Defrag Tools #141 - Larry Osterman - API Contracts

Play Defrag Tools #141 - Larry Osterman - API Contracts

The Discussion

  • User profile image

    I maybe missing the whole point. If the API is only on phone and say your on a desktop. What is to stop you from virtualizing a phone (i.e.) kind of like HyperV but on an workstation OS. COM has come a long way since OLE first came out. It could work the other-way phone emulating Desktop. With the Universal API Contracts sky's the limit on a convergence of device types. Regardless it is a very elegant solution to an extremely complex problem.

    On the invisibility of the contracts to developers. I agree that 90+ % of Devs are not going to care about nuts and bolts. Face it this is airing on Defrag Tools. It is important to me that invisible means behind the scenes and not a BlackBox. You never know what innovations are possible as long as  you have access to all the pieces when you need them.

  • User profile image
    Deactivated User

    It should be API commitments, as with Desktop APIs commitment, Band APIs commitment, devicefamily API commitment, .., Universal API commitment. It also clears up that as an app dev I should let Larry worry over any concerns with it as it's on the platform side of things.. :)

    Except for Firefox devs, if anyone else cooks up things resembling JavaScript then they fail to get traction. ECMAscript implementations are headed toward LLVM/Vulkan like bytecode execution, without any effort to converge Parrot VM with it. So much ABI fragmentation..

    How about using Band SDK from C++, without .net languages? Instead of C++/CX's -> operator, Kenny Kerr's moderncpp seems to be using . operator. C++ is good, projections need rework. So many topics worth hearing Larry's ponderings over.. Hope his time commitments work favoring us.

Add Your 2 Cents