Boston Web Performance Group Message Board › Tools for validating mod_pagespeed
I've been exploring mod_pagespeed and thinking about how to quantify its risks and rewards on a production e-commerce site.
I have developed some tools, might open source them if there's interest. Have rough working versions of:
Producing what it should? spreadsheet-driven tests: put your control and experimental mod_pagespeed settings into a CSV, do asserts on headers, sizes, HTML....
Does page look the same? : rasterize before/after pages, calculate degree of change. Result: a programatic "yup, renders the same" with certainty, not eyeballing.
Better? How much? - Capture #requests, dom ready, page load time, bandwidth etc. Enhancement: dial in simulate real-world latency/bandwidth connections with dummynet.
If anyone would like to try them and give me feedback/collaboration when adopting or doing ongoing CI checks, please drop me a line. I'm in Waltham.