ryandonsullivan

Thanks for sharing. The work our team put into this was pretty fantastic. Lots of real user tests and a number of real solid takeaways. Name, WordPress might be easier to use than Wix and Squarespace both

Yeah, no doubt that a sketchy IP will basically negate any kind of authentication you have in place. Off to read your article...

Cool tool. I'll have to play around with that and see how it goes. May end up adding it to the article :)

That's a totally fair point. I'm not entirely sure why most third parties are still using ~ in their documentation but it still seems to be the norm. I do like the definitive nature of -all.

Thanks for posting this. I know it's on a WordPress site but definitely applies to way more people than just that audience. I hope to build the article out even more and get super specific with various vendors like Mandrill, Amazon SES, Sendgrid, etc.

That's definitely fair feedback. One thing to keep in mind is, that with an outlier of 4 minutes and change, the graph would have to be blown up to a massive scale in order to see any variation in response times from the other hosts. Otherwise the other 6 hosts barely would have come up off the x-axis. The good news is that we have the data to produce exactly what you're talking about, but in terms of practicality for this article it was a little tough to justify. That said, I hear you and putting that together will be a fun new challenge.

Hi Sean, it surprised me enough that I ran the test multiple times and came up with the same result over and over. I connected with Ben over the weekend and would love to talk with you guys some more!

Another thing to look at is using nginx as a web server. It's a lot more lightweight than Apache. Paired with a solid SSD drive and varnish caching, and you can get some really great performance out of WordPress.

Yep. These are all great points. This is the first set of many many tests to come but it was good to get a baseline. If you check the timestamps of the full tests, you'll see that all of them were run back to back to back within the course of about 2 hours total time, so there shouldn't be much variation. That said, locking down ALL of the potential variables is definitely something I'm striving for long term.

For this particular article I didn't record any uptime. It's definitely a metric worth measuring, but I wanted to get super specific with this test. Lots more testing to come :)