Actually, most of Microsoft is still a huge sales organization. Any technical role that is part of a subsidiary and isn’t in support is under sales, and they have been relentlessly driving Azure because that’s their top priority.
That looks awesome. I do have personal concerns about GPL, though (which is why I avoid Qt - everything I do these days is either MIT or Apache licensed).
The other day I saw someone commenting on aMLC SD cards and their increased tolerance to sudden power loss and generally rougher conditions. Can’t find it, though. Anyone have a link?
I use Remote Desktop to work on my Windows machines from my Mac on a daily basis. Although the official client works fine, I prefer to use Jump Desktop due to its built-in SSH tunneling, and switch between my Surface (which sits underneath my secondary display) and machines on Azure with touchpad swipes.
RDP works fine at 5K+4K resolution (my iMac has a secondary display) and everything from Windows is rendered in HIDPI. I do however have the remote connections set at 16 bit for a little less latency (and the content on servers is less interesting anyway).
This works well enough for me to do everything (development, office work, etc.), and also Skype calls (with the audio left on the remote device - again, my Surface is sitting on my desk, and it has killer microphones and speakers).
It depends greatly on which part of a company you’re at.
At Microsoft, senior ICs (IC4 and above) make more than managers (typically M3s) in engineering teams but often the tables are reversed in services and sales teams, because managers in customer-facing teams have business targets with direct bearing on their unit/sub revenue.
I’ve been in both kinds of teams, and there is zero correlation between actual management skills and where HR rates them at, so YMMV.
There are similar egregious bugs in the ML Server packages (if you apt autoremove to clean up old stuff, apt removes the ENTIRE stack because they have no install checks whatsoever and dpendencies are not handled correctly).
Other teams ao Microsoft (like .NET) seem to have their act together, so I suppose this is just the R/ML team.