bug#33428: 27.0.50; edebug-print-{length,level} customization types

bug#33428: 27.0.50; edebug-print-{length,level} customization types

[Stephen Berman]

The doc strings of edebug-print-length and edebug-print-level imply that
nil is a valid value and indeed it is.  Yet trying to set the value to
nil via `M-x customize-option' fails with the error "This field should
contain an integer", because nil does not fit the customization type
:integer used in the definitions of these user options.  It's been like
this since they were changed from defvars to defcustoms more than 20
years ago (though "If non-nil" was added to the doc strings only a
little less than 10 years ago), so should the fix (below) go into
master, or is it on the order of a documentation fix and suitable for
the release branch?

diff --git a/lisp/emacs-lisp/edebug.el b/lisp/emacs-lisp/edebug.el
index 15f68a62ac..b50b0383da 100644
--- a/lisp/emacs-lisp/edebug.el
+++ b/lisp/emacs-lisp/edebug.el
@@ -192,11 +192,11 @@ edebug-continue-kbd-macro
 
 (defcustom edebug-print-length 50
   "If non-nil, default value of `print-length' for printing results in Edebug."
-  :type 'integer
+  :type '(choice integer (const nil))
   :group 'edebug)
 (defcustom edebug-print-level 50
   "If non-nil, default value of `print-level' for printing results in Edebug."
-  :type 'integer
+  :type '(choice integer (const nil))
   :group 'edebug)
 (defcustom edebug-print-circle t
   "If non-nil, default value of `print-circle' for printing results in Edebug."


In GNU Emacs 27.0.50 (build 13, x86_64-pc-linux-gnu, GTK+ Version 3.22.30)
 of 2018-11-18 built on rosalinde
Repository revision: 5d140800398287c20230dda79162a7c40016d88e
Repository branch: master
Windowing system distributor 'The X.Org Foundation', version 11.0.12001000
System Description: 8.3

bug#33428: 27.0.50; edebug-print-{length,level} customization types

[Eli Zaretskii]

> From: Stephen Berman <stephen.berman@gmx.net>
> Date: Mon, 19 Nov 2018 11:32:51 +0100
> 
> The doc strings of edebug-print-length and edebug-print-level imply that
> nil is a valid value and indeed it is.  Yet trying to set the value to
> nil via `M-x customize-option' fails with the error "This field should
> contain an integer", because nil does not fit the customization type
> :integer used in the definitions of these user options.  It's been like
> this since they were changed from defvars to defcustoms more than 20
> years ago (though "If non-nil" was added to the doc strings only a
> little less than 10 years ago), so should the fix (below) go into
> master, or is it on the order of a documentation fix and suitable for
> the release branch?

It's fine with me to put this on the release branch.

Thanks.

bug#33428: 27.0.50; edebug-print-{length,level} customization types

[Stephen Berman]

On Mon, 19 Nov 2018 20:37:05 +0200 Eli Zaretskii <eliz@gnu.org> wrote:

>> From: Stephen Berman <stephen.berman@gmx.net>
>> Date: Mon, 19 Nov 2018 11:32:51 +0100
>> 
>> The doc strings of edebug-print-length and edebug-print-level imply that
>> nil is a valid value and indeed it is.  Yet trying to set the value to
>> nil via `M-x customize-option' fails with the error "This field should
>> contain an integer", because nil does not fit the customization type
>> :integer used in the definitions of these user options.  It's been like
>> this since they were changed from defvars to defcustoms more than 20
>> years ago (though "If non-nil" was added to the doc strings only a
>> little less than 10 years ago), so should the fix (below) go into
>> master, or is it on the order of a documentation fix and suitable for
>> the release branch?
>
> It's fine with me to put this on the release branch.
>
> Thanks.

Done in commit d667318a7f.

Steve Berman

bug#33428: 27.0.50; edebug-print-{length,level} customization types

[Stephen Berman]

On Mon, 19 Nov 2018 23:15:21 +0100 Stephen Berman <stephen.berman@gmx.net> wrote:

> On Mon, 19 Nov 2018 20:37:05 +0200 Eli Zaretskii <eliz@gnu.org> wrote:
>
>>> From: Stephen Berman <stephen.berman@gmx.net>
>>> Date: Mon, 19 Nov 2018 11:32:51 +0100
>>> 
>>> The doc strings of edebug-print-length and edebug-print-level imply that
>>> nil is a valid value and indeed it is.  Yet trying to set the value to
>>> nil via `M-x customize-option' fails with the error "This field should
>>> contain an integer", because nil does not fit the customization type
>>> :integer used in the definitions of these user options.  It's been like
>>> this since they were changed from defvars to defcustoms more than 20
>>> years ago (though "If non-nil" was added to the doc strings only a
>>> little less than 10 years ago), so should the fix (below) go into
>>> master, or is it on the order of a documentation fix and suitable for
>>> the release branch?
>>
>> It's fine with me to put this on the release branch.
>>
>> Thanks.
>
> Done in commit d667318a7f.
>
> Steve Berman

Oops, and closing the bug.