Hacker News Reader

Comment by btrask | Hacker News Reader

One of the comments points out the site in question is purely built around this feature (the Web Speech API): https://twitter.com/hammerlockddt/status/1304658792467116033

However, I think the original complaint is still legitimate. Why is Web Speech a thing, and what browser developers are poised to support such a feature, aside from Google? If Google wants to open source their speech tech, why don't they just do that?

I don't really blame Google, because they're just building a better browser than anyone else can build. But the effective result is definitely "embrace, extend, extinguish."

Perhaps it's not too late to refactor the web into a layered architecture so that it's possible to replace parts of it. But as long as we keep adding every feature under the sun, that goal is getting farther and farther away.

View on HN

I disagree that Google is building a better browser then anyone else can build. Extension support on Firefox is better then Chrome for my use case and Firefox containers are orders of magnitude better then Chrome profiles, in my humble opinion. With Chrome if i want to open up 4 different AWS profiles/accounts I need 4 separate Chrome profiles with 4 different browser windows. With Firefox I can have 4 containers/aws profiles all in the same browser window. Furthermore, Firefox for android supports ublock origin which makes browsing the web orders of magnitude better. I may have an edge case for how I use a browser and I realize that, however, that just goes to show that we shouldnt have a browser monoculture. Different people use technology in different ways.

I still use and mostly like FF too, but just look at their marketshare (and their own developer confidence, with Servo and Brave). Chrome destroyed them and I think it was mostly on technical merit, from a general user's perspective.

Honestly Google did everything they could to help FF (to the tune of $400M/year) and Mozilla squandered it. Really sad.

Edit: Before this comment gets downvoted I want to try to explain it. When Brendan Eich wanted to make a new browser, he could've based it on Firefox/Gecko, which he had decades of experience with, but he didn't. What a vote of no confidence. Then Rust was invented because they thought creating a new language was the best way to catch up to Chrome's security. Even if you like Rust, you have to admit that was a bad decision for Mozilla's management.

Mozilla was the good guys, but being the good guys doesn't automatically mean you'll win. Too much wishful thinking can easily cloud one's judgment.

I'm only writing this because I care.

They paid everyone under the sun to bundle chrome with their installers, they are pushing chrome through every one of their properties, and last but not least they deliberately introduce incompatibilities.

I would like to see some proof for the last statement.

> Chrome destroyed them and I think it was mostly on technical merit

That most certainly had nothing to do with the large advertisement on the Google-owned most visited web page in the world and the way Google artificially limited their products on competing browser.

Please, look away regulators. Nothing to see here.

Edit: I have even forgotten to mention his they forced manufacturers to preinstall Chrome on every Android phones, an action for which they were fined 4.4 billions by the EU. Clearly the idea that Chrome won on technical merits need to die. Chrome market share was made through aggressive marketing and leveraging position in adjacent markets where Google is dominant.

I also disagree that Google are building a better browser - better at/for what?

That was always Microsoft's position ... to paraphrase ... "our stuff is the best, because the most people use it, (and democracy implies most popular = best)". It was a bulls++t position for MS then, and it's a bulls++t position for G now.

Like MS was, Google is very good at pushing people to their products and keeping people. They are rather poor at writing stable, efficient, useful software - Chrome and Android are HHUUGGEE, and routinely updated with new security flaws, new APIs that add yet more security flaws, and new bugs. The purpose of chrome or android etc seems to be to force people to use it, and to force people to stay on it.

> I don't really blame Google, because they're just building a better browser than anyone else can build. But the effective result is definitely "embrace, extend, extinguish."

Google is the only tech company I know of that serves crappier versions of their biggest apps to non-Chrome browsers. Their main search page, Youtube, Gmail, hangouts. Probably more.

If you fake your user agent to Chrome, you get the "good" version of everything and it generally works fine. The differences on some sites, like their main search page, are quite stark.

But if you fake UA, you get CAPTCHA all the time and they won't even let you login to Gmail.

I'm done with Google at this point. I'm not going to be part of the masses marching us towards a worse situation than we had with IE in the 90's

Would you say premature standardization is a part of the problem?

It seems marginally better that Google at least asks others in an open forum. Microsoft basically dictated defacto standards by leveraging their dominant OS.

My preference is for vendor prefix like namespacing and waiting for at least two independent implementations before any serious standardization. But for web dev on the cutting edge I can see the temptation to play with new toys.

> Would you say premature standardization is a part of the problem?

More than that. It's a joke to try to standardize something that is still an open research problem. If something is going to be a standard, there should be at least one off-the-shelf algorithm you can implement that is roughly on par with the competition.

I'm not saying there shouldn't be any room for innovation, but... Well, maybe there just isn't. A standard probably has to be a lowest common denominator.

The problem is that the browser market is not actually a direct market for browsers. People don't buy browsers. They buy OSes, computers, phones, tablets, cars, fridges and other devices that happen to have browsers in them.

So the incentives are very skewed, the available resources too. Google will always out-compete Firefox in a relatively short time on whatever metric Mozilla comes up with. (If Google wants to of course. That's why Apple is doing all this privacy marketing thing and providing subpar but privacy-aware products.)

Ideally the browser and the web ought to be unbundled. People who pay for the Internet access should have a voice in where the Web goes. The ISPs should transfer some money to browser vendors. Sounds crazy? Yes, but that would at least provide different market forces. (And maybe browser vendors should be prohibited from accepting any other sources of revenue. Okay, but what about pull requests? Yeah, good question, this was never an easy problem.)

> waiting for at least two independent implementations before any serious standardization

What's the time limit on that wait? Should a vendor wait years to get a feature through other implementations? Should product development be stuck on other parties (potentially competitors) implement the necessary features required to develop the product?

Over time, this would result in a very large number of features limited to "vendor" namespace, how is that, in practice, different from having the features standardized (but largely unimplemented by other vendors)? And how is that better for users?

What if the vendor wants to pay someone to get those features quickly into alternate open source implementations, wouldn't HN complain then that the vendor now has direct influence in the other product's development?

It seems to me like given a set of requirements as simple as "quickly develop new web products that (may) require new web features" Google just can't win.

I love it. Just recently used their experimental OCR API that is proposed as a standard. Basically instant OCR of any image in the browser with a simple API call. Bring more APIs!

Reply on news.ycombinator.com