• Re-thinking the Windows development cycle

    Ed Bott has a(nother) great piece out on ZDNet: Microsoft, stop feeding bugs to a billion Windows 10 users. Here’s how. He wraps a cogent argument around what we’ve all been bellyachin’ about for… six?… years now.

    I’d like to go one tiny step further, and suggest that Microsoft revamp the outward face of its development cycle. It’s simple, really. Here are the buckets we should have to get Windows from the dream stage to hard, cold reality:

    Canary (or Developer) Channel – the primordial stew, not necessarily associated with a specific version

    Beta Channel – for testing a new version before it’s released, just as you would expect

    Preview Channel – combines the new “Release Preview Channel,” the current “Semi-Annual Channel (Targeted)” half-baked status, and the new “Preview Cumulative Update” releases.

    Stable Channel – when the product’s ready.

    I don’t see much distinction between “Release Preview Channel,” “Semi-Annual Channel (Targeted)” and “Preview Cumulative Update” levels. There’s a lot of tongue-wagging going on, but in the real world it’s a simple choice – do you want the new stuff early, or are you willing to wait until it actually, you know, works?

    A lot of people inside Microsoft spend a lot of time (and a lot of money!) splitting hairs on all of the distinctions. What Microsoft’s customers care about is much more straightforward.