The average size of Web pages is now the average size of a Doom install • Hacker News Board

Re: The average size of Web pages is now the average size of a Doom install

#411

Post by rayiner » Sat, Apr 23, 2016, 4:28 AM UTC

timr wrote: post #339

dasil003 wrote: post #321

Earlier quoted context omitted.

I disagree. This is a backward-looking argument which ignores the unique benefits of the web which made it inevitable that it would evolve into an application platform, regardless of how tortured the results may feel. The web is the first truly cross-platform development environment. It is not controlled by a single vendor, and anyone implementing a new computing device must support the web (just stop for a second an…

"The web is the first truly cross-platform development environment." No it isn't. Not even close. It's maybe the first cross-platform development "environment" of which millenials are widely aware. But it's only an "environment" in the most ecological sense -- it's a collection of ugly hacks, each building upon the other, with the sort of complexity and incomprehensibility and interdependency of organisms you'd expec…

The web makes Java look clean and elegant. Which is saying something. But it's an open question whether anything as popular as the web could have been any less crappy. CPU cycles and memory are way cheaper than social cooperation.

Re: The average size of Web pages is now the average size of a Doom install

#412

Post by dinkumthinkum » Sat, Apr 23, 2016, 4:33 AM UTC

JustSomeNobody wrote: post #111

d33 wrote: post #49

Earlier quoted context omitted.

I would say that a possible solution is also to better rank websites that mention the checksums of their external resources and make web browsers keep them in cache much longer... if pretty much every website uses jQuery, perhaps we should ship jQuery with the web browser?

We absolutely, most certainly, should NOT ship libraries with browsers. This is a horrible idea. Browser vendors would then have too much influence on development libraries.

I'm not sure that is true. No one is saying external libraries cannot be loaded. It is actually a good idea to be able to rely on certain popular libraries already on the client; this makes a lot of sense. You can still have the existing mechanism so I don't think browser vendor are going to hold everyone hostage.

If you like, there could be mutually agreed upon standard repository that browser vendors routinely update from.

Sure, your less popular experimental library won't be in the list, that's what "It probably won't happen but it is hard to defend the position that it is a bad idea, I think.

Re: The average size of Web pages is now the average size of a Doom install

#413

Post by dinkumthinkum » Sat, Apr 23, 2016, 4:37 AM UTC

pikzen wrote: post #388

morgante wrote: post #370

Earlier quoted context omitted.

And why do you think web development has become the dominant form of development?

Because literally any child can start developing, and then grows up and pulls every library under the sun because the language they use is a piece of crap, without a single thought about the consequences? Thankfully node.js is bringing this clusterfuck to the desktop too.

It's more about deployment and dependency problems.

The dependencies we are talking about are transparent to the user, save only for some download performance issues which are pretty minimal, if not overstated in many cases, compared to the former issues faced by native applications.

Re: The average size of Web pages is now the average size of a Doom install

#414

Post by Skinney » Sat, Apr 23, 2016, 5:02 AM UTC

dasil003 wrote: post #379

timr wrote: post #367

Earlier quoted context omitted.

Just off the top of my head: Delphi, Smalltalk, Java, Flash, Mono, and many different cross-library thunks like wxWindows, Qt, GTK+, and so on. Two seconds of googling will find you dozens.

Those aren't cross-platform the way the web is cross-platform, they all depend primarily on one vendor's implementation. None of those ever had any hope of crossing the chasm to ubiquity, they only gained as much traction as their vendor had market clout on a limited set of platforms for a limited time period.

Both the JVM and the CLR have multiple implementations. Both JVM and CLR have a standard, as do Java and C#. The primarily implementation for both is also open source. So no, they don't primarily depend on one vendor's implementation.

Re: The average size of Web pages is now the average size of a Doom install

#415

Post by Skinney » Sat, Apr 23, 2016, 5:25 AM UTC

TeMPOraL wrote: post #389

supergauntlet wrote: post #341

Earlier quoted context omitted.

This is disingenuous - he's obviously referring to image and text articles, not complicated web apps. Obviously such a website is going to take longer to load, just as complex native applications take a while to load. But to claim that because Word takes a while that it's okay that A WEBSITE WITH LITERALLY JUST TEXT AND IMAGES is 18 megs big and takes 6000 ms to load is fucking stupid.

Also, Word takes a while to load , but then works fast and is responsive - as well-written native apps are by default - and doesn't suck up your system resources. Something which can't be said about many websites "with literally just text and images".

You're comparing a well-written native app with a poorly written website. I'm currently working on a web-app that requires an about 600kb (minified, yes I know) app.js file.

It takes a while to load (not really, 100ms + download), but after that it is silky smooth due to client side rendering (faster than downloading more html) and caching.

On the most demanding page, heap usage is a little more than 20mb.

Sure, there are a lot of websites which are slow and huge memory hogs. But that goes for many native apps as well.

Re: The average size of Web pages is now the average size of a Doom install

#416

Post by 15155 » Sat, Apr 23, 2016, 5:43 AM UTC

wallacoloo wrote: post #307

newjersey wrote: post #104

Earlier quoted context omitted.

What if the checksum was the same and you accepted the cache hit if the checksum agrees and get your own copy if it doesn't? Maybe the application should get to declare a canonical URL for the js file instead of the browser? So something like https://ajax.googleapis.com/ajax/libs/jquery/1.12.2/jquery.m... sha-256="31be012d5df7152ae6495decff603040b3cfb949f1d5cf0bf5498e9fc117d546"> Would this cause more problems than i…

This seems a little redundant - why not just ? If you wanted to explicitly fetch from google if the client doesn't have a cached copy, then instead do The first would seem preferable though, as loading from an external source would expose the user to cross-site tracking.

If accompanied by the exact same sha-256 hash idea, loading from any external source cannot expose the user to any additional risk.

SHA + CDN url list (for whitelisting/reliability purposes - public/trusted, and then private for reliability) would be ideal.

Re: The average size of Web pages is now the average size of a Doom install

#417

Post by recursive » Sat, Apr 23, 2016, 5:48 AM UTC

snowwrestler wrote: post #405

chriswarbo wrote: post #363

Earlier quoted context omitted.

> I don't get the point of those settings. You seem to prefer white-on-black text, but the "Only with High Contrast themes" override means that it will almost never apply. The idea of defaults is to be default. I don't mind sites specifying the use of particular colours, and it's useful e.g. for syntax highlighting, for text above background images, for transparent images assuming a particular background colour (e.g.…

So basically you reversed the defaults in order to make it easier to find sites with partial style declarations, so you can yell at them. Seems like a big waste of time to me.

It's not for everyone.

Re: The average size of Web pages is now the average size of a Doom install

#418

Post by tmd83 » Sat, Apr 23, 2016, 5:56 AM UTC

Razengan wrote: post #304

jessriedel wrote: post #15

I'm skeptical that developers talking to each other about how bad web bloat is will change anything. They will still face the same incentives in terms of ad revenue, costs of optimization, etc. Here's a random idea that might have more potential: create an adblocker browser plugin that also colors URLs based on how slow they are expected to load, e.g., smoothly from blue to red. The scores could be centrally calculat…

I think a more radical departure is needed. It's about time to acknowledge that the web is increasingly being used to access full-blown applications more often than "webpages." Web browsers have more or less become mini operating systems, running elaborate virtual machines. There's way too much complexity for everyone involved — from web devs and browser devs to the users and the people who maintain the standards, th…

I also agree that the web is full of bloat and a mishmash of compromised features. I think you can design a web thats more powerful yet simpler if you actually solve the problems in a more forward thinking manner.

For example there has been so many things for powerful layouts whereas everyone knew 10 years ago we need powerful layout solutions (flexbox or whatver) and now we have grid frameworks and years of craft on older css enhancement that has to be supported. They keep adding features here and there to sort of address lots of problems where individually those features might be cheaper but the overall cost of implementing both by browsers and the us regular developers is much higher.

Here are couple of the things I want from the web and quite a few of them are there already if not in super ideal forms. Powerful layout thats simple enough to use, concept of webpage, a bundle (http2? all your resource together), making the whole partial rendering (ajaxified page) a natural concept. Even making the UI/markup delivery being made separate from content (you can do that with all sorts of library but I think it should be at the core). Security concepts that are easier to implement (CSRF, url tampering etc.).

One of the idea I had is that browsers make a new engine that does the right things from the start and hopefully thats a much lighter engine and if you serve new pages they are really fast and if you serve old pages there is an optional transpoiler kind of thing that translates to the new version of the fly. Now it won't be terribly good to start with so its optional but essentially the old version is frozen and the more people start to only use the new engine (with transpoiler).

Re: The average size of Web pages is now the average size of a Doom install

#419

Post by bryanrasmussen » Sat, Apr 23, 2016, 6:34 AM UTC

tracker1 wrote: post #296

bryanrasmussen wrote: post #239

Earlier quoted context omitted.

the reason why I prefer to use a cdn is because it is a games theory example come to life here - if everyone used the cdn version then any user coming to your site would most likely have the cdn version in their cache and thus performance would go up, but if you use the cdn version and your competitors don't their performance is slightly better than yours and so on and so forth. Games theory indicates that in most ga…

but there isn't 1 CDN, or 1 version... if you need two libraries, but the canonical CDN for jquery is on one, and your required extension is on another... that's two DNS lookups, connections, request cycles, etc. So you use the one that has both, but one is not canonical, which means more cache misses. That doesn't even count the fact that there are different versions of each library, each with it's own uses, and dis…

that's 2 dns lookups on any user that hasn't already done that somewhere in the past and had it cached.

The question then is how likely are they to have done that in regards to your particular cdn and version of the library.

I agree that a lot of possible cdns, versions and so forth decreases the value of the common CDN approach, but there are at least some libraries that have a canonical CDN (JQuery for example) and not using that is essentially being the selfish player in a games theory style game.

Since I don't know of any long running tracking of CDN usage that allows you to predict how many people who visit your site are likely to have a popular library in their cache it's really difficult to talk about it meaningfully (I know there are one-off evaluations done in one point in time but that's not really helpful).

Anyway it's my belief that widespread refusal to use CDN versions of popular libraries is of course beneficial in the short run for the individual site but detrimental in the long run for a large number of sites.

Re: The average size of Web pages is now the average size of a Doom install

#420

Post by khedoros » Sat, Apr 23, 2016, 7:52 AM UTC

PretzelFisch wrote: post #393

khedoros wrote: post #309

Earlier quoted context omitted.

That sounds ambitious, interesting, and like a lot more work than keeping a webpage running. My first thought is "cross-platform nightmare". My second thought is that they'd have to rethink their expectations of continuous deployment/release and a/b testing. We're either looking at making individual pages maintain binaries for the platforms they support (implying support of only those platforms that make sense to the…

This native delivery is pretty much a solved problem, just look at how firefox and chrome auto update.

Firefox and Chrome don't do "move fast and break things"-style continuous deployment, can't push updates to a user seamlessly (i.e. without restarting the program), and can be downloaded and compiled on platforms that Mozilla and Google don't want to officially support.

Native delivery of a monolithic browser based on an open-source codebase is a fixed problem. Trying to do the same with a website, using current techniques, would cause issues to both their current workflow and my expectations as a user that websites don't currently have.

I'm not saying that it's impossible to do, I'm just saying that it's not a good fit for current trends in web development, and I'm not convinced that it would be great for the users either.