Future Puppy is its own Base

Current users of Puppy will easily recognize future puppy. Just like now, Puppy 7 (or whatever) will install correctly on a wide variety of hardware configurations (the widest variety of any distro). And just like now, future puppy will install in a wide variety of ways, Live-CD, frugal hard disk, usb stick, and so forth. Also just like now, the Puppy interface will be friendly, fun, and fast, making use of traditional gui elements to create an interface that anyone can just sit down and use.

The biggest development from current puppy will be in the Puppy Package Manager, because it will contain numerous selections of fine software written for KDE, Gnome, X, and windows. Users will have a choice of several different word processors, for instance, or media players, etc. As there is not only one style of user so there will be not only one style of program. The KDE and Gnome programs will run as stand alone programs, lean and fast as with everything else in puppy. An extremely important point is that all of the programs will be tested and confirmed to work perfectly with Puppy, with as few dependencies as possible, but all of them satisfied.

How these programs will be prepared is a point worth pondering. I believe that it should be done this way: every program available should be compiled in Puppy (except programs such as Opera which are not available as source)--and then they should be extensively tested to function properly. This is how it has to be done, I think, for the future puppy to be as good then as the current puppy is now.

As I think about it, I see more clearly that this is the way it would have to be done. It would not help to make the future Puppy binary compatibile with Ubuntu or even slackware again, because Ubuntu binaries, just for instance, are compiled to run with the rest of Ubuntu, and that means, among other things, the full Gnome or KDE, but that is precisely what we do not want, to drag along all that baggage (HAL and Dbus, for example). The puppy way is to think things through and then create a solution, as, for example, in running as root, even though it is unorthodox and controversial. If you become dependent on someone else then you lose the ability to do those kinds of things and to be unconventional.

It would certainly be harder to build the library of puppy programs in the way I am advocating, but the end result would be, in my opinion, better and more interesting, and along the way, I think, there would be more to be learned, like, for example, what to do if a program doesn't compile the first time ;-)