Mark Jaquith

<p>When your posts get a lot of comments, performance can be a concern. The generation of the comment HTML and the delivery of all that HTML can slow down your page loads. <strong>This plugin has pages load without comments, and then uses a secondary AJAX request to pull down the comments.</strong> Navigating through the pages of comments is also handled via AJAX. On my test install with 100 comments per page, the <code>onload</code> event fires 2x faster. The HTML is generated by the server 6x faster. The HTML downloads 3x faster. The uncompressed size of the HTML is reduced by a factor of 17.</p>
<p>Initial development of this plugin was commissioned by Google — many thanks!</p>


AJAX Comment Loading

Loads comments with a secondary AJAX request, including comment paging. Speeds initial loading of the page.

eddr

secretchord.io

<p>This plugin offers a way to queue Ajax calls, skipping the loading of the the bulk of WP system, saving considerable system resources in the form of CPU load and runtimes. This operation can be set for any AJAX call.</p>
<p>This is done by catching the AJAX calls early in the WP lifecycle, via an MU plugin. The calls are santized ( user should define fields ) and ar stored in a queue for later processing. The queue system is based on Automattic’s own Action Scheduler plugin.</p>
<p>Processing the calls : the calls are processed in bulk via a hook that is run by CRON.</p>
<p>The original motivation for this development was to handle the multitudes AJAX calls that the PixelYourSite evokes when the facebook conversation API is in use. The conversation API requires the server to send the tracking information instead of the client browser. This means possibly lots of calls even if the site is beyond walls of cache.</p>
<h4>Who is it for</h4>
<p>At this stage the plugin is a developer level plugin and is intended to be used as such. It requires defining fields for sanitation.<br />
However, it comes with a basic implementation for the latest PixelYourSite API calls and can be easily activated without further configuration for most cases.</p>
<h4>The flow</h4>
<ul>
<li>AJAX call evoked in the client browser -></li>
<li>MU plugin catches the call if the configuration includes the specific action of the call -></li>
<li>Processor ( built-in or user defined ) checks the data and sanitizes it -></li>
<li>Data is stored in DB to be handled in queue. The data includes the call’s data and cookies -></li>
<li>Data is processed in batch via a CRON hook. The processing is done via the same “Processor”.</li>
</ul>
<h4>Basic terminogoly</h4>
<p><em>Processor</em> : A class/instance that is used to define the <em>data map</em> of fields for sanitation and later on, for batch processing. In many cases, the Processor will be quite simple. Please check the built-in Processor class <em>“class-pys.php”</em> under the “qacp/Processors” directory.<br />
<em>Data Map</em> : The map of fields and their respectable sanitation functions. There’s an example in the “qacp/qacp-how-to.txt” file.</p>
<h4>Things to Remember</h4>
<ul>
<li>The Processor should define clearly what are the received fields and their sanitation functions.</li>
<li>The action and processor should be registerd. There’s an example in the “qacp/qacp-how-to.txt” file.</li>
</ul>
<h4>Hooks</h4>
<ul>
<li>qacp/processors/active_processors</li>
<li>qacp/processors/actions</li>
<li>qacp/processors/identify_processor</li>
<li>
<p>qacp/processors/instance_by_action_name</p>
</li>
<li>
<p>qacp_before_batch_processing</p>
</li>
<li>qacp_before_process_entry_{$_action_name}</li>
<li>qacp_before_process_entry</li>
<li>qacp_process_one_entry_{$_action_name}</li>
<li>qacp_process_one_entry</li>
<li>qacp_mark_event_finished</li>
<li>qacp_after_process_entry_{$_action_name</li>
<li>qacp_after_process_entry</li>
<li>qacp_after_batch_processing</li>
</ul>


Queue Ajax Calls Performantly

Provides a way to queue Ajax calls in a considerably more performantly via an MU plugin.  Allows processing these calls as a batch.

AJAX Comment Loading vs Queue Ajax Calls Performantly