Beyond the above, here are some things to consider:
One thing to consider with regards to the future of Tcl and Tk is where docs should go. For instance, like other languages, the Tcl docs often expect the user to be familar with the standard C library and sometimes even Unix. Even if the user is using Linux (which is less and less true each year), (LV interestingly enough, I find just the opposite in my experience - more and more people are using Linux/FreeBSD/MacOSX/Solaris/and other derivitives. To say more would be quite a different topic of discussion) the user may very well not be familar with C - in fact, they may be learning Tcl to avoid learning C. Work to make the Tcl docs self-defining (i.e. one need not turn to other documents to learn how to use any particular Tcl command) might be worth considering.
Donal Fellows: While I cannot speak for the whole TCT on this matter, I can say definitely for myself that I am very supportive of improving the documentation of Tcl. I know I am too close to see all the things that need doing, so comments and clarifications from other people are really welcome. Even just pointing out that something is unclear is useful (doubly so if done by submitting a bug or patch on SourceForge, since then we're less likely to lose track of it!)
LV I agree with Donal and the anonymous poster that one of the things that would be quite useful would be community contributed improvements to the documentation. Only when people with different perspective step forward to say "hey, what do you mean here? How does that apply in the fill in the blank environment?" will there be improvement. Also, people already familar with the various environments are going to need to step forward to research the answers to these questions. RS Community-contributed additions to the documentation are collected under Tcl syntax - Tk syntax help
HJG The helpfile that comes with Tcl/Tk is quite comprehensive, but still could be improved a lot. E.g. every page should have at least one short, self-contained example, ready for cut&paste.
With the apparent downturn in publishing relating to Tcl books, users need to turn to the internet to learn the language. One of the things that would be useful to have in the Tcl community would be an effort to write a Tcl tutorial for the beginning programmer, without being platform specific, and with the only prerequisites being the ability to read and write (questions and code) in the language of the tutorial and the ability to somehow get a compatible version of tcl (tclkit/tcl BI/whatever) onto the platform to be used by the user. For some first steps in this Wiki, see Beginning Tcl.
There have, over the past year, been community members who have expressed concerns about the health and future of Tcl. Jeffrey Hobbs words today (Sep 13,2001) in comp.lang.tcl really express things well:
Now however, with the open system that we have, more people are working on lots of subtler aspects of the core. Anyone with a good knowledge of C (or good perseverance) can become familiar with one or more parts of the core. Examples are people like Rolf Schroedter who knows more about serial I/O than any previous core author; Miguel Sofer, who is examining the deep internals and byte compiler with a fine toothed comb; Andreas Kupries, who was key in rewriting the I/O core for further stability and cool new features; David Gravereaux, who worked doggedly on the utmost stability for threads (for all that he didn't want to ;) ); Mo Dejong, who helps to ensure things compile right; Vince Darley, who gave us VFS amongst other new tidbits; Peter Spjuth and Eric Melski, authors of new core widgets; .... the list goes on.
This is definitely a living language and moving further forward. Getting into writing the core isn't as hard as some think. Sure, you don't jump into the execution engine on your first day, but there's lots more to it than that. And the wealth of comment and cleanliness of the code .... compare the code base to any other major scripting language, and you'll see why Tcl is the one that won the ACM Software Engineering award for John Ousterhout.
A simple thought. Where is the "business" of TCL going? Who is going to determine it? And what criteria are they going to use when doing so? See also Business Case.
See http://aspn.activestate.com/ASPN/Mail/Message/tcl-core/3594952 for a view, at the beginning of 2008, of where Tcl will be going this year.