Hacker News Reader

Comment by duopixel | Hacker News Reader

Google is trying to throw it's weight at something it's competitors already do better. Even if it does catch on, we all lose with the social sharing craze that is littering the web. More clutter, slower loading pages, and gimmicks to get to you upvote a site.

This has happened before, first it was the syndication format craze with icons for RSS, RSS 2.0, atom, xml, etc. Then it was the aggregator craze (Digg, Reddit, StumpleUpon, etc) and now it's the social craze (Facebook, Twitter, etc).

There's a clear need for sharing what you like, from the perspective of the user and the publisher. I've put these buttons into my design, but I'd rather see the browsers' favorites revamped into a searchable database that allows easy sharing and get rid of this madness.

View on HN

we all lose with the social sharing craze that is littering the web. More clutter, slower loading pages, and gimmicks to get to you upvote a site.

Couldn't agree more. I've made a deliberate effort to not buy into the craze with swombat.com. Anything that can't be restyled as a text link which blends into the no-graphics minimalist style of the site just ain't gonna happen.

One small concession I've made is that I show the reddit upvote widget for people who have come directly via Reddit. Like that's helped me. ;-)

I love the way Good implemented the Twitter/Facebook sharing feature. Very minimal.

Good?

I'm guessing http://www.good.is/

For what it's worth, +1 is the best-implemented widget of its type, in my opinion. Facebook is a big baddie in this space; their widget implementations are horribly naive and drastically bog down pages.

Facebook engineering is right here, friend. If you have an idea for making our widget implementations less horribly naive, we're all ears.

Yeah, me too. I don't work on widget stuff in particular, but anecdotally looking at Techcrunch earlier today (and again just now), the Like buttons appear a few seconds before the +1 buttons...

how soon something appears on the page doesn't necessarily mean it's better. blocking scripts / document.write and other "non-best practice" things may appear first on the page.

smart loading is not about showing first, especially in the context of widgets i.e. not the main content of the site.

The biggest and most persistent issue seems to be that multiple copies of fairly heavy Javascript and CSS resources get loaded and parsed, one per per widget, on top of a relatively heavy document per button in the iframe itself, resulting in a rather quick inflation of the overall delivered page. PageSpeed says that for a simple test page with 10 "Like" buttons, a whopping 1.2MB of JS is parsed and run [1]. This doesn't include the additional overhead for the CSS; the total delivered size for 10 widgets is in the hundred of kilobytes. Anecdotally, I was able to reduce the heap size (as reported by Chrome) of some of our slower (to render) pages by a full 50% just by removing Facebook widgets, resulting in notably snappier page loads. Putting a bunch of Facebook widgets on a page just obliterated its load speed. We had issues with some pages actually crashing or hanging user browsers, and we consistently came back to the Facebook widgets as the biggest offender. Each iframe has to be parsed out into a document, its inline script run, its external scripts loaded, parsed, and run, and attached styles loaded and applied. Multiply by 20 buttons on a page and it got really slow really quickly.

What really impressed me about the Google implementation is that it collects the information about all the buttons on the page, and then fires off a single call to retrieve all information necessary to build them out (the /rpc call that seems to happen after just about everything else). That happens later in the page lifecycle, so the +1 button seems to take a bit longer to come up, but because it's all firing off at once it seems to result in a smoother overall page load experience. While I don't have a good unit test in front of me, I've often felt like pages stall out on Facebook widgets, and as far as I can tell, that's not really a concern with +1. The actual iframe loads are quite quick, the iframes are small, and are obviously built to render ASAP. The heavy lifting occurs much later. The Facebook widgets have seemed to be to be built with the assumption that there will ever only be one per page (heavy iframe document, multiple external scripts that will have to load in serial), resulting in pretty terrible bloat when you had multiples.

That said, in all fairness, in some quick tests tonight, the Facebook widgets seem to be far better-behaved than they were a few months ago, so it may be that my complaint is no longer relevant.

[1] http://pagespeed.googlelabs.com/#url=http_3A_2F_2Fdl.dropbox...

Edit: Not really a complaint, but an observation - the like.php widget has an XHTML 1.0 doctype, which requires closing tags for non-empty elements. An HTML4 or HTML5 doctype would make that valid, though.

> Google is trying to throw it's weight at something it's competitors already do better.

I don't quite understand how you can do a like button "better". If by that you mean the number of current users, I disagree that Google should stop competing just because there are more people on FB.

As of today, Google doesn't have a social platform to leverage +1. Without it, your upvotes are effectively going into a blackhole. Nobody will see them, so what's the point of upvoting and putting the button on your site?

But this is the creation of a social platform. They've just turned their search results, one of the most used things on the web, into a place where you will now see recommendations from friends if you search for topics where they've already +1'd it. It creates pretty much the same value as a Facebook like. You and I may dislike it and it does cause clutter, but Facebook likes are very popular with users and drive a lot of traffic. Google is doing the same on what I believe is a larger group of people and what I know is giving much more relevant information (i.e., I don't get informed about my friends' opinion on sneakers until I actively look for sneakers).

That's my question too, shouldn't they have launched a website to display those +1s? Or at least release an API so that others can use it?

That would be Google Profiles, but only time will tell if it's going to take off.

> Nobody will see [your upvotes]

If I understood the article correctly, that isn't true. According to Google, your friends/contacts will see your +1s in their search results, meaning they'll have an additional hint at choosing the best result for them.

Arn't they a signal for Google search and Ad rank? The SEO impact of that is pretty powerfull

"Google is trying to throw it's weight at something it's competitors already do better."

Google has one thing to offer its competitors don't - they can use the +1 button as a signal for their search algorithms.

As good as sharing a site on Facebook/Twitter is, it causes only a "momentary" increase in traffic. On the other hand, if the +1 button actually helps a site improve its rankings in Google, it could cause a sustained increase in traffic, something most websites will look very kindly on.

On the other hand, the fact that adding a +1 will cause much less of a single, highly visible high-impact traffic spike will mean its effects are harder to feel, which might make people consider it less important.

All in all, the war over widgets is just beginning.

I'm not quite clear in all of this whether clicking +1 will only tend to promote that content in the search results of your friends (Buzz connections) or whether it may have a wider SEO effect.

Can anyone clue me in?

Google said +1 won't influence search ratings. This is just a mark for your friends.

Except all your +1's are also published on your mandatory public google profile. It is private but public.

> More clutter, slower loading pages, and gimmicks to get to you upvote a site.

I've been pretty happy with the improvement caused by http://www.disconnect.me/, especially on my relatively high-latency WiMax connection at home. It still lets you load e.g. facebook.com, but only when you actually go there on purpose, not as a widget in another page.

Android's sharing intent does it right. We need a browser supported standard for sharing things much like they support RSS now. That way it'll be integrated and we won't need sites mucking up their design with this crap.

Exactly. This is implemented fairly well in a few browser extensions, but it ought to be included in core functionality, as even 'sharing' links to your own email becomes a more prevalent paradigm than bookmarks/favorites.

Off topic, but I wish Android's sharing intent would be extended into a sort of "open with..." functionality. It's great for sharing stuff, but I've had numerous occasions where I wanted to open something up in a different program (example: YouTube page to YouTube app, RSS feed to Reader, etc), but had no simple way of doing so on my phone. If you don't have a default app set, it will ask you when you open the links, but once you've made your selection, there's no easy way to switch.

+1 isn't a social sharing button. That is what Buzz is for. +1 is more akin to Digg/HN but integrated with SERPs instead of a separate destination.

Does it make more sense for +1 to be part of the browser, as opposed to the website?

slower loading pages

Strongly agree in that respect. Using the browser on my (android) phone has become increasingly unpleasant in recent months, because the address bar flashes up with every little data packet that comes in. Don't want to single out any particular firm or group of firms, but the signal:noise ratio seems to be in freefall lately.

I don't know where Web2.0 has gone wrong in recent years, but I thought a key ethos was usability.

I don't read techcrunch anymore because pageloads on my netbook have hit 20 seconds before I can scroll the page. Hacker news IIRC is still less than 250ms. Facebook is about 2 seconds till usability, but is still loading shit in the background, which hits it on par with MSN and Yahoo.

I don't get why a tech site like techcrunch would think it's acceptable to have such an absurdly long load time. Facebook lost me when page loads for every page took too long and they were only seconds. 1/3 of a minute back when I was learning to make my own websites was considered death, you wouldn't expect anyone to sit through the load.

I've found AdBlock takes care of the cruft nicely. Whenever I see a widget like that, I just add it to my filter list, and suddenly that website loads a lot quicker.

So, you want Delicious?

In the browser.

I used to use a Firefox add-in for it, myself. Worked flawlessly in my usage

Couldn't agree more about the clutter. Try loading techcrunch; it blocks Chrome for 10 seconds on a sandy bridge i7!

It's this sheepish behavior of bloggers that lowers the quality of their content as well. Just gathering random people's friends isn't the smartest strategy to monetize content. Good quality content+seamless payments should replace this model.

Reply on news.ycombinator.com