That vision thing ...

[Andrew Cagney <ac131313 at cygnus dot com>]

(If you know something about Australian politics ``The Vision Thing'' is
a backhanded slap at someone that that has no sense of vision :-)

Hello,

For the curious :-) Where next.

Firstly, there is always the ongoing and important need to improve
things like language support (namely C++), symbol-table performance,
native threads, code cleanups, bug fixes and the like.  I believe that
that there is now plenty active and ongoing development in these area. 
Me giving opinions on it would just be distracting.

I also think that these ongoing developments are going to be the
principal factors that determine the time frame for each succeeding GDB
release.  They are also the things that attract the most attention and
also get the most praise :-)

Longer term I can see two significant developments:

	o	libgdb

		The basic motivation is the desire
		to make it easier (and more robust) to
		integrate GDB into a GUI.

		Here the first step is MI (along with
		ui-out/ui-file).

	o	multi-arch et al.

		Here the objective is to remove
		the assumption that GDB has, at any
		time, just one thread/language/isa
		/abi/vm/target/....  The motivation
		being to make it possible for GDB to
		handle complex targets that contain
		multiple architectures and the like.

		Here that first step is multi-arch.

So?

I'd just like to encourage people to pursue what they are currently
working on.  Just keep in mind that there is likely to be more
significant change (especially through multi-arch) lurking in the
background.  For my part I hope to be concentrating on mult-arch et al,
however, I will most likely occasionally prod libgdb.

As with any bit of open source, there is of course, no timeframe ...

Comments, fits of laughter, ...

	Andrew