A bug that Microsoft won't fix, because Linux is not supported.
</charitable>
A bug that Microsoft won't fix, because Linux is not supported.
</charitable>
case @useragent
when 'Windows'
send file fast
when 'OSX'
send file fast
else
sleep 3600resolve_renderer_module(user_agent.os)
where the osx/Windows ones are already loaded in RAM, but it has to try (and fail) to load the linux one every time somebody tries.
This is classic Hanlon's razor stuff.
I wonder if people said "Hanlon's Razor" when they first heard about the VW emissions scandal.
NOTE: I'm just thinking out loud here. I don't have any particular bias against Microsoft.
I don't think that was possible, because the engine specifically performed [0] better during the test and worse during normal operations. So what would Hanlon's Razor be? Whoops, the testing procedure accidentally happened to make our engine perform better?
[0] With the definition of performance being emissions output, not power or torque.
Changing the version didn't seem to matter (I couldn't find the latest one for any Linux style OS easily, but I tried "Mozilla/5.0 (X11; OpenBSD i386) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/56.0.2924.87 Safari/537.36" with the same results). However, Android type user agent strings like "Mozilla/5.0 (Linux; Android 4.0.4; Galaxy Nexus Build/IMM76B) AppleWebKit/535.19 (KHTML, like Gecko) Chrome/18.0.1025.133 Mobile Safari/535.19" load fine.
Not knowing the code I have no idea what this could mean, but I suspect you're 100% right here, there's some server side rendering that does different things based on user agents, and it's having trouble on the "other UA" category for some reason.
Never attribute to malice that which is adequately explained by stupidity