jQuery provides an extremely versatile DOM query mechanism. If you’re wondering what’s happening under the covers, and if there are less- and more-performant ways to formulate your query . . . the answer is YES. In many cases you can help jQuery by specifying a tag name, i.e.:
is faster than:
. . . with the difference being that the query engine can narrow by tag name first. It’s smart like that.
So, you shoud be in the habit of specifying a tag name. The only exception: getting by ID. If you’re getting by ID, do not include a tag name, or anything else besides the ID selector.
Note that the second example here (with the class selector only) is perfectly valid, and if you need to query for multiple element types, by all means go ahead — jQuery will still return the correct results.
There is a lot of good plugin development going on for jQuery. Here are two which have come up on the list lately:
HighlightFade by Blair Mitchelmore is a very thorough jQuery implementation of the “yellow fade” technique. Options include the type of fade (linear, sinusoidal, exponential), the duration, and a function to call on completion. The plugin weighs in at 3.5K (uncompressed, so it could be made smaller), and should handle all your yellow-fading needs.
Tablesorter by Christian Bach is browser-side table sorter (click on a column heading to sort the table accordingly). It’s fast (in the demo, sorts 400 rows in about .3 seconds), and it auto-detects the column type to provide the appropriate kind of sorting. For example, it automatically recognizes date vs. text vs. URL contents, and sorts accordingly. For URLs, it disregards the http/https/ftp/file prefix for sorting purposes, which is more likely to give you the sort you actually want. You can also define your own auto-detects, which usually involves 5-10 lines of code. The plugin is about 10.5K uncompressed.
Did you know there are lots more jQuery plugins listed on the wiki at http://proj.jquery.com/plugins/? Note that the URL to the wiki will change when jQuery goes 1.0.