Hacker News new | past | comments | ask | show | jobs | submit login
Are Current Web Design Trends Pushing Us Back to 1999? (2011) (sixrevisions.com)
132 points by DanBC on May 22, 2013 | hide | past | favorite | 85 comments

I'll out myself as a luddite: I hate the modern web. I hate nearly everything about it other than the fat high-speed link I use to connect to it. I hate the animation, the video, the ever more insidious and unblock-able pop-ups. I hate that technical posts that used to be just some text I could quickly skim are now often videos I have to watch (I'm looking at you MSDN Channel 9!). I absolutely hate video. It's a terrible medium for communicating information to a person who is trying to multitask. It should be strictly reserved for porn and other entertainment. I hate Flash. I hate how Flash ads sit there and suck up my battery life. I hate Javascript that does the same. If I've just got my browser open to a page and I'm working in Emacs, Safari should be sitting at 0.0 (maybe 0.1)% CPU usage, not waking up my CPU constantly to do whatever shit it thinks it needs to do.

Text is great. Civilization pretty much peaked when we figured out how to blast text from the other side of the planet to your screen in 150-200 milliseconds. It's been downhill from there.

You just want to read text? How pedestrian. Don't you want to visit a blog post, stare at a blank screen for a couple seconds while a couple dozen HTTP requests are shot off for the JS, and then grabbed using an XHR? Or even better, a WebSocket? It's a modern convenience few people would want to give up!

What I want to do is watch auto-play video ads, because reading is too hard.

What's especially awesome is when they take a minute to load, after you've switched to a different tab and have no idea which site is playing them. You get to hunt through all your tabs just to discover who's playing such an awesome ad!

auto-play video ads should be disabled at the OS level

Grab adblock and flashblock. And noscript if you're feeling hardcore. The web gets way better.

Add https://userscripts.org/scripts/review/149647 as well to disable autoplay of html5-videos.

Nothing stops HTML5 video from autoplaying, sadly.

Look above you friend.

Very true.

I've used a variety of web browsers, but I've settled on Links as the best one. I can easily watch YouTube videos and in VLC, so no problem there with regards to video. The only thing I fire up Firefox for is my work-related activity which requires Flash and HTML5 to run.

But most of the time, it's

  links2 -g
and off I go. A lot of websites look really bad on it though, with formatting errors and crap HTML. Visiting the mobile versions of those sites really clears everything up, though.

Thanks for the reminder about links2 with the -g switch!

Stallman[1] goes one stage further "I fetch web pages from other sites by sending mail to a program that fetches them, much like wget, and then mails them back to me". I like the 'one firehose' approach (mail, presumably in emacs, for everything). Could add RSS feed watching in I suppose.

[1] http://stallman.org/stallman-computing.html

I use links2 -g a lot, but as you pointed out, many of sites are pretty bad on it, having pages and pages of crap before the content starts. That said, there are very few ads and paywalls and multimedia distractions.

I do wish it supported animated gifs. Most of the time when I see a gif these days, I don't mind it flashing.

This. It's so refreshing to look at a man page. They're not perfect and rarely contain examples, but we have SO and newsgroups for that. The worst, imo, are online slideshows.

You can very easily achieve all these things, but you're missing the point. The web isn't text on steroids, it's the infancy of the replacement for tv.

"Many were increasingly of the opinion that they'd all made a big mistake in coming down from the trees in the first place. And some said that even the trees had been a bad move, and that no one should ever have left the oceans."

- Douglas Adams.

> The web isn't text on steroids, it's the infancy of the replacement for tv.

The web started as a digital version of books and newspapers. Now it's trending towards a replacement for TV? How awful.

> The web started as a digital version of books and newspapers. Now it's trending towards a replacement for TV? How awful.

The web started as hypermedia with support for very limited media forms. As it supports more media, it fills more roles; becoming hypermedia that includes in the kind of media it supports types which allow it to serve as a replacement for radio, TV, etc., doesn't stop it from having the book/newspaper like functionality.

The modern web is a much better platform for the book/newspaper-like functionality that the early web was used for than the early web was, as well as competing with radio and TV for that kind of functionality.

Awful? Why, because people who like things you don't can also get things they like from the web?

>The modern web is a much better platform for the book/newspaper-like functionality that the early web was used for than the early web was

Really? Do you ever select a few words on a web page, then Ctrl-C it? You did not notice over the last 10 years as browsers have gotten more complex that you now have to use a lot more trial and error to get the words you wanted? How just a tiny movement of the mouse while dragging can have drastic effects on the selection? How text far away or in a different column sometimes gets selected for difficult-to-fathom reasons?

Never noticed how on Safari if the text is next to an image or in a title, it sometimes takes Safari a really long time to transfer the text to the clipboard, sometimes with beachballing involved?

Ever use Ctrl-F to search a web page? Never noticed how "mystery hits" are a lot more frequent than they were 10 years ago? (Where the search stops at some place where there is no visible occurance of the string you searched for.) Never noticed how Ctrl-F is not smart enough to make sure that the hit it is showing you is not obscured by one of those elements that respond to scrolling by remaining at a fixed position relative to the window?

Never noticed how even though links have a relatively short half life, hardly anyone bothers to make a local copy of the pages they have bookmarked because it is such a hard technical problem to write code that makes a local copy of all the "assets" that might be referenced by a page? (Contrast with Usenet, which, for all of its faults, made it easy to make a local copy of an article or a thread of articles.)

Never tried to bookmark a page or send someone a link to a page and notice that you cannot because it is one of those web sites where every "page" has the same URL?

I'd call you an elitist if I didn't agree with you.

Couldn't agree more, especially if you're a voracious speed reader, watching video is like trying to suck ice cream through a coffee stirrer.

Apparently Microsoft's designers agree, since emphasizing text via typography is one of the goals of Metro [1]. I've been working with it lately and have come to appreciate it.

[1]: http://en.wikipedia.org/wiki/Metro_(design_language)

You spoke my mind. Especially the CPU usage bit, I can't wait for Apple to pull the iOS-style multitasking into OS X.

That's quite ridiculous; iOS style multitasking is severely crippled.

XNU does multitasking fine already; there's no need to screw it up. If you're sick of Safari's memory and CPU usage, just switch to a different browser, but to say something as absurd as wanting task-switching on a desktop OS just for the sake of CPU cycles is short-sighted and a prime reason for the fall of productivity and change for the sake of change.

iOS style multitasking is great for battery life. And for the foreseeable future: 1) I'm not going back, voluntarily, to being tethered to a wall socket; 2) Lithium-ion batteries aren't getting more capacious very quickly.

How is it great for battery life? When applications are inactive, they're already not using CPU cycles. Task-switching involves focusing context, which does eat up CPU time, whereas with standard UNIX multitasking, switching between tasks doesn't switch context.

You could remove the context switch from the equation, but then you'd end up with a setup similar to the current multitasking system, so it's all for nought.

I'm not suggesting that we rip out the existing multitasking ability in OS X. Just the option to freeze SOME apps when I switch away from them, and unfreeze them when I switch back. I can already do this with "killall -STOP Safari" and "killall -CONT Safari". The only thing Apple needs to do is to make it automatic.

I wish I could find something that can read web pages to me while I multi task. I can listen to a video while doing something else, but I can't read and do anything else.

You shouldn't have that attitude. Because since things will progress further in the direction you don't like, it'll be a shame to lose your voice once you quit the web.




Chrome settings > advanced > content > plugins > click to play

That'll fix most of your issues, aside from publishers using video instead of text when text is preferable

Amen, brother. Amen.

One thing not mentioned was the excessive use of fixed elements that would have probably been rendered as frames back in 1999. They're not exactly the same, but most of the usage for fixed elements is replicating what a designer or developer would have used frames for, such as affixing a menu to stick to a certain area of the page no matter where the user scrolls.

Edit: Also modal popup windows everywhere, even where they aren't really needed (Twitter and the Chrome Extension Web Store are 2 that are questionable) instead of those js popup windows we used to see as browsers tend to block those. Maybe it's just me, but when you get to the point of having to put an entire page's content into a modal window, maybe it should be a separate page.

Good point on the modal popup windows! That could be a whole part of the article.

alert() was, after all, a standard way to notify the user about something (overused, yes). Then we switched to pop-in alerts. Accessibility-wise, I think it is a big loss. The appearance is no more standard, focus management and navigability suck, its semantics are lost, and you cannot block it any more ("prevent this page from opening more pop-ups").

Most of the time the content of a pop-in will be as irrelevant as the past pop-up's, but with full HTML available: welcome colours, pictures and videos!

These drive me absolutely crazy. I hate them, and they're literally everywhere now. Every website I go has some shitty fixed bar at the top that obscures text and is constantly nagging me to like their page or whatever. Developers, please just say no. Iframes died last decade for a good reason.

The worst is the new USA Today redesign. Each article page is presented to seem as if it were a modal popup, but it's not. They have an "X" button in the upper right corner that just begs you to click it, and when you do it just sends you to the home page. It's a neurotic design.

lol that is too true!

The main difference is that in 1999 you could run lynx and see nothing but pure plain text. Today, browsers no longer fetch text — they fetch scripts that fetch scripts that fetch text.


New idea. Fetch script that emulates lynx in page, fetch script that fetches text to display as plain text.

Plugin maybe?

I always had a hard time using lynx, since the early 90s

Even at that time, there were always some form scripting or frames that killed the experience in one way or another.

and i usually needed lynx to browse sites like distro repositories and the likes.

Have you used the internet without an adblocker? It's a horrible horrible place :(

Ghostery (https://chrome.google.com/webstore/detail/ghostery/mlomiejdf...) fixes most of the speed problems.

All the time and the experience is radically unlike what you describe because I simply don't visit ad-bloated sites. Those practices tell you they're just trolling for page-views – don't reward that lack of respect and you're fine.

I agree with the larger point, especially the bloat. However...

> Do we want to follow someone on Twitter just because they have 30,000 followers?

Yes. Social proof is not going away because it's part of human nature, not a technological gimmick.

> Modern-Day "Best Viewed With" Badges

It's also hard to blame designers for disliking IE, or pointing out that various experimental demo sites will work best with certain browsers, or may be a bit heavy on resources.

> Yes. Social proof is not going away because it's part of human nature, not a technological gimmick.

Technological gimmicks might be expected to tune better though. If the only people I've followed on twitter so far are artists and programmers, it seems to me that it would be a reasonable bet that I'm not going to be interested in superstars. It's odd that people would be expected to click through to follow just based on # of followers.

Since we are already at sharing personal stories: I follow only people who have a balanced follower/following ratio. I refuse to be interest in people who are not interested in me. The number of people who make postings that contain non replaceable information is ridiculously low, and I don't see the point why I should participate in non-reciprocal interactions. But I have a list of about a handful of people who's tweets I have to keep up with just because they are industry thought leaders.

It's odd that people would be expected to click through to follow just based on # of followers.

Why is this so odd? This is a fundamental aspect of human behavior. The # of twitter followers is in indication (not perfectly I might add) of the viability of credibility. This is no different than case stories, customer references, etc. which are all the cornerstones of successfully marketing entities.

Without knowing how many people saw and didn't follow, or followed and then stopped following - without the information that pushes the hypothesis back in the other direction - numbers alone are... well, they're not entirely meaningless - you can take some guesses - but they're not particularly useful.

Like - If a superstar can get 30k people to follow them, it means less than some random artist getting all the 30k people who know about them to follow them. Because in the latter case you can guess that the weight pushing the hypothesis in the other direction's going to be quite small.

In semantic terms what you're really relying on in the former case to reduce the seize of the contrary hypothesis is the assumption that society is more or less uniform with respect to its interests along the relevant axis - which I find highly questionable.

Then think of where these stickers are going to turn up -you'd expect people to have better first hand evidence available.

So, I find it a bit odd. I'd have assumed it'd get discounted into irrelevance and people would just make up their own minds based on their knowledge of the person.

I'd have assumed it'd get discounted into irrelevance and people would just make up their own minds based on their knowledge of the person.

You give the human brain way too much credit.

In a way, it's also no different from displaying comments based on how many people like it....

Social proof seems more important in the US than in Europe. Also, whether designers deserve blame or not is irrelevant. What the practice does is turn away potential visitors[1]/customers[2]. If a business is ok with that, well, then... really?

[1] Google displays a get Chrome message every time I visit Google.com.

[2] Apple's product pages work best in Safari. How's that going to help win over the die-hard Windows users?

It wasn't an Apple product page, it was an Apple HTML5 tech demo they'd crafted around Safari to show off what Safari could do. So it wasn't necessarily about turning away customers, it was about showing off.

So you follow everyone with 30K+ followers !?

Add flat UI to the list. The only thing that's saving it is better typography and better sense of spacing (since designers are more involved now).

It's a step back in terms of usability as well. In flat UI, clickable and unclickable often look the same and several other problems.

Very interesting that you should bring flat UIs up. I monitor my father's inbox[1]. I noticed earlier today he got an email to confirm the creation of a Microsoft account he needed. Instead of clicking on the flat blue "Verify xxxx@gmail.com" button to confirm his email address, he replied to the email saying "verification". It then bounced, predictably, and confused him even more.

I haven't spoke to him yet, but I'm guessing that the flat blue button with some text, in the email didn't seem "clickable" to him.

Flat UIs do indeed seem like a serious step backwards in usability.

[1]Phishing protection

What I find funny is that now, when we finally have a proper CSS support for gradients, shadows, rounded edges and so on, we're going into design style that could be easily achieved in the '90s :)

Although not a designer, I have a suspicion that the ability is precisely the driver of flat design. When gradients, shadows, rounded edges, etc, were HARD, then using them showed a degree of effort and technical capability.

In other words, gradients, shadows, rounded edges, etc were a form of credential.

Now that they're easy, they've lost that credentialing effect and (over)using them is akin to a 90's myspace page with an abundance of glitter and flashy clipart.

I personally would like to see a middle ground with tasteful use of gradients and shadows as opposed to the flat design. But then, there's a reason I'm NOT a designer (for example, I like skeuomorphic design, I think it's fun and playful)

This is just wrong, without CSS3 Flat UI would be limited to plain squares and rectangles, even if you hate the style you must at least admit there is a lot more to Flat UI than that.

We're back to mystery meat navigation. That is Web 1.0. I would also argue that a lot of designers implemention flat ui do not have an augmented sense of typography and layout to compensate.

I enjoyed this article and thought it was very insightful. To me the only point that lacked merit was the one about telling users to use a specific browser; the motivation for that practice was very different in the early 90s than it is today.

In the 90s I think it was common for sites to demand that users switch to a certain browser because they wanted to avoid compatibility issues related to CSS and JavaScript. It wasn't (at least according to my observation) usually about providing the user with a strikingly new experience.

By contrast, most modern sites that require a certain browser do so because they're flirting with an interesting technology that offers the user something new. For example the example the author cites was a demonstration, funded by Google, to promote a use of WebGL technology that allowed users to explore sprawling real 3D environments in their browsers.

Of course, restricting a user's choice of browser is never desirable and it limits your audience. Just saying I think the motivation for doing so and the payoffs we experience today aren't comparable to those of the 90s.

I think in the 90's the motivation was the same. Sure for some folks it was just about being lazy and only supporting one browser, but I think for a lot of people it was taking advantage of CSS and Javascript only available in one browser.

In those days a lot of the functionality that we take for granted as cross-browser was only available on the bleeding edge of one or another browser. I think for a mainstream audience, those badges were more appropriate then than they are now.

One big difference is that almost every new site seems to have a Macbook Air featured prominently in the middle of the homepage (at least for webapps anyway).

I was ready to dismiss it as a cheap headline, but I think he's on to something.

That said: sometimes a thing is terrible in its first incarnation, but a later incarnation works well.

Incidentally: (2011)

Whoops, sorry. I've edited that in.

The one thing I think he missed was the use of a ton of javascript snippets from different services used to "enhance" the site. Several seconds are normally added to loading the page.

One more: in the old days you often needed to run a java applet to get full functionality, today you often need to download a mobile app.


Hit Counters: These served very little purpose. The social counters serve the purpose of encouraging the viewer to engage. Also, if a client makes a request and you refuse due to personal preference to the point of losing the client, that's not very professional.

Splash Page: So, CSS3 tech demos have replaced Flash intros? Where? I have failed to see any HTML5, CSS3, canvas, or whatever other new technologies have replaced the annoying Flash intros. The closest I've seen is creative gizmos in headers and footers that don't prevent me from doing what I'm there to do.

Best Viewed Badges: These originally served a purpose in that it was certainly possible that a website would not render correctly in one browser versus another during the browser wars. The modern equivalent of that is an effort to prevent that situation from happening again. Also, pointing to a technology showcase isn't exactly proof of the theory.

Although, I agree with the validation deal. In this day it's rather difficult to have a website that validates with all the third party people doing what they want in the space, I'm looking at you Microsoft. And others.

Cut and Paste Scripts: This one confused me, it seems there is praise and complaint at the same time. "Cleaner and easier to read than before." Too many HTTP requests but too much code not in an external file? Well, which is it then? Personally, I fail to see this as a big deal unless you are seriously abusing the head section of the document.

Marquees: The only difference between them, once you look at it, is design. Plus there is a bit of function difference there as well. As in modern equivalents often serve a purpose while the marquee element had little purpose because it was so limiting.

But I'll end the same way but expand on it, in no way are things as bad today as before. I made websites back then, things are much better today on a level that simply cannot be expressed.

Just because you said the opposite doesn't make it true. Both hit counters and social counters track user interaction. These days with auto follow bots you are probably more at risk for an inflated number.

The best viewed badge is also just people not wanting to build or test in IE. That sucks that people have to do their jobs. The equivalent in programming would be someone that doesn't want to recognize memory or processor limits. You work within constraints, so man up people.

You're right, but I was just stating my opinions on the writer's opinions. Just because he said something that I think the opposite of doesn't make him correct either.

But I seriously disagree with you on hit counters since they could be set by the website's owner to whatever arbitrary number desired, the counter essentially meant nothing. A social counter is easily verified, assuming you trust the third party social network doing the counting.

Encouraging people to upgrade to a better browser is a sign of laziness on the part of the developer? Sure you work within constraints, but sometimes when it comes to browsers your market demands features you can't deliver on the browser of their choice. Therefore, you suggest a better browser. Granted, some people are lazy so there is something to what you say, such as people who refuse to use any CSS prefixes other than -webkit.

The lack of idempotent navigation pisses me off. I should be able to find what happened on my Facebook wall by date, not by "scroll until you get there".

The essay makes valid points.

However the essay design is awkward because it starts by comparing the music industry where about five rich old white male record company execs decide what everyone will be stuck listening to, with the much more democratic / organic / diverse process of web development styles and trends. Dropping the whole music industry analogy from the essay would have made it stronger.

Still a good essay, just saying an essay about style and good taste misses the mark if it doesn't optimize those traits in itself. Like a rant about misspelled words containing misspelled words or something.

Valid points, but instead of complaining, we need to understand why these trends are coming back? For. e.g. back in a days hit counter were a sign of credibility and popularity, popups are to grab user's attention (there is no excuse to auto play video!) same way twitter followers are displayed to show how popular the user is, popups and modal screens are used to say this is a different context (per say). Even though they are misused just for the sake of a developer doing 'cool stuff'. We should figure out is there more creative way to address these ?

The web design needs a re-thinking from all points mentioned in OP. +1 for the counter, these are not only worthless information but also sometimes takes so much time to load the web sites.

The glam trend died and true rock music was starting to be revived.

ROFL at "the glam trend died". Tell that to Motley Crue, Def Leppard, Poison, Dokken, Slaughter, Cinderella and plenty of other 80's era glam-metal bands who have continued to experience success into the 2000's and 2010's, and the whole new generation of glam/sleaze bands which emerged (mostly) from Sweden: Hardcore Superstar, Vains of Jenna, Crashdiet, Babylon Bombs, Gemini Five, etc.

Glam metal will never die!

And as for the idea that grunge somehow represented "real rock"... LOL. Grunge was a completely manufactured phenomenon and should probably go down in history as the most artificial "movement" ever. There's a reason the grunge bands blew into town, had their 5 or 6 years in the limelight and then mostly faded away. I mean, sure, STP or Pearl Jam still have their fans (what band doesn't?) but nobody really gives two shits about grunge anymore.

Anyway, to keep this vaguely on-topic, I'd say this... "modern" web-design does include elements which have been re-purposed, borrowed, slightly-modified, and cribbed from "old" web-design. Just like "post grunge" bands borrow elements from "traditional" grunge (downtuned guitars, etc.) along with elements from old-school glam-metal: the occasional guitar solo, a more melodic sound, more optimistic vocals, etc.

It's the nature of things... everything goes in cycles, and "things" keep re-appearing over and over again, packaged slightly differently, or with a different polish or veneer applied, and maybe with a different name. This isn't necessarily bad per-se, it's just part of the way things evolve.

You seem to be under the mistaken impression that the article was about music history.

What can I say? Other than software, metal (and glam metal in particular) is one of the things I'm most passionate about. If I had a shred of musical talent, I'd probably get out of the software business and join a glam-metal band.

Sadly, I am pretty much tone-deaf and couldn't carry a tune with a wheel-barrow.

Well that would explain your clear bias. This sentence applies equally to your list of glam metal bands: "I mean, sure, STP or Pearl Jam still have their fans (what band doesn't?) but nobody really gives two shits about grunge anymore."

The claim that Motley Crue and Def Leppard are still relevant in modern music culture is pretty silly.

You may well be right! This is one subject on which I won't even pretend to not be biased. But, being as this is HN and not "music fan news", I won't say anything more on the topic. I wouldn't have said anything about it, but I was in a playful mood this morning and felt like having a little fun. :-)

You seem to be under the mistaken impression that the article was using a good music metaphor.

Look at fashion, things always seem to come back around again.

Notice how all his "similarities" (minus twitter follower counts!) pit common and painful pratices from 1999 with demos from the modern day. Seriously. Demos! Which means that, no, we have not actually regressed into 1999.

Elegant trolling, though, I guess, if we are allowed to say that on HN.

I use ad-blocking, tracker-blocking and cookie-blocking. Some sites won't even let you see whether their flash is worth viewing without turning those all off. (I'm looking at you, BusinessInsider.)

I have a tool for them too, called "Block Site". Once they "pass the test" ... gimickry over content ... I blacklist them (fool me once, shame on you...), and any links to those sites are removed from all pages.

Design gravitates toward what the masses respond to. Obviously, from that standpoint, Facebook is "well-designed". For the more discerning, there are bookmarks for sites worth revisiting.

A lot of good points there, and a notable exception to Betteridge's law.

How ironic. I read this as I'm being annoyed by gmail accidentally breaking chat (it no longer can remember that I read a chat AND KEEPS FLASHING TO DRAW MY ATTENTION TO ITS STUPIDITY), and plus.google.com has a layout redesign that I simply loathe.

Designers will always want control, but they seem to use that control to do things that I do not like.

what's wrong with hitcounters? I still like them.

The page views from users and bots for a poorly specified interval has very little meaning. Much like the social network "followers" badges we have today.

I read the whole article. So the answer to the title question is no? Kinda confused on what the ultimate point is. Stay simple and clean no matter what toys you're using?

Title on HN could do with (2011) appending.

Yes, sorry, I edited that in now!


Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact