Hi,
>  From: Kevin Kenny <kennykb@[...].org>
>    
You make excellent points Kevin, and your arguments are well considered.
... are you sure you're in the right community? ;)
>  The TIP 257 prose needs significant rework - to propose TclOO
>  as a thing in itself rather than by comparison with XOtcl.
>  But the implementation stands on its own merits. Yes, it lacks
>  the features of ITcl and XOtcl - but that's what I want.
>  Tcl's advantage has always been minimalism; we can always
>  add features, but taking them out is a real challenge.
>  It certainly provides enough to do useful work with it.
>  For example, I have a reference implementation of TIP 308
>  that is well along.
>    
I think TIP 257 needs to be fixed, not necessarily discarded or replaced 
by another proposal.  I think this discussion is really about:
(1) What are the viable OO TIPs for 8.6, given the intended timelines 
(TIP 257 definately, and some feel TIP 279)?
(2) Which of the viable TIPs is preferred by the Tcl community and which 
is preferred by the TCT?
(3) Does the preferred TIP need to be fixed, and if so how can it be 
fixed to reasonably address the needs and concerns of the Tcl community?

There are some people still stuck on "(0) Propose OO TIPs because you're 
not happy with the existing ones".  That discussion happened in 2006, 
saw the birth of TIP 279, and concluded with 18 months of resounding 
silence.  At the end of it we have TIP 257 and TIP 279.

I think TIP 257 is viable.  I don't know about TIP 279, but I get the 
impression it isn't - there is no actual reference implementation, and 
no apparent documentation of TclCOOL.  The next logical step IMHO is to 
have the respective supporters slog it out about viability.

I think both TIPs need to be fixed before anyone can express a 
reasonable preference.  Most statements at this point are based on 
viability ("I prefer TIP 257 because its the only viable one") or 
ideology ("We should / shouldn't have a framework").  TIP 257 still 
looks like much as it did 18 months ago, is still as confusing, and 
still requires extensive familiarity with XOTcl to understand it.  I 
don't see how a preference can be expressed, much less a vote cast, 
until the TIP is cleaned up to be a succinct, stand-alone and complete 
specification of the behaviour of the object system.

I think TIP 257 can be fixed:
- The prose needs rework before a meaningful vote can be taken
- OO is about methods and state, and both need to be fully specified.  I 
have my preferences for what I'd like to see in that spec, but honestly 
knowing where I stand is more important than getting my way (I say that 
now ... ;/ ).
- The TIP is not minimal enough.  Java, C++, VB, PHP and Perl cover the 
top 68% of language use (once you exclude C), and they don't have mixins 
or guards. 

I think that TIP 279 can be fixed, but I'm not putting much effort into 
thinking about it until it proves to be viable, and I'm not in a 
position to contribute to its viability.

I think that's the end of this mail.  Regards,
Trevor



-------------------------------------------------------------------------
This SF.net email is sponsored by the 2008 JavaOne(SM) Conference 
Don't miss this year's exciting event. There's still time to save $100. 
Use priority code J8TL2D2. 
http://ad.doubleclick.net/clk;198757673;13503038;p?http://java.sun.com/javaone
_______________________________________________
Tcl-Core mailing list
Tcl-Core@[...].net
https://lists.sourceforge.net/lists/listinfo/tcl-core