Re: Cygwin Book?

[Warren Young <warren at etr-usa dot com>]

Christopher Faylor wrote:
> I have gotten periodic requests to write a cygwin group from various
> publishers but it has always seemed like a daunting task to me.

The trick is in deciding what to cover.

It seems to me that just getting Cygwin installed could be stretched to 
maybe fill a chapter.  The hardest part is just finding the packages you 
need in the tree, and because you can do it iteratively, it doesn't come 
to much of a practical problem.  If I were writing it, I'd probably make 
this Appendix A, not Chapter 1.

You could fill a book with chapters that are basically "how to use Linux 
on Windows", but really, aren't 99% of Cygwin users *ix transplants 
anyway?  Everyone knows how to use the tools, which is why they've 
sought out Cygwin in the first place.  I guess there are a few who get 
Cygwin foist upon them as a prerequisite for something else -- some 
embedded systems compilers are like this, for instance -- but I'd bet 
this is a tiny minority of users.

I point all this out because I think I know what would be the most 
useful book, and you, cgf, are indeed one of the few people who can do 
it justice: a book on how Cygwin works and why it is the way it is.  Not 
just cygwin1.dll internals, but how setup.exe packages work, the way 
various POSIX features are distorted by the Windows lens (symlinks, 
mounts, IPC, fork, PIDs, permissions...), etc.

The Cygwin story is one of compromises, accommodations, and probably 
even some outright hackery.  This is the story that those of us who wish 
to understand Cygwin need to read.