Bump ZSH_VERSION (from 4.3.6-dev-0) or a zsh release?

Bump ZSH_VERSION (from 4.3.6-dev-0) or a zsh release?

[Rocky Bernstein]

I'm sort of ready to put out the first zsh debugger release. However
it would be real helpful if the current ZSH_VERSION number were
incremented. Even better though would be a general release.

What's the current release schedule look like?

Thanks.

Re: Bump ZSH_VERSION (from 4.3.6-dev-0) or a zsh release?

[Peter Stephenson]

"Rocky Bernstein" wrote:
> I'm sort of ready to put out the first zsh debugger release. However
> it would be real helpful if the current ZSH_VERSION number were
> incremented. Even better though would be a general release.
> 
> What's the current release schedule look like?

I should at least make a test release for 4.3.7 imminently and see how
it's going.  Although there are incomplete bits, I don't know of
anything that's worse than last time and lots of things are better, so
this would be a good time.

pws

Re: Bump ZSH_VERSION (from 4.3.6-dev-0) or a zsh release?

[Rocky Bernstein]

Thanks. This is great.

In a quick scan of README and NEWS I don't see mention of that new
option added to turn of mulitiple function definitions with a single
body. Nor do I see mention of the change to trap DEBUG to have it
skip/return, nor mention of the addition of funcfiletrace, or the fc
changes (allowing fc inside a non-interactive shell, behavior of print
-s in subshells).

Possibly these aren't "important" changes, but if you want to
summarize, basically this version has support that make interactive
debugging possibe. ;-)

On Fri, Oct 10, 2008 at 4:48 AM, Peter Stephenson <pws@csr.com> wrote:
> "Rocky Bernstein" wrote:
>> I'm sort of ready to put out the first zsh debugger release. However
>> it would be real helpful if the current ZSH_VERSION number were
>> incremented. Even better though would be a general release.
>>
>> What's the current release schedule look like?
>
> I should at least make a test release for 4.3.7 imminently and see how
> it's going.  Although there are incomplete bits, I don't know of
> anything that's worse than last time and lots of things are better, so
> this would be a good time.
>
> pws
>

Re: Bump ZSH_VERSION (from 4.3.6-dev-0) or a zsh release?

[Peter Stephenson]

On Fri, 10 Oct 2008 08:50:08 -0400
"Rocky Bernstein" <rocky.bernstein@gmail.com> wrote:
> In a quick scan of README and NEWS I don't see mention of that new
> option added to turn of mulitiple function definitions with a single
> body. Nor do I see mention of the change to trap DEBUG to have it
> skip/return, nor mention of the addition of funcfiletrace, or the fc
> changes (allowing fc inside a non-interactive shell, behavior of print
> -s in subshells).

You're right, these should at least be summarised.

Index: NEWS
===================================================================
RCS file: /cvsroot/zsh/zsh/NEWS,v
retrieving revision 1.20
diff -u -r1.20 NEWS
--- NEWS	8 Oct 2008 11:37:34 -0000	1.20
+++ NEWS	10 Oct 2008 13:20:09 -0000
@@ -37,6 +37,26 @@
 Colouring of prompts is now supported within the shell by prompt
 escapes.  The prompt theme system has been updated.
 
+Various changes have been added to make debugging of shell code easier:
+  - As noted in README, the option DEBUG_BEFORE_CMD is now set by default.
+  - In DEBUG traps, $ZSH_DEBUG_CMD gives the code for which the trap is
+    called as a string.
+  - "setopt ERR_EXIT" in a DEBUG trap causes the code not to be executed.
+  - $ZSH_SUBSHELL indicates the subshell level at which code is being
+    executed.
+  - The zsh/parameter module has various additional arrays similar to
+    the existing $funcstack and $functrace, namely $funcsourcetrace
+    and $funcfiletrace.  The consistency and informativeness of
+    the output of all these arrays has been improved.
+  - Prompt escapes %x and %I show the source file and line number in
+    debug prompts (compare %N and %i which show names and line numbers
+    in the execution environment).
+  - The option NO_MULTI_FUNCDEF can turn off multiple definition of
+    functions at once, a rarely used feature that can cause problems
+    with misplaced "()".
+  - The "fc" builtin has been enhanced to make non-interactive use possible
+    and output consistent when the history is manipulated with "print -s".
+
 The completion style accept-exact-dirs has been added.  When true, this
 suppresses attempts to complete non-final directory segments of a filename
 path when the directory exists.  (For example, /home/pws/src/zsh/<TAB>

-- 
Peter Stephenson <pws@csr.com>                  Software Engineer
CSR PLC, Churchill House, Cambridge Business Park, Cowley Road
Cambridge, CB4 0WZ, UK                          Tel: +44 (0)1223 692070