5 ms·

I feel browsing the web is long overdue having a robot butler to do it for us and come back with the things we wanted, leaving the "People in Norwich are using

I feel browsing the web is long overdue having a robot butler to do it for us and come back with the things we wanted, leaving the "People in Norwich are using this genius trick"-style nonsense on the cutting room floor. It ought to be trivially possible, going on from things like grease monkey and ad blocking scripts to add a bit of user intent guessing / content filtering / simplified view logic.

thomascgalvin 5y ago
If it was trivially possible, Firefox would already be doing it. But there's always an arms race with stuff like this. Our robot butler would be able to figure out how to parse the web today, but the advertisers would completely change the structure of pages tomorrow. Facebook used to have their ads wrapped in `<div class='advertisement'>`. Now, they're buried a thousand divs deeps, and the ads and content both have nonsensical, non-semantic class names.
- discordance 5y ago
  It's technologically possible, but not legally possible and not convenient. People in the know have the right extensions to block stuff and get past paywalls using the right snapshot services etc. Firefox can't officially offer this, but they can let people do whatever they want if they run their own extensions. If this was common place or convenient for the masses then it would be actively discouraged or outlawed, similar to what happened with Napster and MP3s.
- thizzbuzz 5y ago
  If it was trivially possible, Firefox would already be doing it. I am thoroughly unconvinced of this, as a Firefox user for two decades. There are so many simple power-user features that Firefox has chosen to ignore.
- throwaway81523 5y ago
  > If it was trivially possible, Firefox would already be doing it. Sorry but Firefox is mostly on the dark side these days. That shouldn't be too surprising since most of its money comes from google. It does have a setting to stop embedded video from playing sound unless you send an event, but it is too easy for obnoxious pages to evade and play sound against your wishes.
- waltbosz 5y ago
  I also noticed super deep div structure, and weird random class names. I assume it's to prevent automation and scraping. They're really hard to run querySelector against.
  - WesolyKubeczek 5y ago
    Then you render it to text which gives you a decent markdown approximation, and render it back to what you want.
s-lambert 5y ago
Isn't this what Web 3.0 was supposed to be before crypto-fans stole the term? > I have a dream for the Web [in which computers] become capable of analyzing all the data on the Web – the content, links, and transactions between people and computers. A "Semantic Web", which makes this possible, has yet to emerge, but when it does, the day-to-day mechanisms of trade, bureaucracy and our daily lives will be handled by machines talking to machines. The "intelligent agents" people have touted for ages will finally materialize. - Tim Berners Lee
npsimons 5y ago
"Five minute precis" has been the dream since Gibson's "Count Zero." Unfortunately, getting something anywhere near worthwhile isn't easy.
gpm 5y ago
This is what Google has been trying to figure out forever isn't it? Things like the automagic answers block at the top of the search results, google assistant, etc. Unfortunately we aren't very good at making robot butlers yet, and copyright severely limits what they can do (provided they are running as a third party service) in the first place.
- LegitShady 5y ago
  would anyone with any sense trust google to be the sole arbiter of what they read and see anyways?
_dain_ 5y ago
>a robot butler to do it for us and come back with the things we wanted, That's what an RSS reader does.
Mezzie 5y ago
It's technologically possible, but the 'user intent guessing' makes this messy. The Venn diagram of people who understand information seeking behavior and science well enough to design this and people who have a deep enough understanding of development to design it in a way that it doesn't break every 3 seconds is two circles. So you'd need a team, and as somebody with a tech and a library background, good luck getting those two groups to communicate and work together. Doing it without the information seeking background just recreates the issues with the modern web where your 'robot butler' comes back with a ton of irrelevant information, and doing it without the tech background just gets you a bunch of librarians high-fiving themselves over 'solving' a problem using outdated tech and is used by 20 people because it doesn't scale well.

thomascgalvin 5y ago

If it was trivially possible, Firefox would already be doing it. But there's always an arms race with stuff like this. Our robot butler would be able to figure out how to parse the web today, but the advertisers would completely change the structure of pages tomorrow. Facebook used to have their ads wrapped in `<div class='advertisement'>`. Now, they're buried a thousand divs deeps, and the ads and content both have nonsensical, non-semantic class names.

discordance 5y ago
It's technologically possible, but not legally possible and not convenient. People in the know have the right extensions to block stuff and get past paywalls using the right snapshot services etc. Firefox can't officially offer this, but they can let people do whatever they want if they run their own extensions. If this was common place or convenient for the masses then it would be actively discouraged or outlawed, similar to what happened with Napster and MP3s.
thizzbuzz 5y ago
If it was trivially possible, Firefox would already be doing it. I am thoroughly unconvinced of this, as a Firefox user for two decades. There are so many simple power-user features that Firefox has chosen to ignore.
throwaway81523 5y ago
> If it was trivially possible, Firefox would already be doing it. Sorry but Firefox is mostly on the dark side these days. That shouldn't be too surprising since most of its money comes from google. It does have a setting to stop embedded video from playing sound unless you send an event, but it is too easy for obnoxious pages to evade and play sound against your wishes.
waltbosz 5y ago
I also noticed super deep div structure, and weird random class names. I assume it's to prevent automation and scraping. They're really hard to run querySelector against.
- WesolyKubeczek 5y ago
  Then you render it to text which gives you a decent markdown approximation, and render it back to what you want.

s-lambert 5y ago

Isn't this what Web 3.0 was supposed to be before crypto-fans stole the term? > I have a dream for the Web [in which computers] become capable of analyzing all the data on the Web – the content, links, and transactions between people and computers. A "Semantic Web", which makes this possible, has yet to emerge, but when it does, the day-to-day mechanisms of trade, bureaucracy and our daily lives will be handled by machines talking to machines. The "intelligent agents" people have touted for ages will finally materialize. - Tim Berners Lee

npsimons 5y ago

"Five minute precis" has been the dream since Gibson's "Count Zero." Unfortunately, getting something anywhere near worthwhile isn't easy.

gpm 5y ago

This is what Google has been trying to figure out forever isn't it? Things like the automagic answers block at the top of the search results, google assistant, etc. Unfortunately we aren't very good at making robot butlers yet, and copyright severely limits what they can do (provided they are running as a third party service) in the first place.

LegitShady 5y ago
would anyone with any sense trust google to be the sole arbiter of what they read and see anyways?

_dain_ 5y ago

>a robot butler to do it for us and come back with the things we wanted, That's what an RSS reader does.

Mezzie 5y ago

It's technologically possible, but the 'user intent guessing' makes this messy. The Venn diagram of people who understand information seeking behavior and science well enough to design this and people who have a deep enough understanding of development to design it in a way that it doesn't break every 3 seconds is two circles. So you'd need a team, and as somebody with a tech and a library background, good luck getting those two groups to communicate and work together. Doing it without the information seeking background just recreates the issues with the modern web where your 'robot butler' comes back with a ton of irrelevant information, and doing it without the tech background just gets you a bunch of librarians high-fiving themselves over 'solving' a problem using outdated tech and is used by 20 people because it doesn't scale well.