viper-mode undo bug introduced between Nov 10 and Nov 14

viper-mode undo bug introduced between Nov 10 and Nov 14

[Jim Meyering]

Hi, I noticed that "undo" in viper-mode stopped working properly (in
built-from-git master emacs) back in November and filed a bug for it.

https://bugs.gnu.org/22295

I still haven't found time to delve into it, so am posting here in the
hopes that someone else will get to it.

Re: viper-mode undo bug introduced between Nov 10 and Nov 14

[Eli Zaretskii]

> From: Jim Meyering <jim@meyering.net>
> Date: Thu, 14 Apr 2016 13:46:32 -0700
> 
> Hi, I noticed that "undo" in viper-mode stopped working properly (in
> built-from-git master emacs) back in November and filed a bug for it.
> 
> https://bugs.gnu.org/22295
> 
> I still haven't found time to delve into it, so am posting here in the
> hopes that someone else will get to it.

Phillip, could you please take a look at that bug?

Thanks.

Re: viper-mode undo bug introduced between Nov 10 and Nov 14

[Jim Meyering]

On Fri, Apr 15, 2016 at 12:19 AM, Eli Zaretskii <eliz@gnu.org> wrote:
>> From: Jim Meyering <jim@meyering.net>
>> Date: Thu, 14 Apr 2016 13:46:32 -0700
>>
>> Hi, I noticed that "undo" in viper-mode stopped working properly (in
>> built-from-git master emacs) back in November and filed a bug for it.
>>
>> https://bugs.gnu.org/22295
>>
>> I still haven't found time to delve into it, so am posting here in the
>> hopes that someone else will get to it.
>
> Phillip, could you please take a look at that bug?

Thanks for asking for help.
This is still a problem. However, now there is an additional bug in viper mode:

When I run e.g., "!!" to tell it to replace the text of the current
line with the output of some shell command, I see the usual "!" prompt
asking me for that command in the minibuffer. Let's say I type "cat"
then enter. So far so good.

However, if I type "!!" again, this time I see "!(default cat) ", and
let's say I want to accept that default and hit return. Then emacs
attempts to run the command "(default cat)". Oops. Even if I type
something, it is only appended, so if I type "cat" again, emacs runs
"(default cat) cat".

This appears to have been introduced some time on or before May 6th.

Re: viper-mode undo bug introduced between Nov 10 and Nov 14

[Eli Zaretskii]

> From: Jim Meyering <jim@meyering.net>
> Date: Tue, 10 May 2016 21:33:15 -0700
> Cc: Phillip Lord <phillip.lord@russet.org.uk>, emacs-devel <emacs-devel@gnu.org>
> 
> However, now there is an additional bug in viper mode:
> 
> When I run e.g., "!!" to tell it to replace the text of the current
> line with the output of some shell command, I see the usual "!" prompt
> asking me for that command in the minibuffer. Let's say I type "cat"
> then enter. So far so good.
> 
> However, if I type "!!" again, this time I see "!(default cat) ", and
> let's say I want to accept that default and hit return. Then emacs
> attempts to run the command "(default cat)". Oops. Even if I type
> something, it is only appended, so if I type "cat" again, emacs runs
> "(default cat) cat".
> 
> This appears to have been introduced some time on or before May 6th.

On which branch do you see this new problem?  Does it happen on
emacs-25, the release branch?

In any case, please file a separate bug report about that, as it
doesn't seem to be related to the undo problem (AFAIU).

Re: viper-mode undo bug introduced between Nov 10 and Nov 14

[Jim Meyering]

On Wed, May 11, 2016 at 12:54 AM, Eli Zaretskii <eliz@gnu.org> wrote:
>> From: Jim Meyering <jim@meyering.net>
>> Date: Tue, 10 May 2016 21:33:15 -0700
>> Cc: Phillip Lord <phillip.lord@russet.org.uk>, emacs-devel <emacs-devel@gnu.org>
>>
>> However, now there is an additional bug in viper mode:
>>
>> When I run e.g., "!!" to tell it to replace the text of the current
>> line with the output of some shell command, I see the usual "!" prompt
>> asking me for that command in the minibuffer. Let's say I type "cat"
>> then enter. So far so good.
>>
>> However, if I type "!!" again, this time I see "!(default cat) ", and
>> let's say I want to accept that default and hit return. Then emacs
>> attempts to run the command "(default cat)". Oops. Even if I type
>> something, it is only appended, so if I type "cat" again, emacs runs
>> "(default cat) cat".
>>
>> This appears to have been introduced some time on or before May 6th.
>
> On which branch do you see this new problem?  Does it happen on
> emacs-25, the release branch?
>
> In any case, please file a separate bug report about that, as it
> doesn't seem to be related to the undo problem (AFAIU).

I traced the new problem to commit
088acab3831b45e0e0749705226b8680076df4b6, aka
emacs-24.5-rc3-fixed-9210-g088acab. When I revert that change and
rebuild from emacs.git's master branch, that particular problem (which
also caused repeated "/"-based searching to mistakenly use the
"(default ...)" prefix) works once again.

If it is still necessary tomorrow, I'll create a separate bug report.

Re: viper-mode undo bug introduced between Nov 10 and Nov 14

[Eli Zaretskii]

> From: Jim Meyering <jim@meyering.net>
> Date: Thu, 12 May 2016 22:01:48 -0700
> Cc: emacs-devel <emacs-devel@gnu.org>, Phillip Lord <phillip.lord@russet.org.uk>
> 
> > On which branch do you see this new problem?  Does it happen on
> > emacs-25, the release branch?
> >
> > In any case, please file a separate bug report about that, as it
> > doesn't seem to be related to the undo problem (AFAIU).
> 
> I traced the new problem to commit
> 088acab3831b45e0e0749705226b8680076df4b6, aka
> emacs-24.5-rc3-fixed-9210-g088acab. When I revert that change and
> rebuild from emacs.git's master branch, that particular problem (which
> also caused repeated "/"-based searching to mistakenly use the
> "(default ...)" prefix) works once again.
> 
> If it is still necessary tomorrow, I'll create a separate bug report.

Yes, please file a separate bug report.

For the record, that commit on the master branch switched viper to
lexical-binding, so reverting it is probably out of the question.  We
need to look for a solution that keeps the lexical-binding.

Thanks.

Re: viper-mode undo bug introduced between Nov 10 and Nov 14

[Jim Meyering]

On Wed, May 11, 2016 at 12:54 AM, Eli Zaretskii <eliz@gnu.org> wrote:
>> From: Jim Meyering <jim@meyering.net>
>> Date: Tue, 10 May 2016 21:33:15 -0700
>> Cc: Phillip Lord <phillip.lord@russet.org.uk>, emacs-devel <emacs-devel@gnu.org>
>>
>> However, now there is an additional bug in viper mode:
>>
>> When I run e.g., "!!" to tell it to replace the text of the current
>> line with the output of some shell command, I see the usual "!" prompt
>> asking me for that command in the minibuffer. Let's say I type "cat"
>> then enter. So far so good.
>>
>> However, if I type "!!" again, this time I see "!(default cat) ", and
>> let's say I want to accept that default and hit return. Then emacs
>> attempts to run the command "(default cat)". Oops. Even if I type
>> something, it is only appended, so if I type "cat" again, emacs runs
>> "(default cat) cat".
>>
>> This appears to have been introduced some time on or before May 6th.
>
> On which branch do you see this new problem?  Does it happen on
> emacs-25, the release branch?
>
> In any case, please file a separate bug report about that, as it
> doesn't seem to be related to the undo problem (AFAIU).

I traced the new problem to commit
088acab3831b45e0e0749705226b8680076df4b6, aka
emacs-24.5-rc3-fixed-9210-g088acab. When I revert that change and
rebuild from emacs.git's master branch, that particular problem (which
also caused repeated "/"-based searching to mistakenly use the
"(default ...)" prefix) goes away.

If it is still necessary tomorrow, I'll create a separate bug report.