tcl-core

[TCLCORE] CFV: TIP 283

[miguel <ms...@us...>]

As promised, here's the call for votes on

TIP #283: Modify Ensemble Command Resolution Behaviour

This is something of "8.5 or never", it modifies the semantics before
ensembles are officially released. Functionality of code that followed
recommended practice (fully qualify everything) is unchanged.

My Vote:

283: YES

Miguel Sofer

Re: [TCLCORE] CFV: TIP 283

[miguel s. <mi...@ut...>]

miguel wrote:
> As promised, here's the call for votes on
> 
> TIP #283: Modify Ensemble Command Resolution Behaviour
> 
> This is something of "8.5 or never", it modifies the semantics before
> ensembles are officially released. Functionality of code that followed
> recommended practice (fully qualify everything) is unchanged.

Uhh ... forgot to give a voting period, sorry.

Voting period ends at [clock format 1162857600]

Sorry

Miguel

Re: [TCLCORE] CFV: TIP 283

[Donal K. F. <don...@ma...>]

miguel wrote:
> As promised, here's the call for votes on
> TIP #283: Modify Ensemble Command Resolution Behaviour

TIP#283: YES

This TIP makes ensembles reliably behave like the way I wanted them to
originally, but didn't know how to achieve using the API available at
the time.

Donal.

Re: [TCLCORE] CFV: TIP 283

[<dr...@hw...>]

miguel <msofer=40users.sf.net> wrote:
> As promised, here's the call for votes on
>=20
> TIP =23283: Modify Ensemble Command Resolution Behaviour
>=20

TIP =23283: YES
--
D. Richard Hipp  <drh=40hwaci.com>

Re: [TCLCORE] CFV: TIP 283

[<dg...@ni...>]

Quoting miguel <ms...@us...>:
> TIP #283: Modify Ensemble Command Resolution Behaviour

283: NO

While I welcome the attempt to address the issues
raised in Tcl Bug 1436096, this proposal solves those
issues by exposing TCL_EVAL_INVOKE style command
dispatch to scripts not really prepared for all that implies.
The attempts to finish the implementation patch have exposed
these problems.

We ought to be exploding/cleaning up/eliminating semi-magical
internals stuff like TCL_EVAL_INVOKE, not letting them loose
on script programmers.

A revised proposal that addressed the same problems but
in a cleaner way with more thorough testing I believe I could
support.  This just doesn't seem ready yet.  I'm not confident
we've even discovered all the potential difficulties yet.

DGP

Re: [TCLCORE] CFV: TIP 283

[Andreas K. <and...@ac...>]

> As promised, here's the call for votes on
>
> TIP #283: Modify Ensemble Command Resolution Behaviour
>
> This is something of "8.5 or never", it modifies the semantics before
> ensembles are officially released. Functionality of code that followed
> recommended practice (fully qualify everything) is unchanged.

TIP #283: NO

The ickyness and breakages regarding [info level 0] which were found with the
implementation after the vote started make me very uneasy to let this through
before this has been sorted.

--
	Andreas Kupries <and...@Ac...>
	Developer @ http://www.ActiveState.com
	Tel: +1 778-786-1122

Re: [TCLCORE] CFV: TIP 283

[Joe E. <jen...@fl...>]

Miguel wrote:
> As promised, here's the call for votes on
> TIP #283: Modify Ensemble Command Resolution Behaviour

TIP #283: NO.

[ I was going to vote PRESENT on this one, since I didn't
  understand the full ramifications.  I *still* don't understand
  the full ramifications, but it's starting to look like it's
  not such a great idea; it opens up a barrel of worms that's
  perhaps better left closed. ]


--Joe English

  jen...@fl...

Re: [TCLCORE] CFV: TIP 283

[miguel <ms...@us...>]

miguel wrote:
> As promised, here's the call for votes on
> 
> TIP #283: Modify Ensemble Command Resolution Behaviour

As both author and sponsor of the TIP, I would like to suspend the vote 
and withdraw the tip pending a clarification of the issues raised.

The patch was prepared and tested with the patch in [Bug 1577492], and 
submitted to a vote. That bugfix was later reverted for reasons 
independent of this TIP.

The adaptation of this TIP's implementation to function without that 
patch caused the [clock] auto initialisation to fail. That issue is easy 
enough to fix, but it worked as a canary. The failure has uncovered 
deeper issues that need to be thought about and resolved properly before 
considering this modification to the ensemble's resolution behaviour.

Should a withdrawal not be possible during a vote, I herewith change my 
vote to "NO". In any case, I expect to resubmit a similar TIP in the 
near future, in time for 8.5 consideration.

Please accept my apologies.

Miguel