No affiliation with the service. Just a happy user. Great UI. 2FA implementation with TOTP, app passwords and backup codes. Custom domains support SPF and DKIM. Very cheap pricing at $10/year for "all-you-can-eat" with some soft limits. Looks like a one-person project but developer has been active with regular updates on the both the service status and issue tracker.
Keras still has CNTK and Theano backends in addition to Tensorflow. Given that both frameworks are not being developed anymore, is there any point in maintaining those backends in the future?
F# language has a concept called "computation expressions" which are related to monads. They are also sometimes called workflows. But the article is using "workflow" in the more common language way which was confusing to me at first.
I loved OCaml as well but the lack of libraries pushed me to another ML variant: F# on .net platform. With .net core and F# tooling around it gradually maturing, it is a solid cross-platform functional language.
While the title of the blog post says "for .net core", it is much more than that. The blog post clarifies it is actually targeting the new .net standard so it works on all platforms including .net core (Windows, Mac, Linux), .net framework on Windows, Windows UWP and Xamarin (iOS, Android).
Wait a few months. Then target your library development at .net standard 2.0. It will work across the classic .net framework (Windows), the new cross-platform .net core (Windows, Linux, OSX) as well as the Xamarin (iOS, Android, Windows UWP).
Virtual card numbers are not a new thing. For example, some Citibank cards allow you to create separate virtual numbers for different merchants. However, it looks like Final makes the process much easier.
Cortana isn't a Windows help client, rather it is an intelligent assistant and search engine. To give you a very common example, you can ask Cortana about today's weather. So not sure how that is related to "things aren't quite obvious".
Lumia device sales for the period of April 26 to June 30 was 5.8m. Doing a simple extrapolation to full 3 months gives an estimate of 8m Lumia smartphones in the quarter.
HSA's requirements are far broader than those provided in Metal. A7's shared memory implementation appears to be far too basic to be able to implement HSA-type API. For example, on-demand paging, user-mode queueing and ability for GPU to do CPU callbacks. Many systems can have shared memory but are not HSA compliant.
OCaml is a very nice language, and the compiler is pretty good but the only issue is that the implementation doesn't allow you to use more than one CPU core because of some issue with the GC design. F#, which is heavily OCaml inspired, has good parallel support.