Shit-ass Websites

Last night John Gruber pointed out how bad The Next Web was to load in a Web browser. Here’s what Gruber found using Safari’s inspector on The Next Web:

I measured a few of their articles using Safari’s web inspector, and Cody wasn’t exaggerating. One article at TheNextWeb weighed in at over 6 MB and required 342 HTTP requests. 73 different JavaScript scripts alone. Absurd. I did a reload on the same page a few minutes later and it was up to 368 HTTP requests but weighed “only” 1.99 MB.

It got me thinking about some of the other sites I visit, so I did some tests loading the homepage of each site and here’s what I found. There are three stats for each site — the number of http requests, the size of the page and how long it took the page to download. To be fair, I also included The Loop.

Clearly most of the Web sites I tested are pretty good. At least they don’t come close to The Next Web.

The Loop: 38 requests; 38.66KB; 1.89 secs

Daring Fireball: 23 requests; 49.82KB; 566 milliseconds

Macworld: 130 requests; 338.32KB; 8.54 secs

Ars Technica: 120 requests; 185.99KB; 2.08 secs

Apple: 46 requests; 419KB; 1.39 secs

CNN: 196 requests; 269.41KB; 4 secs

BGR: 368 requests; 2.74MB; 35.33 secs

AppleInsider: 141 requests; 649.39KB; 5.64 secs

Facebook: 137 requests; 993.54KB; 11.19 secs

MacStories: 119 requests; 2.16MB; 2.13 secs



  • Anonymous

    I had removed BGR from my bookmarks and stopped visiting their site because the long load times made me think there was some bad coding that was messing up Safari. I was apparently only half right: their bad coding must equally mess up all browsers. Funny, their tone towards Apple was also a little off-putting, as well. There seems to be a connection between Apple-related sites and good programming; maybe the need to be neat, clean, and have good design goes hand in hand with good coding. LoopInsight, for example.

    • Anonymous

      I agree mostly. However, MacStories is 2.16MB.

      • Anonymous

        Too bad for them. That may affect their future popularity. But their load times are nothing like BGR.

  • Anonymous

    I think the benchmark should be that websites should load almost instantly on nearly any connection.

  • Anonymous

    There seems to be a connection between Apple-related sites and good programming

    Hmm true that.

  • Anonymous

    What happened to your site, Jim? I just got a shitty-ass GoDaddy page filled with gaudy ads for several minutes. Eventually I was able to get through to the actual article after several re-laods.

    • http://twitter.com/hellodeibu Dave

      John Gruber linked to his site, that’s what happened ;-)

    • http://www.loopinsight.com Jim Dalrymple

      We are going through a server move tonight. DNS has been changed, so that’s what you saw.

  • http://twitter.com/tristandyer Tristan Dyer

    I don’t think it is an exclusively Apple idea that websites should be svelte.

    Microsoft.com: 78 requests, 213.34 KiB, 1.5 seconds

  • Bumble33

    Why? Are we really that impatient that we cant wait a matter of a few seconds? if you think of all the money spent just so we can have web page loadings of say 1 or 2 seconds rather than 5 or 6, and all because we dont like to wait, it sounds a bit pathetic. I’m sure we could all think of better ways to spend developer time – coming up with apps that do what you want them to do more easily, operating systems that dont crash etc. I will admit that I do get impatient when a page doesnt load quickly, but thats my problem for expecting things instantly. I can wish a kettle to boil water in 3 seconds but it aint gonna happen, so I have to wait for that so why not a web page?

    • belton braces

      Reading that empty comment just cost me 6 valuable seconds of my life that I want back.

      Incidentally, your 3 second kettle: http://www.amazon.co.uk/Tefal-Quick-Cup-Water-Seconds/dp/B000SK9G52

      • Anonymous

        The 3 second kettle was the best technical tip I got all day.  Sadly, they don’t seem to sell it in the US :-(

    • Anonymous

      Why?

      Let’s say you are on AT&T’s cheapest data plan and all you use it for is to visit a website like The Next Web, just twice a day, then you’d reach your limit five days before the end of the month.Or maybe you live in a remote area and at best you can only get a 2GB monthly internet plan. Remember to not visit more than 17 webpages a day or you’ll be paying the big bucks.

    • http://twitter.com/mikeharper mh

      When “we” could mean thousands (if not hundreds of thousands) of users loading a few pages each, then you’re talking about hours of load time that a user could be navigating your site and buying things or clicking on ads. Instead, they’re waiting for your site to load.

  • Maverick

    @Bumble33 The issue isn’t necessarily the wait time, it is the amount of data that the end user is forced to use viewing these sites. Since an increasingly growing number of people are viewing web pages on smartphones and tablets using cell phone carriers’ draconian data plans, every KB counts. I have a 2GB per month data cap on my plan. It would not take long to blow through that if I visited sites that demanded 6MB each time I loaded it.

  • http://www.informationworkshop.org Mark Hernandez

    I track over 250 tech-related sites in NetNewsWire.  When I saw this article, the first site that popped into my head as the most overloaded is http://macdailynews.com/  which, well, try the main page and then try an article.  :-)

    • Anonymous

      That’s awesome.  Big flashy ads surrounding the plain text.  It’s almost like Vegas!

      • http://www.theuniversalsteve.com Anonymous

        That’s why I use GlimmerBlocker and have disabled Flash in Safari. Sites like MacDailyNews are readable for me. It’s just text with mostly plain columns on either side. Nothing moves or blinks.

  • Anonymous

    I measured TNW several times, with and without cache in Chrome and the max I could get out of the requests was 192, tested both on homepage and article’s. Although I think my internet connection is contributing to this score (70mbit dedicated connection)

  • Anonymous

    The total time to load is subject to variable latency, and it would be better to average a bunch of different tests together. Your point is well made though.