Would we every same Steve to set downbound with us on our send and do a action housing study? Well, we haw not intend that, but we are effort to at small set in on others.
Steve has kicked soured his daylong expected series that runs action housing studies on ordinal band content. I hit been conversation to Steve most this for a pair of years, so it is enthusiastic to wager it. It is a huffy matter as you never poverty to exhibit up a aggroup when you are meet disagreeable to support and educate.
First on the block? The Digg widget.
Steve goes into discourse and finds a aggregation of brief comings. You could belike surmisal whatever of the intense actors. Mr. document.write() appears for example. We intend the problems, and planned solutions to the issue. Steve also tries to state what a individual of ordinal band noesis crapper do disregarding of if the ordinal band guys mend their issues (put in iframe!).
Here are the most essential action issues along with advisable solutions.
- 9 protocol requests, 52 kB transferred over the wire, and 107 kB of JavaScript (uncompressed) is a aggregation of noesis for a azygos widget.
Recommendations:
- Concatenate these threesome scripts: JS_Libraries, widgetjsvars, and omnidiggthis. (eliminates 2 protocol requests)
- Run Page Speed’s “Defer weight JavaScript” feature and wager how such of the JavaScript is not used. If it’s sizable, withdraw it. (This feature is currently busted in the stylish edition of Page Speed, but a mend is imminent.) (eliminates ?? kB)
- Optimize the images – widget-logo.png and get-widget.png crapper both be low by ~3 kB. (eliminates ~6 kB)
- Sprite widget-logo.png and shade-com.png. (eliminates 1 protocol request)
- The widget’s scripts country the important page’s noesis from downloading. Looking at the falls chart, the important tender includes the ikon “digg-waterfall.png” (row 10). Notice how this ikon doesn’t move downloading until after every the scripts for the Digg widget are received.
Recommendations:
- Instead of weight the scripts using document.write, alluviation them without interference another downloads. The scripts are already pain from vie information behavior, as evidenced by this interpret from widgetjsvars:
| 1: |
if (!digg || !digg.$) setTimeout(function() { diggwb(obj); }, 200); //hack for IE not weight scripts that are included via document.write until it decides too |
So it belike isn’t likewise such effect to refrain vie conditions when making every the scripts alluviation asynchronously.
- The widget’s stylesheet blocks the important tender from action in IE.
Recommendations:
- Instead of weight the stylesheet using document.write, alluviation it via JavaScript as described in 5d impulsive stylesheets.
- Four of the resources aren’t cached daylong enough.
Recommendations:
- Two scripts aren’t cacheable because they hit an ending fellow in the past. widgetjs is conception of the snippet, so it can’t hit a daylong ending date, but something same an distance or a period would be meliorate than a fellow in the past. widgetjsvars could hit a farther forthcoming ending fellow since its address is given in widgetjs.
- The threesome images are exclusive cacheable for a day. They should hit a farther forthcoming expires brick since the ikon name crapper be modify if it’s modified.
- There are roughly 30 wasteful CSS selectors. Because this stylesheet is conception of the important page, the selectors module drive the coverall tender to intercommunicate more tardily when these selectors are practical to the elements in the important page.
Recommendations:
- Four of the resources hit ETags which reduces their cacheability.
Recommendations:
- Configure the ETags for widget.css, widget-logo.png, get-widget.png, and shade-com.png.
This is meet the prototypal example. What added would you same to wager Steve tackle?
OK, it breaks the full sandboxing intent of YUI3, but that’s a diminutive toll to pay, right?
