Sitespeed.io + Jenkins = LOVE

Sitespeed.io can now output JUnit XML that you can use in your continuous integration tool. With that, you can for every build test your site against the rules! In the documentation you will see how to do it for Jenkins & Travis (Travis don't have full support yet for JUnit XML but you can still make sure that builds get broken if you violate the rules).

You can specify your own limits of when you want the tests to fail: for the overall score and the limit for individual tests. And you can of course skip tests. Using Jenkins you will get information per page and per test, so you easy can see what is failing.

Feed the beast with a list of URL:s

For some sites crawling the site isn't the best thing; you want to test a couple of the most used pages (some of them longer down in the path tree). Now you can do that by supply a list of URL:s to sitespeed to test.

The list is a plain text file with one URL on each line and the first URL will give the name of the whole test. Here's an example of a file:

http://peterhedenskog.com/
http://peterhedenskog.com/work/
http://peterhedenskog.com/music/

Check the documentation of how to do it. Note: The parameter to supply a file is f meaning that the old functionality "follow only this path when crawling" has changed to parameter c so please make sure you change that if you use that functionality.

New functionality

  • DOM elements on summary page

    DOM elements

    You can now see the average & median number of DOM elements on the site summary and for a specific page.

  • No expire header

    Missing expire

    See how many assets on your site that is missing an expire header (I hope it is none!).

  • Cached assets

    Cached assets

    How many of your assets are being cached?

  • New page summary

    New page summary

    The summary has been redesigned, now you can see cache, content, time, content-type and domain metrics.

Small fixes

  • The max size of a document for getting a warning has been changed, now using statistics from HTTPArchive.
  • The primed cache values has been removed from site summary & site detailed report, because they wasn't actually showing correct values all the time.
  • The Yslow rule ynumreq has been removed and replaced by three new ones in order to get clearer JUnit XML results: cssnumreq, cssimagesnumreq & jsnumreq
  • You can now see the response headers info on the assets page.

Bug fixes

  • If a max age HTTP cache header was missing the cache time part, the cache time was set to 0 instead of checking if an Expire header exists.
  • If an asset was missing expire headers, it inherited the last one assets with an expire date/header.
r>

See the changelog for changes done in the past and the next milestone what will come in the next release.