Re: ModPagespeed changes Cache-Control

[Joshua Marantz]

Hi -- looks like you are doing everything right, and when I visited http://www.jettes-fodterapi.dk/wp-content/uploads/2016/04/Behandlinger.jpg I see 

1. Cache-Control:
   max-age=345600

in the Chrome developer tools. I think you were just seeing the short TTL as a transient cache-control that PageSpeed issues to avoid having proxy-caches hold onto the unoptimized version too long.

-Josh

[Kim Frederiksen]

It seems to us that Mod Pagespeed drops the prioritize critical css optimizations after a short while if the site has no views. Can anyone elaborate on the expiration of this? With a site that has no changes, but little traffic, we simply can't get pagespeed insights to maintain the 99/100 score that it gets after a few actual page loads. A few hours after it drops again to a mobile score of 75/100 and desktop of 91/100 which indicates render blocking CSS.

Again, thanks for your help.

[Joshua Marantz]

I think there may be 5-minute cache-ttl for the critical CSS data hardcoded.  I filed https://github.com/pagespeed/mod_pagespeed/issues/1311 to track this.  Feel free to subscribe.  It seems like this shouldn't be hard for us to make configurable.

The reason for this conservatism is that the CSS may stay constant, but the HTML might change, causing previously non-critical CSS to become critical, which would result in a flash-of-unstyled-content.  However it would self-correct once more beacons are sent, so at least we should leave it up to the site owner.

-Josh

--
You received this message because you are subscribed to the Google Groups "mod-pagespeed-discuss" group.
To unsubscribe from this group and stop receiving emails from it, send an email to mod-pagespeed-di...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/mod-pagespeed-discuss/d3bdea83-2350-444c-9d04-6be1ef5d6845%40googlegroups.com.

For more options, visit https://groups.google.com/d/optout.