Geoff Chappell - Software Analyst
A conversation in a bar, of all places, alerted me to real-world interest in undocumented functionality in an area of Windows that I had somehow managed not even to have looked at: the Desktop Window Manager (DWM). Whether through ignorance or wishful thinking, I had assumed this area was one in which Microsoft wanted or even needed third-party co-operation, such that all or most of the functionality would be documented. It seems instead to be yet another instance of Microsoft (understandably and even naturally, if not entirely commendably) documenting only enough for what Microsoft needs from others for what Microsoft wants to present as new in Windows but not as much as might help others develop innovative enhancements of Windows. Of course, all I mean to present for now is some basic bookkeeping, such as anyone might take as a first step, while I assess whether this is an area I want to move into studying properly. If you would like me to study it properly, write to me with your encouragement (and reasoning).
While looking at this stuff, and noting that I have been asked before, I have started a catalogue of KERNEL32 versions and functions. Of course, this means I must at least touch on the new KERNELBASE.DLL for Windows 7 and eventually write some explanation of this change to the architecture.
Meanwhile, this cataloguing, which could of course be done by just about anyone, is all I can nowadays justify as “research” for free publication. I must stress again that when advocates of open-source software grumble about closed-source software, they’re really just excusing the industry’s laziness. Studying software without source code is not fundamentally difficult work—intensive, yes, if you want good results from it, but not fundamentally difficult. Neither is it work that produces nothing that nobody wants to know. It’s just not work for which the industry finds commercial support.