Levy and Griver - What didn't make it in?

Hmmmm. the idea that MS has limited resources, or at least a project has limited resources sounds strange. is there any projects in the company that would get more developers quicker then others? im guessing that windows, especially longhorn, would have a lot of developers, and probably a lot more in the "magic box"? 

You know what?  I'm cool with the "staying hungry" part of things.  But I don't see strong enough emphasis on security yet.  My perception of Windows security is very strongly point not in favor of the current implementations.  That is one area that I think it's CRITICAL to have a large programming staff assigned.

Huh? What don't you see? We've trained every developer, PM and tester in the company around security. We have automated tools that we run at every build. Every spec has a security review. We have people whose full time job is to provide security reviews at a higher level. WS03 has had many fewer security updates than any Windows OS before.

I'm curious what you think we should be doing - if we're missing something, maybe I can drive some of that to happen.

I'd be interested in seeing the "preview" version of tool as a web download. Good to hear it's still on the plate

Can anyone say why product development can't be more of an iterative process, with new features and updates being introduced bit by bit rather than in large, one-off releases? I'm thinking here specifically of Visual Studio.NET, where we've received very little in terms of fixes and improvements for several years now. And when we did, it was packaged as a new edition, with a new box, or a new upgrade CD.

In the "old" days (i.e. before the explosion of the Internet), it still made sense to push new releases which would appear in our local computer stores. Nowadays however, we have the ability to distribute software via the Internet, so the physical limitations are reduced.

Once new features are tested and available, out they go to the world. Doesn't this make sense? But, at the same time, would a progressive approach also cause compatibility problems, and general confusion? Not to mention deployment issues, where a new feature X could rely on feature Y which can't be shipped because feature Z doesn't work on machines with feature A (I ran out of letters!)? 

Any thoughts, as always, gratefully received

miseldine wrote

But, at the same time, would a progressive approach also cause compatibility problems, and general confusion? Not to mention deployment issues, where a new feature X could rely on feature Y which can't be shipped because feature Z doesn't work on machines with feature A (I ran out of letters!)? 

I would think absolutely there would be problems with half baked new features being released willy nilly into the wild. I used to work for a POS company. They had a product that ran under DOS and Windows (Post 95 days) I know it was ancient history but the permiations of equipment that we were having to test for were swamping us. I got the reputation of being able to make anything work with anything else, but I tell you it was hell. Fast forward some years, now the explosion of hardware and software surely has only exasperated that situation. I love the idea of being able to get things as they are created but in practice I realize that that love would soon enough turn to hate when the unavoidable incompatability would rear its ugly head.

my 2c

One of my favourite aspects of Microsoft is the 'limited resources' factor. The startup mentality. Must be exciting guys, and will definitely help (and stay) if things ever got leaner.

Stay hungry guys, it makes all of our lives and jobs easier.