The Bullshit Web • Hacker News Board

Re: The Bullshit Web

#91

Post by cheezymoogle » Tue, Jul 31, 2018, 7:04 PM UTC

yathern wrote: post #78

cheezymoogle wrote: post #5

I've said this before, but it bears repeating: Moby Dick is 1.2mb uncompressed in plain-text. That's lower than the "average" news website by quite a bit--I just loaded the New York Times front page. It was 6.6mb. that's more than 5 copies of Moby Dick, solely for a gateway to the actual content that I want. A secondary reload was only 5mb. I then opened a random article. The article itself was about 1,400 words long…

I just loaded up a nytimes[1] article too - and only weighed in at 1.0MB. For a 1000 word article. Subsequent reloads dropped it to ~1000KB. I don't think that's too bad, considering there are images in there as well. Now of course, I'm running an ad blocker. I assume the remaining MB that you noticed had come from advertising sources. In which case, bloat isn't the issue, ads are. [1] - https://www.nytimes.com/2018/…

Are you also running noscript? I'm running a DNS sinkhole and still get 3mb on reloads.

Re: The Bullshit Web

#92

Post by iforgotpassword » Tue, Jul 31, 2018, 7:05 PM UTC

yathern wrote: post #78

cheezymoogle wrote: post #5

I've said this before, but it bears repeating: Moby Dick is 1.2mb uncompressed in plain-text. That's lower than the "average" news website by quite a bit--I just loaded the New York Times front page. It was 6.6mb. that's more than 5 copies of Moby Dick, solely for a gateway to the actual content that I want. A secondary reload was only 5mb. I then opened a random article. The article itself was about 1,400 words long…

I just loaded up a nytimes[1] article too - and only weighed in at 1.0MB. For a 1000 word article. Subsequent reloads dropped it to ~1000KB. I don't think that's too bad, considering there are images in there as well. Now of course, I'm running an ad blocker. I assume the remaining MB that you noticed had come from advertising sources. In which case, bloat isn't the issue, ads are. [1] - https://www.nytimes.com/2018/…

That distinction is nonsense. The ads are part of the page and are no more or less bloat then the rest of the useless junk that gets embedded. It's deliberately put there by the NY times, they don't end up there by accident.

Re: The Bullshit Web

#93

Post by jstewartmobile » Tue, Jul 31, 2018, 7:05 PM UTC

cozzyd wrote: post #69

jstewartmobile wrote: post #47

Earlier quoted context omitted.

My GMail tab runs between 225MB and 400MB+. And I used to think Java's memory usage was bad...

If I leave gmail open too long, it starts eating 100% of CPU for some reason (Firefox on Linux).

I'm using Opera on both Linux & Windows 10. It's pretty light on the CPU, but I get the "Out of Memory" screens regularly (on 16GB machines!).

Re: The Bullshit Web

#94

Post by StavrosK » Tue, Jul 31, 2018, 7:07 PM UTC

eterm wrote: post #86

I've been thinking about building either a local proxy or firefox addon that puts everything into "first-party only" mode unless whitelisted. It would probably break the web (at least the genuine parts) less than disabling javascript wholesale which is just too awkward, but it would vastly cut down on the "bullshit" as this article calls it. It would need some care, for example it would probably have to work from roo…

Nah, this breaks too many CDNs and most pages stop working, just as if you had disabled JS.

Re: The Bullshit Web

#95

Post by closeparen » Tue, Jul 31, 2018, 7:08 PM UTC

throwawaymath wrote: post #31

soared wrote: post #21

Earlier quoted context omitted.

I don't think thats a meaningful comparison. Moby Dick is a book, written by 1 guy and maybe an editor or two. NYT employs 1,300 people. When you read a book all you get is the text. NYT has text, images, related articles, analytics, etc. Moby Dick doesn't have to know what pages you read. NYT needs to know how long you spent, on which articles, etc. They need data to produce the product and you can only achieve that…

> NYT needs to know how long you spent, on which articles, etc. They need data to produce the product and you can only achieve that with javascript tracking pixels (Server logs aren't good enough). No they don't. They really don't need to know any of that. They don't even get a pass on tracking because they're providing a free whatever - I pay for a subscription to the NYT. The business, or a meaningfully substantial…

NYT needs to produce and recommend content that people find engaging to continue earning their subscription dollars.

The idea that tracking is purely or primarily there to support a business model of selling user data is a strawman invented by self-righteous HNers. You need to know what parts of your product are effective to make it competitive in today’s marketplace.

Re: The Bullshit Web

#96

Post by soared » Tue, Jul 31, 2018, 7:08 PM UTC

throwawaymath wrote: post #31

soared wrote: post #21

Earlier quoted context omitted.

I don't think thats a meaningful comparison. Moby Dick is a book, written by 1 guy and maybe an editor or two. NYT employs 1,300 people. When you read a book all you get is the text. NYT has text, images, related articles, analytics, etc. Moby Dick doesn't have to know what pages you read. NYT needs to know how long you spent, on which articles, etc. They need data to produce the product and you can only achieve that…

> NYT needs to know how long you spent, on which articles, etc. They need data to produce the product and you can only achieve that with javascript tracking pixels (Server logs aren't good enough). No they don't. They really don't need to know any of that. They don't even get a pass on tracking because they're providing a free whatever - I pay for a subscription to the NYT. The business, or a meaningfully substantial…

Except that the content is directly related to user behavior. If they see no one reads the style section, they'll cut it and move resources to financial news. If they didn't have tracking they'd never know that, be wasting resources and having a comparatively inferior product.

They can't do UX anaylsis, nothing.

Re: The Bullshit Web

#97

Post by slig » Tue, Jul 31, 2018, 7:08 PM UTC

eterm wrote: post #86

I've been thinking about building either a local proxy or firefox addon that puts everything into "first-party only" mode unless whitelisted. It would probably break the web (at least the genuine parts) less than disabling javascript wholesale which is just too awkward, but it would vastly cut down on the "bullshit" as this article calls it. It would need some care, for example it would probably have to work from roo…

uMatrix / uBlock seems to do what you want.

Re: The Bullshit Web

#98

Post by cocoa19 » Tue, Jul 31, 2018, 7:09 PM UTC

Another point for the bullshit web: The GDPR cookie consent notification. I really hate clicking on those notifications to make them go away.

Designers, can you please just add a non-obtrusive link, instead of a pop up that covers half my browser screen?

Re: The Bullshit Web

#99

Post by nrhk » Tue, Jul 31, 2018, 7:09 PM UTC

0xbadcafebee wrote: post #52

ukulele wrote: post #45

Unpopular opinion alert: Maybe the "bullshit" is only bullshit to you, the thorny tech-savvy reader. Maybe businesses have tried the plaintext approach, and their business was improved by adding fonts, stylesheets, API calls, spinners, scripts, high-res images, and god knows what else. Maybe speed improvements are not important beyond a certain point. Maybe 5MB doesn't matter to most people. Maybe micro-optimization…

They're not idiots, they just don't give a crap about waste until it affects their bottom line. (Most humans seem to operate on this principle) A lot of the business-related bullshit is a response to advertising and sales. But a lot is also lazy tech people slapping pieces together until it does a thing, and they don't care that it loads 100x more data than it needs to, because again, doesn't affect their bottom line…

Why would they care, they have to do the work of 5 in the time of one because there's now off the shelf solutions for everything. Too bad if those solutions often come bloated and incorrectly configured, maybe they should hire more engineers. Some business prioritize loading speeds but big ones with brand reputation won't care, they have little to lose.

Re: The Bullshit Web

#100

Post by _louisr_ » Tue, Jul 31, 2018, 7:11 PM UTC

cheezymoogle wrote: post #67

_louisr_ wrote: post #39

Earlier quoted context omitted.

This is both an appeal to people's universal appreciation of efficiency, and a weak denunciation of the modern web. Your argument is 1) that a website's value is the number of words on the page, and 2) that raw text is the highest value data that can be transmitted over the internet, and 3) that inefficiency and wastefulness of bits is a bad thing First off, you need to defend your first two assumptions. Don't websit…

1. A website's value is the amount of information that it provides to the end user. I would argue that the entirety of the works of Shakespeare or the 1911 Brittanica Encyclopedia provides more information than a high-definition picture of Donald Trump grimacing at EU leaders or an autoplaying ad for Doritos Locos Tacos NEW AT TACO BELL. As far as supplemental uses, at the end of the day, people are using plaintext t…

1. Well how is the format of plaintext the best method of getting information to the end user? What if you added a thin indexing layer on top of the plaintext? That would allow people to jump through huge documents with ease, but it's no longer plaintext. Sounds more valuable to me. Where is the line? What's the ideal?

2. Fair enough

3. Referencing "increasing the entropy in the universe" isn't a good argument because the amount of entropy increase due to humans is much, much less than how much entropy is increased by particles being blasted out of all the stars in the universe (unless I fundamentally misunderstand what entropy is). I think that stars blasting out particles is a much larger contributor to entropy than humans not using computer bits effectively.

And also what does entropy as a concept have to do with anything, anyway? Why should human engineering tasks have such considerations? If being super efficient with an abundant resource has a large cost (of some sort), but low efficiency has no business- or environmental- downside, then why be efficient with it?

In your last bit you argue that it's acceptable to send a site that is bigger than even the most bloated HTML/CSS page; I don't think that's true for any site/app that wants to be fast. It slows you down and people notice and stop using your service unless it's required of them.

I think that in general things are not as bad as you make them out to be, and your arguments have some merit but are mostly revealed as nonsense when the rubber hits the road. Universe entropy is completely unrelated to modern software engineering and web sites that have a lot of devs and are not SalesForce are not _that_ bloated.