirming - I'm using it that way right now. Sometimes if you're doing something funky you have to explicitly specify the name of the chromecast. ("Pause the kitchen speakers"). But for the most part, it just works.
It's complicated in order to provide a lower-level API so that libraries can take advantage of different features and provide more opinionated ways of creating custom elements. Polymer is an example of a library which provides a simpler (but slightly more opinionated) API on top of custom elements.
> The value of shadow DOM is also quite doubtful to be honest. It is shadowed from whom and for what purpose? User don't care about DOM structure ... Programmers? They are OK with that.
The Shadow DOM provides style isolation, which is convenient to developers mixing-and-matching components, and can also be a performance boost as browsers can optimize for the fact that styles don't cross the shadow boundary. It also provides a bit more isolation for when you are writing code to discourage components from reaching into each others' private implementation.
I've proposed once @ www-styles concept of style sets: the @set construct.
Style sets mechanism solves two principal problems of CSS:
1. They allow to define local and independent style systems for elements in the same DOM tree.
2. Style sets allow to reduce computational complexity of style resolution. In classic CSS all style rules that used by the document are placed in the single ordered table. Task of finding list of matching selectors for the element has computational complexity near O(n), where n is a number of rules in the table. And style resolution of the whole DOM tree is O(n * m * d) task. Where m is a number of DOM elements and d is an average depth of the tree. Style sets allow to reduce significantly n multiplier in this formula.
I haven't seen any requests from framework/platform developers.
In contrary, the feature to call particular function when particular element appears in the DOM was actual from the time of first framework appearance. People were trying to solve it with DOM mutation events but they did not go through - too expensive, etc.
But Google has created a perverse incentive here. They give me a ton of choices, so many and so vaguely explained that I couldn't possibly understand what they really mean and I get suspicous of everything they offer me.
So what I do is disable everything without even thinking. Later when something isn't working as expected and it turns out it's because of one of those privacy settings, I start to think about whether or not I really need it.
So effectively, they give me an incentive to summarily reverse their opt-out strategy into an opt-in one. I suppose that's fine with them because most people won't do even that.
However, they should be aware of the fact that this is the second easiest thing to do after leaving everything enabled.
Over time, they may be creating a popular culture of "disable everything" just like they created a culture of ad-blocking (where "they" is the whole advertising industry in this case, not so much Google itself)
Google also has an obnoxious "if we can't keep it, you can't either" approach to user data. For example, I have location history turned off for obvious reasons and tried using the Google Maps app. I wound up having to re-enter my destination several times due to accidentally doing something that cancelled the navigation and caused it to instantly forget the destination. As another example, Google's new Allo messaging app supports end-to-end encryption but it's not on by default and activating it blocks local logging - if they can't get a copy of your chat history and use it to work out stuff about you, they won't let you have one either.
If they did support keeping the history but only using it for you, it would be even harder for them to credibly prove that they are not keeping the history for advertisement purposes ("I have disabled you keeping any location history and still you list the places I have been to? How dare you??").
I can understand your point about allo more, though of course storing the local history unencrypted also means that it's going to be accessible to interested third parties in one way or another. If they don't keep it at all, it means that it won't be.
I tried to find information about it online, but the official documentation [0] is less than helpful to say the least. I'm guessing when they say "stuff", that's a euphemism for blob, but fuck if I know, maybe they are just trying to be hip in their documentation. After all, it's also business time, in case you didn't know.
Currently downloading the Chromium source code, to see if that's more helpful, but I kind of doubt it...
However, I decided to give the benefit of the doubt and did a few file-searches, in case the files were moved at some point and the documentation was just not updated.
I think the official documentation is what's found in the help center; this page contains a lot more information about using a custom passphrase for your sync data: https://support.google.com/chrome/answer/1181035?hl=en (top result when searching for "chrome sync encrypted").
It does, but it's poorly documented and quite difficult to use. Use the flag --sync-url="" when launching Chrome, and I believe the server is run through chromiumsync.py.
Yeah, and please also don't send me any messages where you chose autocompletions for the words, or messages where your phone autocorrected the spelling of. Heck, don't even send messages where you pressed the keyboard buttons to spell out the words, it's far to impersonal and dehumanizing to be limited to the options your phone presents; people should communicate like they were meant to: by talking to each other face-to-face.
---
With Smart Replies (as with other communication-helping technology), meaning comes from how and when you choose to use them, and there's no reason you shouldn't choose to use them when it communicates what you were intending. If I was going to say "Yes!" and that's a Smart Reply option, why shouldn't I save the time and choose it? It was still my action that sent that message, still my choice to send it, still the meaning that I wanted to communicate. Smart Replies save time for the simple stuff, so you can better spend your time on other messages that you might write manually.
tl;dr: The browser attempts to learn your preference for each site automatically based on how you interact with videos. You can see what it's calculated by visiting chrome://media-engagement