Hello,

A few more insights on the pipe story (copy from bug 1958930)

--------------------------------------------------------------------------------------------
--------------

As to my prior interrogations, I have found the missing piece in the
puzzle: Yes, Tclx creates raw pipes, but No, it doesn't leave them
unadorned to the core ! Indeed, what I have just realized, is that
Tcl_MakeFileChannel does introspect the type, and in the pipe case calls
TclpCreateCommandChannel with an empty pid list ! This explains everything,
including the resort to the bad TerminateThread when the reader thread is
blocked in ReadFile etc...

David, do you agree with the following conclusion:
 - TclX's [pipe] fully interoperates with the (now obsolete) machinery in
tclWinPipe.c
 - Hence it can be ported to the core (TIP304) with no change and no
worries
 - It does have the old TerminateThread issue.
 - This could be solved in post-NT era by direct wait on handles
 - But this last move is orthogonal to TIP 304 and can wait a bit.

-Alex

-------------------------------------------------------------------------
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