Live data from Hacker News

The web sucks if you have a slow connection

danluu.com

351–360 of 622 posts

Re: The web sucks if you have a slow connection

#351
post #224

Earlier quoted context omitted.

Hah! A Jevons Effect[1] in a web site's bandwidth! [1] When an increase in the efficiency with which a resource is used causes total usage to increase. https://en.wikipedia.org/wiki/Jevons_paradox

I'm not sure why this gets a special term. It sounds like basic supply and demand. If you decrease the price of something by increasing the efficiency of production, you will obviously capture more of the demand curve. What am I missing?

The supply and demand "law" refers to the observation that the price of a good settles at a point where the available supply (which increases as the price goes up) matches the demand (which decreases as the price goes up).

Jevons Paradox is only tangentially related. It is based on the observation that sometimes using a resource more efficiently results in higher overall consumption. For example, say 40 kg of lithium is needed for the batteries of an electric car. At some point, 4000 tonnes are produced annually, enough for 100,000 electric cars per year. Now a new battery comes on the market that needs only 20 kg of lithium. Should the lithium producers be worried that the lithium demand will drop, since only 2000 t will be needed for the 100,000 electric cars? Maybe. But if Jevons Paradox comes into play, the annual production of electric cars might triple as their cost drops due to lower lithium usage, and the new demand will then settle at 6000 tpa. So, paradoxically, reducing the amount of lithium in each battery could be good news for lithium producers.

Whether or not Jevons Paradox occurs depends on the elasticity of supply-demand curves, in this case the curves for lithium and for electric cars.

Re: The web sucks if you have a slow connection

#352

Earlier quoted context omitted.

High packet loss and slow connections frequently go hand in hand (TCP over cellular modems). This effect is exacerbated if the web site you're connecting to changes congestion control and TCP ramp up settings.

Exactly, but it's not a modern website problem.

How is it not a modern website problem, if that modern website is being viewed over wireless connections, and follows Google's lead in disabling slow start TCP?

Re: The web sucks if you have a slow connection

#353

Earlier quoted context omitted.

I think this is the same anecdote: http://blog.chriszacharias.com/page-weight-matters

One wonders how a user that takes 2 minutes to load 98KB is actually able to watch a video. Even by the most optimistic estimations, a video that is a few minutes long at 480p will weigh in at 10 megabytes, meaning it'll take them OVER 3 HOURS to download the entire thing. You would probably be able to browse (slowly), read comments, but not actually do much else.

Why do you need 480p? 144p/240p was the standard YouTube less than 10 years ago.

Re: The web sucks if you have a slow connection

#354
post #344
post #224

Earlier quoted context omitted.

Hah! A Jevons Effect[1] in a web site's bandwidth! [1] When an increase in the efficiency with which a resource is used causes total usage to increase. https://en.wikipedia.org/wiki/Jevons_paradox

Wow, uncanny resemblance to Giffen goods: https://en.wikipedia.org/wiki/Giffen_good

How so? I don't really see the similarity

Re: The web sucks if you have a slow connection

#355
post #39

Something that sticks out looking at the table. How can some sites simply FAIL loading? I mean, there is something inherently wrong with our web today, where if my internet is very slow and _could_ load a page in 80 seconds if I just leave it like that, the server itself could have configured the timeout to be 60 seconds. So I can never load the page?! The assumption is here that both points of the connection is base…

I heard this somewhere on HN, but websites should fail like escalators, not like elevators. Too many people design sites like elevators though.

I love this visual, but elevators have way more safety mechanisms than poorly-designed web sites. Check out the source HTML of instagram.com for a great example. /-:

Re: The web sucks if you have a slow connection

#356
It only sucks if you've experienced a fast connection.

We generally don't target hardware from 98, why should we target bandwidth from 98? Current smartphones and computers are really powerful, and most applications are targeted towards those devices. Native apps don't have this insane requirement to support hardware from 2 decades ago.

The web is so much more than text in 2017. And before you whine about the ads and useless stuff, go read a tabloid and whine about the waste of paper, or try to watch tv and whine about the electricity and time you're wasting watching advertisements.

Media has and always will be like that.

The time spent on backwards compatibility and optimizations are usually not worth it anyway.

Do I think mostly text sites should be 5mb? Obviously not.

Re: The web sucks if you have a slow connection

#357

Earlier quoted context omitted.

Hmm, either your experience, or memory, of the nineties / dial-up, is different than mine and all of my (quickly polled friends). We all built a library of painfully-obtained 320x200 horrible overcompressed video... ahh, the memories of RealPlayer. Some video was down to 180p... heck, I still have videos that were one or two megabytes in size in my archives - that I darn tootin' well waited hours to download :)

Please don't mention RealPlayer again. I feel ill and queasy thinking of that piece of software, its million rewrites and the CPU hog they released for Linux (at least on my lame hardware). I remember using "download accelerators" to try and grab files faster back in the day. Who knew they were just doing 4 simultaneous downloads of ranges of the same file eh? Ah I feel old

As I recall, the download managers were primarily useful because they could resume files that didn't finish.

Incredibly useful for enormous files like visual basic 6, which I spent like...a month or so downloading in the late 90s.

Re: The web sucks if you have a slow connection

#358

Earlier quoted context omitted.

Not sure what your comment has to do with mine. I used Internet in the 90s and dial-up specifically as recently as 2004 - and have quite a good memory of the experience. Internet at dial-up speed was an extremely valuable commodity, to the point of disabling images in the browser and only downloading the most essential things (which is about as far from a youtube video you can get) - like documents and zipped install…

Hmm, either your experience, or memory, of the nineties / dial-up, is different than mine and all of my (quickly polled friends). We all built a library of painfully-obtained 320x200 horrible overcompressed video... ahh, the memories of RealPlayer. Some video was down to 180p... heck, I still have videos that were one or two megabytes in size in my archives - that I darn tootin' well waited hours to download :)

Can confirm, I remember downloading anime clips at 6 or so hours apiece! Not even full episodes, clips! And it was amazing!

And before there was YouTube there was flash videos/animations on Newgrounds. The people I hung with back then were into animutations.

Re: The web sucks if you have a slow connection

#359

Earlier quoted context omitted.

Yay, more browser fingerprinting data points!

Well, to be fair, as the spec notes, you can already fingerprint on speed by timing how long an AJAX call takes. Also, "on a shit home DSL connection" doesn't really distinguish me from millions of other UK residents.

Yes it does, when used in combination with the other data points.

That said, I always get "unique" on those fingerprinting tests. You can't be "extra unique," so I guess I don't mind it.

Re: The web sucks if you have a slow connection

#360

Earlier quoted context omitted.

Practically speaking, I think it's much more appropriate to just assume the admins are lazy.

I'll respectfully disagree with you. It takes more work to have a bloated JS mess of a site, than it is to have a small, simple, clean site. If they were lazy, they wouldn't have gotten to that spot in the first place.

If you're hand coding it either way, maybe but, at least in my personal experience, it's much faster/less effort to drop bootstrap and jquery on the page and get to something acceptable looking then to hand code just the 50 lines of js/CSS I actually need. Obviously there are many benefits to the latter approach, especially in the long run, but it's definitely not the lazier approach.
Post reply on HN