[lzm]

It's a shame their site is so slow. It often takes 10+ seconds to open a single MDN page, so sometimes I just click the w3schools link if I don't feel like waiting.

If only they used plain HTTP instead of HTTPS...

[andos]

Indeed. Here's some statistically ignorant experiment I just did. Loading the Function [1] document (not the whole page, just the document) takes around 1.6 s here:

  Connecting:  0.837 s
         SSL:  0.258 s
     Sending:  0     s
     Waiting:  0.767 s
   Receiving:  0.001 s

   Xfer Size: 14.26  KB (gzip)
      onload:  3.37  s  (primed cache)
              12.78  s  (empty cache)

The SSL overhead is really a big deal, but waiting 0.8 s for what is basically static content just makes no sense. Meanwhile, in w3schools world, it takes around 0.2 s to get the “same” content [2]:

   Blocking:  0     s
    Sending:  0     s
    Waiting:  0.204 s
  Receiving:  0.037 s

  Xfer Size: 23.94  KB (plain)
     onload:  1.09  s  (primed cache)
              1.82  s  (empty cache)

I don't know if I can compare the networking numbers here, but I definitely can compare the onload result — MDC is annoyingly slow. Their content is excellent, they just need to improve the website performance to become the top Google result.

[1] https://developer.mozilla.org/en/JavaScript/Reference/Global...

[2] http://www.w3schools.com/js/js_functions.asp

[modeless]

The slowness is probably a big part of their ranking problem.

[lmorchard]

Believe it or not, HTTP vs HTTPS is not the problem. As I mentioned in another thread:

We're rewriting the wiki basically from scratch [1]. There are problems and growing pains. Pull requests welcome!

[1] https://github.com/mozilla/kuma