> On Apr 22, 2015, at 7:52 AM, Otto van der Schaaf <[email protected]> wrote: > > 2) An external piece of software (or ATS plugin). For example, > thePagespeed plugin could do this. > > The pagespeed plugin is able to scan and fetch referenced web assets in the > html (and does so in its current state). An enhancement would be to use ATS's > native fetching capabilities, currently the plugin relies on Serf.
Sweet. Yeah, we (someone) should use TSHttpConnect and/or HttpFetchSM to fetch those objects :). Sounds like this plugin has the foundation to do what the old prefetch feature does, and much, much more. Cheers, — leif > > The plugin is also able to determine which css/js/images are critical for > rendering a page above the fold. That knowledge might be nice in combination > with a server PUSH capability. > > > > 2015-04-21 23:44 GMT+02:00 Leif Hedstrom <[email protected] > <mailto:[email protected]>>: > There’s a feature in ATS, which is poorly understood, and unsupported, which > intends to fetch content based on parsing HTML. It’s our belief that this is > either better done via either of > > 1) HTTP/2 and the LinkRel headers. Server PUSH might also help or > eliminate the need for this. > > 2) An external piece of software (or ATS plugin). For example, the > Pagespeed plugin could do this. > > > Eliminating this feature would also remove the configuration file > prefetch.config. > > > This is tracked in the Jira > > https://issues.apache.org/jira/browse/TS-3167 > <https://issues.apache.org/jira/browse/TS-3167> > > > Unless we hear otherwise within 1 week, this will be implemented for 6.0.0. > > — Leif > >
