In response, I will typically write a quick email about how/why I can't support iOS and Safari that usually boils down to two points:
- Less significant: Safari is not as well-maintained as Chrome/Firefox and supporting it takes more work than supporting those other browsers.
- More significant: the top two browsers used by users of the Manager are Chrome and Firefox, I'm a one-man-show, and I don't have time to try to support more than two different browsers.
As I do every January, I crunched the numbers on how people used the Manager in the previous year and, as a result of increased Safari usage in 2017, I'm going to be changing that policy in 2018.
The post below explains how the numbers look in the near-/recent-term and how browser-usage looks during 2017.
For those who DO NOT feel like reading the analysis and the background, the executive summary of this blog post is that the Manager is going to start officially supporting iOS devices running Safari in 2018.
For anyone else who wants to see the numbers and read the analysis, hit the jump for the full story.
Thanks for using the Manager!
Current StateAt present, I do try to sort of...unoffically support Safari and iOS devices, but the fact of the matter is that most users of the Manager are logging in from Windows PCs and running the latest version of Chrome, and I don't really look at Safari when I'm working on new releases of the Manager.
Here's a quick breakdown of the top 25 OS/browser configuration types that the Manager handled during the last 30 days of production traffic.
UA stands for user agent:
|Production user agent stats: 2017-12-17 - 2018-01-26|
The chart shows only the top 25 most popular OS/browser configurations, so this does not represent all traffic handled by the Manager, just the most common.
I am a one man operation here, and my resources are spread pretty thin, so I have to support the largest chunk of the user base as efficiently as I can.
When I look at how most people are using the Manager, what I see is that, the most popular browser/OS configs account for 1934 user sessions during the 30 day window, and only 152 of those sessions were created by an iOS/Safari user.
Firefox accounts for about 1.5% of recent sessions, which is hardly anything, but Safari is the browser is a little bit less than one percent of the sessions, so, looking at numbers like that, it's been hard for me to justify spending time working on support for Safari.
Future StateAll of that having been said, when I crunched the numbers for 2017, I was surprised to find that though Safari barely shows up in the browser/OS combination stats, it is nevertheless the number two most-popular browser for all users of the Manager (i.e. ignoring the OS/device).
In 2016, it was Firefox, but this year, Safari surpassed Firefox as the runner up. Here's how that looks in Google Analytics:
Now, I know for a fact that the Manager looks all types of broken in Safari: I have a Mac Mini and, occasionally when someone writes in about a problem in Safari, I load up the site and man, it looks terrible.
Which is to say that, in spite of how bad/broken the Manager looks in Safari, it's still ahead of Firefox by six points.
Basically, at this point, I realize that I pretty much have to support Safari, so I'm going to a.) pick up an iPad (probably an Air 2) and b.) an iPhone 7 and start using them to QA new releases of the Manager.
(I'm less worried about desktop Safari, which is less of a problem child than the actual mobile devices, according to those of you who have written in about Safari/iOS.)
ConclusionSo, to conclude, while the Safari/iOS combo accounts for only a small sliver of traffic, the browser itself is the number two browser of choice for users of the Manager and I feel like, in 2018, I need to start acknowledging that.
Since the challenge is less about the browser and more about the devices, I'm going to bite the bullet and pick up some devices.
Finally, my operating theory here is that since a good number of users are already using Safari and suffering through a sub-optimal experience, addressing that issue directly helps a large chunk of dedicated users have a better user experience.