gpg encryption warning

gpg encryption warning

[Colin Baxter]

On gpg encrypting (symmetrical) a file I get the warning:

mouse-minibuffer-check: Wrong type argument: window-valid-p, #<window 40>

after clicking ok in the encryption dialogue buffer. The actual
encryption seems to work, with password and repeat password prompts.  I
am using emacs-25.1.50.1, git-pulled and built today. I do not get the
warning with emacs-24.5.1.

Colin.

I'm sorry if I've already posted this - I'm having a bad gnus day with lots of
fat finger work.

gpg encryption warning

[Colin Baxter]

On gpg encrypting (symmetrical) a file I get the warning:

mouse-minibuffer-check: Wrong type argument: window-valid-p, #<window 40>

after clicking ok in the encryption dialogue buffer. The actual
encryption seems to work, with password and repeat password prompts.  I
am using emacs-25.1.50.1, git-pulled and built today. I do not get the
warning with emacs-24.5.1.

Colin.

Re: gpg encryption warning

[martin rudalics]

 > On gpg encrypting (symmetrical) a file I get the warning:
 >
 > mouse-minibuffer-check: Wrong type argument: window-valid-p, #<window 40>
 >
 > after clicking ok in the encryption dialogue buffer. The actual
 > encryption seems to work, with password and repeat password prompts.  I
 > am using emacs-25.1.50.1, git-pulled and built today. I do not get the
 > warning with emacs-24.5.1.

Thanks for the report.  Can you tell me who calls
‘mouse-minibuffer-check’ here (and why)?

martin

Re: gpg encryption warning

[Colin Baxter]

On Wed, Apr 20 2016, martin rudalics wrote:

>> On gpg encrypting (symmetrical) a file I get the warning:
>>
>> mouse-minibuffer-check: Wrong type argument: window-valid-p, #<window 40>
>>
>> after clicking ok in the encryption dialogue buffer. The actual
>> encryption seems to work, with password and repeat password prompts.  I
>> am using emacs-25.1.50.1, git-pulled and built today. I do not get the
>> warning with emacs-24.5.1.
>
> Thanks for the report.  Can you tell me who calls
> ‘mouse-minibuffer-check’ here (and why)?
>
> martin

Hi Martin,

I don't fully understand what you mean - I'm a newbe, sorry. If I call
the command epa-encrypt-file (either via M-x or the tools ->
Encryption/Decryption menu) I'm presented with a Keys buffer asking me
to select recipients for encryption. I want to do symmetric encryption
so I select no one and click the OK button with my mouse. At this point
I get the "mouse-minibuffer-check" warning in the Message buffer, and a
password prompt in the echo buffer. If I wish, I can then ignore the
warning and enter a password and proceed as normal. However, emacs-24.5
does not give any warnings at all when an identical procedure is followed.

Let me know if I can do anything else in order to provide perhaps more
useful information.

Thanks.

Colin.

Re: gpg encryption warning

[martin rudalics]

 > Hi Martin,

Hi Colin,

 > I don't fully understand what you mean - I'm a newbe, sorry. If I call
 > the command epa-encrypt-file (either via M-x or the tools ->
 > Encryption/Decryption menu) I'm presented with a Keys buffer asking me
 > to select recipients for encryption. I want to do symmetric encryption
 > so I select no one and click the OK button with my mouse. At this point
 > I get the "mouse-minibuffer-check" warning in the Message buffer, and a
 > password prompt in the echo buffer. If I wish, I can then ignore the
 > warning and enter a password and proceed as normal. However, emacs-24.5
 > does not give any warnings at all when an identical procedure is followed.
 >
 > Let me know if I can do anything else in order to provide perhaps more
 > useful information.

Our aim will be to fix this unless someone else beats us to it.  This
bug is apparently not common and maybe caused by your setup.  So hardly
anyone will care.

The first step is to determine the call chain for this error.  In the
'Options' menu select 'Enter Debugger on Error' and try to repeat your
scenario (I hope it is fairly deterministic).  This should pop up a
buffer with the backtrace.  Send me this information.  Normally we
should do this on the list but let's postpone that until we either have
a fix or we are stuck with a good recipe of how to reproduce the bug.

martin

Re: gpg encryption warning

[Colin Baxter]

Hi Martin,

If it helps, I can confirm that I get no error message is I edit and
then save (encrypt) a gpg file that's in an un-encrypted buffer. The
issue seems to be only with first-time encryption.

Emacs bisection is new to me and I'd have to be "talked through it"; you may,
understandably, not want to entrust a newbe with that job.

Best wishes,

Colin.

Re: gpg encryption warning

[martin rudalics]

 > If it helps, I can confirm that I get no error message is I edit and
 > then save (encrypt)

When you do that (I hardly ever encrypt) do you have to re-enter the password?

 > a gpg file that's in an un-encrypted buffer. The
 > issue seems to be only with first-time encryption.
 >
 > Emacs bisection is new to me and I'd have to be "talked through it"; you may,
 > understandably, not want to entrust a newbe with that job.

It's a trivial job, all you need is a fairly fast machine (one that does
a reasonable complex "make" of emacs in less than 10 minutes, preferably
muss less.  In the directory where you git pull emacs you'd simply have
to do

git bisect start
git bisect bad
git bisect good ...

where ... stands for the hash of the version that works (probably
24.5.1).  Then "make" emacs as you usually do and if the result is good
mark it as good via

git bisect good

and if it's bad as

git bisect bad

One description of git bisect is this

https://git-scm.com/docs/git-bisect

but there might be better ones.

One caveat: Often a specific commit breaks building emacs and then you
have to skip this commit and usually go back until you find one that
builds.

I don't bisect more often than once in a month because on my machine it
usually takes me half a day to finish.

martin, who doesn't want to talk you into it ;-)

Re: gpg encryption warning

[Yuri Khan]

On Fri, Apr 22, 2016 at 7:10 PM, martin rudalics <rudalics@gmx.at> wrote:

> One description of git bisect is this
>
> https://git-scm.com/docs/git-bisect
>
> but there might be better ones.

I learned Git while preparing some patches for Wine. They have a nice
bisection walkthrough at:

https://wiki.winehq.org/Regression_Testing#Running_the_Bisection

Re: gpg encryption warning

[Colin Baxter]

On Fri, Apr 22 2016, martin rudalics wrote:

>> If it helps, I can confirm that I get no error message is I edit and
>> then save (encrypt)
>
> When you do that (I hardly ever encrypt) do you have to re-enter the
> password?

Yes. You enter the password and then again for confirmation.


>

>
> martin, who doesn't want to talk you into it ;-)


Ok, I'll have a go. I should have some time this week.

Best wishes,

Colin.

Re: gpg encryption warning

[Colin Baxter]

On Fri, Apr 22 2016, martin rudalics wrote:

>> If it helps, I can confirm that I get no error message is I edit and
>> then save (encrypt)
>
> When you do that (I hardly ever encrypt) do you have to re-enter the password?
>
>> a gpg file that's in an un-encrypted buffer. The
>> issue seems to be only with first-time encryption.
>>
>> Emacs bisection is new to me and I'd have to be "talked through it"; you may,
>> understandably, not want to entrust a newbe with that job.
>
> It's a trivial job, all you need is a fairly fast machine (one that does
> a reasonable complex "make" of emacs in less than 10 minutes, preferably
> muss less.  In the directory where you git pull emacs you'd simply have
> to d
 ---- snip --------


Well bisecting is not going well. It is taking forever, which is not
helped by work commitments. I'll keep going whenever I have the time,
but should anyone wish to take over they'd be welcome!

Best wishes,

Colin.