Re: [Bug-zile] Renaming tests

[Gary V. Vaughan]

Hi Reuben,

On 18 Mar 2012, at 23:30, Reuben Thomas wrote:

> On 15 March 2012 11:51, Gary V. Vaughan <address@hidden> wrote:
>> 
>> Found a little time for GNU hacking today, so...
> 
> Great!
> 
>> I cherry-picked all the tests patches from lua to master, after which
>> master now has 3 Zile failing test cases:
>> 
>>  tests/interactive/backward-char.el
>>  tests/interactive/backward-delete-char.el
>>  tests/interactive/backward-kill-word.el
>> 
>> The first two were fixed in the lua branch by 07c644b, and which I
>> attempted to port to master with 903ce3a.  The idea is simply to make
>> \C-? a valid read syntax for 0x7f and then add global key bindings so
>> that we can use it in execute-kbd-macro without worrying about what
>> terminfo says about the kbs cap.
> 
> Unfortunately you appear to have committed the bindings but not the
> read-syntax support. All three tests still fail, and git log -p only
> shows the new bindings, not any new support for the read-syntax.

Right, I wasn't sure how to add \C-? read-syntax support to the C version,
since all the keypress code is different, and didn't have time to work it
out.  We can always disable those tests until one of us has time to add
the read-syntax.

> Also, Emacs now fails five tests:
> 
> interactive/backward-char
> interactive/backward-kill-word
> interactive/backward-word
> interactive/end-of-line
> interactive/fill-paragraph
> 
> I presume these fail on the Lua branch as well, since it's the same
> Emacs! Any ideas?

No, all tests pass with Emacs-24.0.50 for me on the lua branch.  I guess
something is different between our respective Emacs builds?

>> The other one is a mystery, and doesn't fail on the lua branch, although
>> it uses \M-? inside execute-kbd-macro (the other two use \C-? and \M-\C-?,
>> so maybe I uncovered a new master branch bug with reading '?' inside
>> execute-kbd-macro...
> 
> I'm confused: of the three failing tests,
> interactive/backward-kill-word uses \C-\M-?, while the other two
> (interactive/backward{,-delete}-char) use \C-?. Hence, all three use
> \C-?.

My mistake then.

>> I also found a display bug in interactive mode on the C branch, but I'll
>> raise a new ticket at Savannah for that.
> 
> Fixed; thanks.

Not at all.  Thank you!

Cheers,
-- 
Gary V. Vaughan (gary AT gnu DOT org)