GNU bug report logs - #45033
28.0.50; New option gnus-registery-register-all-p

Previous Next

To add a comment to this bug, you must first unarchive it, by sending
a message to control AT debbugs.gnu.org, with unarchive 45033 in the body.
You can then email your comments to 45033 AT debbugs.gnu.org in the normal way.

Toggle the display of automated, internal messages from the tracker.

Message #5 received at submit <at> debbugs.gnu.org (full text, mbox):

From: Eric Abrahamsen <eric <at> ericabrahamsen.net>
To: bug-gnu-emacs <at> gnu.org
Subject: 28.0.50; New option gnus-registery-register-all-p
Date: Thu, 03 Dec 2020 16:10:24 -0800

[Message part 1 (text/plain, inline)]

The Gnus registry keeps track of all received messages in a series of
databases, and allows the user to split follow-up messages into the same
group as the parent, as well as associate

It currently creates database entries for all messages, which slows down
summary buffer creation, as well as Gnus' saving process.

This patch introduces a new customization option,
`gnus-registry-register-all-p', which, when nil, won't create database
entries automatically. The registry will only contain entries that the
user (or other packages) has created manually. If the user isn't using
automatic splitting, this can provide a significant speedup for general
Gnus usage.

Discussions on gnus.general lead me to believe that universal splitting
isn't used all that often, and that most users are very surprised to
find that they have a 50MB registry file on disk. Therefore I've
defaulted this option to nil, which is a change from previous behavior.
It would be perfectly easy to default to t if this seems inappropriate.

Eric

[0001-New-option-gnus-registry-register-all-p-default-to-n.patch (text/x-patch, attachment)]

Message #8 received at 45033 <at> debbugs.gnu.org (full text, mbox):

From: Lars Ingebrigtsen <larsi <at> gnus.org>
To: Eric Abrahamsen <eric <at> ericabrahamsen.net>
Cc: 45033 <at> debbugs.gnu.org
Subject: Re: bug#45033: 28.0.50; New option gnus-registery-register-all-p
Date: Fri, 04 Dec 2020 11:11:12 +0100

Eric Abrahamsen <eric <at> ericabrahamsen.net> writes:

> This patch introduces a new customization option,
> `gnus-registry-register-all-p', which, when nil, won't create database
> entries automatically. The registry will only contain entries that the
> user (or other packages) has created manually. If the user isn't using
> automatic splitting, this can provide a significant speedup for general
> Gnus usage.

Makes sense to me.  However, the option shouldn't have a name ending in
-p -- that's for predicate functions only.  `gnus-registry-register-all'
would be a better name.

> Discussions on gnus.general lead me to believe that universal splitting
> isn't used all that often, and that most users are very surprised to
> find that they have a 50MB registry file on disk. Therefore I've
> defaulted this option to nil, which is a change from previous behavior.
> It would be perfectly easy to default to t if this seems inappropriate.

I think defaulting it to t makes more sense (besides being more
backwards-compatible) -- the registry isn't on by default, after all.

-- 
(domestic pets only, the antidote for overdose, milk.)
   bloggy blog: http://lars.ingebrigtsen.no

Message #11 received at 45033 <at> debbugs.gnu.org (full text, mbox):

From: Eric Abrahamsen <eric <at> ericabrahamsen.net>
To: Lars Ingebrigtsen <larsi <at> gnus.org>
Cc: 45033 <at> debbugs.gnu.org
Subject: Re: bug#45033: 28.0.50; New option gnus-registery-register-all-p
Date: Fri, 04 Dec 2020 09:35:52 -0800

[Message part 1 (text/plain, inline)]

Lars Ingebrigtsen <larsi <at> gnus.org> writes:

> Eric Abrahamsen <eric <at> ericabrahamsen.net> writes:
>
>> This patch introduces a new customization option,
>> `gnus-registry-register-all-p', which, when nil, won't create database
>> entries automatically. The registry will only contain entries that the
>> user (or other packages) has created manually. If the user isn't using
>> automatic splitting, this can provide a significant speedup for general
>> Gnus usage.
>
> Makes sense to me.  However, the option shouldn't have a name ending in
> -p -- that's for predicate functions only.  `gnus-registry-register-all'
> would be a better name.

Ah, I hesitated about that. Thanks.

>> Discussions on gnus.general lead me to believe that universal splitting
>> isn't used all that often, and that most users are very surprised to
>> find that they have a 50MB registry file on disk. Therefore I've
>> defaulted this option to nil, which is a change from previous behavior.
>> It would be perfectly easy to default to t if this seems inappropriate.
>
> I think defaulting it to t makes more sense (besides being more
> backwards-compatible) -- the registry isn't on by default, after all.

Makes sense. Here's a new version of the patch with option name edited
and docs adjusted.

[0001-New-option-gnus-registry-register-all.patch (text/x-patch, attachment)]

Message #14 received at 45033 <at> debbugs.gnu.org (full text, mbox):

From: Eric Abrahamsen <eric <at> ericabrahamsen.net>
To: Lars Ingebrigtsen <larsi <at> gnus.org>
Cc: 45033 <at> debbugs.gnu.org
Subject: Re: bug#45033: 28.0.50; New option gnus-registery-register-all-p
Date: Fri, 11 Dec 2020 09:17:43 -0800

Eric Abrahamsen <eric <at> ericabrahamsen.net> writes:

> Lars Ingebrigtsen <larsi <at> gnus.org> writes:
>
>> Eric Abrahamsen <eric <at> ericabrahamsen.net> writes:
>>
>>> This patch introduces a new customization option,
>>> `gnus-registry-register-all-p', which, when nil, won't create database
>>> entries automatically. The registry will only contain entries that the
>>> user (or other packages) has created manually. If the user isn't using
>>> automatic splitting, this can provide a significant speedup for general
>>> Gnus usage.
>>
>> Makes sense to me.  However, the option shouldn't have a name ending in
>> -p -- that's for predicate functions only.  `gnus-registry-register-all'
>> would be a better name.
>
> Ah, I hesitated about that. Thanks.
>
>>> Discussions on gnus.general lead me to believe that universal splitting
>>> isn't used all that often, and that most users are very surprised to
>>> find that they have a 50MB registry file on disk. Therefore I've
>>> defaulted this option to nil, which is a change from previous behavior.
>>> It would be perfectly easy to default to t if this seems inappropriate.
>>
>> I think defaulting it to t makes more sense (besides being more
>> backwards-compatible) -- the registry isn't on by default, after all.
>
> Makes sense. Here's a new version of the patch with option name edited
> and docs adjusted.

If there aren't any more comments on this, I'll go ahead and push?

Message #17 received at 45033 <at> debbugs.gnu.org (full text, mbox):

From: Lars Ingebrigtsen <larsi <at> gnus.org>
To: Eric Abrahamsen <eric <at> ericabrahamsen.net>
Cc: 45033 <at> debbugs.gnu.org
Subject: Re: bug#45033: 28.0.50; New option gnus-registery-register-all-p
Date: Sat, 12 Dec 2020 12:00:04 +0100

Eric Abrahamsen <eric <at> ericabrahamsen.net> writes:

> If there aren't any more comments on this, I'll go ahead and push?

Sure, go ahead.

-- 
(domestic pets only, the antidote for overdose, milk.)
   bloggy blog: http://lars.ingebrigtsen.no

Message #20 received at 45033 <at> debbugs.gnu.org (full text, mbox):

From: Eric Abrahamsen <eric <at> ericabrahamsen.net>
To: Lars Ingebrigtsen <larsi <at> gnus.org>
Cc: 45033 <at> debbugs.gnu.org, 45033-done <at> debbugs.gnu.org
Subject: Re: bug#45033: 28.0.50; New option gnus-registery-register-all-p
Date: Sat, 12 Dec 2020 09:07:27 -0800

Lars Ingebrigtsen <larsi <at> gnus.org> writes:

> Eric Abrahamsen <eric <at> ericabrahamsen.net> writes:
>
>> If there aren't any more comments on this, I'll go ahead and push?
>
> Sure, go ahead.

Thanks, done.

Message #28 received at 45033 <at> debbugs.gnu.org (full text, mbox):

From: "Basil L. Contovounesios" <contovob <at> tcd.ie>
To: Eric Abrahamsen <eric <at> ericabrahamsen.net>
Cc: Lars Ingebrigtsen <larsi <at> gnus.org>, 45033 <at> debbugs.gnu.org
Subject: Re: bug#45033: 28.0.50; New option gnus-registery-register-all-p
Date: Fri, 18 Dec 2020 23:16:11 +0000

Eric Abrahamsen <eric <at> ericabrahamsen.net> writes:

> From 1b94a5bc72ac56a132f73b5d1732800ac0ee830d Mon Sep 17 00:00:00 2001
> From: Eric Abrahamsen <eric <at> ericabrahamsen.net>
> Date: Thu, 3 Dec 2020 15:58:57 -0800
> Subject: [PATCH] New option gnus-registry-register-all
>
> * lisp/gnus/gnus-registry.el (gnus-registry-register-all): If nil,
> the registry won't automatically create new entries for all seen
> messages. Defaults to t to preserve previous behavior.
            ^^^^^^^^^^^^^

[...]

> +If @code{gnus-registry-register-all} is non-nil (the default), the
                                           ^^^^^^^^^^^^^^^^^^^^^
> +registry will perform splitting for all messages.  If it is nil,
> +splitting will only happen for children of messages you've explicitly
> +registered.

[...]

> ++++
> +*** New user option 'gnus-registry-register-all'.
> +
> +If non-nil (the default), create registry entries for all messages.
      ^^^^^^^^^^^^^^^^^^^^^
> +If nil, don't automatically create entries, they must be created
> +manually.

[...]

> +(defcustom gnus-registry-register-all nil
                                         ^^^
> +  "If non-nil, register all articles in the registry."
> +  :type 'boolean
> +  :version "28.1")

Was the user option meant to be initialised to t?

Thanks,

-- 
Basil

Message #31 received at 45033 <at> debbugs.gnu.org (full text, mbox):

From: Eric Abrahamsen <eric <at> ericabrahamsen.net>
To: "Basil L. Contovounesios" <contovob <at> tcd.ie>
Cc: Lars Ingebrigtsen <larsi <at> gnus.org>, 45033 <at> debbugs.gnu.org
Subject: Re: bug#45033: 28.0.50; New option gnus-registery-register-all-p
Date: Sat, 19 Dec 2020 10:04:05 -0800

On 12/18/20 23:16 PM, Basil L. Contovounesios wrote:
> Eric Abrahamsen <eric <at> ericabrahamsen.net> writes:
>
>> From 1b94a5bc72ac56a132f73b5d1732800ac0ee830d Mon Sep 17 00:00:00 2001
>> From: Eric Abrahamsen <eric <at> ericabrahamsen.net>
>> Date: Thu, 3 Dec 2020 15:58:57 -0800
>> Subject: [PATCH] New option gnus-registry-register-all
>>
>> * lisp/gnus/gnus-registry.el (gnus-registry-register-all): If nil,
>> the registry won't automatically create new entries for all seen
>> messages. Defaults to t to preserve previous behavior.
>             ^^^^^^^^^^^^^
>
> [...]
>
>> +If @code{gnus-registry-register-all} is non-nil (the default), the
>                                            ^^^^^^^^^^^^^^^^^^^^^
>> +registry will perform splitting for all messages.  If it is nil,
>> +splitting will only happen for children of messages you've explicitly
>> +registered.
>
> [...]
>
>> ++++
>> +*** New user option 'gnus-registry-register-all'.
>> +
>> +If non-nil (the default), create registry entries for all messages.
>       ^^^^^^^^^^^^^^^^^^^^^
>> +If nil, don't automatically create entries, they must be created
>> +manually.
>
> [...]
>
>> +(defcustom gnus-registry-register-all nil
>                                          ^^^
>> +  "If non-nil, register all articles in the registry."
>> +  :type 'boolean
>> +  :version "28.1")
>
> Was the user option meant to be initialised to t?

Bah, it was originally nil, then was supposed to be t, and it looks like
I made all the changes except the final crucial one. Thanks for pointing
this out -- I'll take another look and make sure I haven't screwed
anything else up, then fix it.

Message #34 received at 45033 <at> debbugs.gnu.org (full text, mbox):

From: Eric Abrahamsen <eric <at> ericabrahamsen.net>
To: "Basil L. Contovounesios" <contovob <at> tcd.ie>
Cc: Lars Ingebrigtsen <larsi <at> gnus.org>, 45033 <at> debbugs.gnu.org
Subject: Re: bug#45033: 28.0.50; New option gnus-registery-register-all-p
Date: Sat, 19 Dec 2020 15:25:05 -0800

Eric Abrahamsen <eric <at> ericabrahamsen.net> writes:

> On 12/18/20 23:16 PM, Basil L. Contovounesios wrote:
>> Eric Abrahamsen <eric <at> ericabrahamsen.net> writes:
>>
>>> From 1b94a5bc72ac56a132f73b5d1732800ac0ee830d Mon Sep 17 00:00:00 2001
>>> From: Eric Abrahamsen <eric <at> ericabrahamsen.net>
>>> Date: Thu, 3 Dec 2020 15:58:57 -0800
>>> Subject: [PATCH] New option gnus-registry-register-all
>>>
>>> * lisp/gnus/gnus-registry.el (gnus-registry-register-all): If nil,
>>> the registry won't automatically create new entries for all seen
>>> messages. Defaults to t to preserve previous behavior.
>>             ^^^^^^^^^^^^^
>>
>> [...]
>>
>>> +If @code{gnus-registry-register-all} is non-nil (the default), the
>>                                            ^^^^^^^^^^^^^^^^^^^^^
>>> +registry will perform splitting for all messages.  If it is nil,
>>> +splitting will only happen for children of messages you've explicitly
>>> +registered.
>>
>> [...]
>>
>>> ++++
>>> +*** New user option 'gnus-registry-register-all'.
>>> +
>>> +If non-nil (the default), create registry entries for all messages.
>>       ^^^^^^^^^^^^^^^^^^^^^
>>> +If nil, don't automatically create entries, they must be created
>>> +manually.
>>
>> [...]
>>
>>> +(defcustom gnus-registry-register-all nil
>>                                          ^^^
>>> +  "If non-nil, register all articles in the registry."
>>> +  :type 'boolean
>>> +  :version "28.1")
>>
>> Was the user option meant to be initialised to t?
>
> Bah, it was originally nil, then was supposed to be t, and it looks like
> I made all the changes except the final crucial one. Thanks for pointing
> this out -- I'll take another look and make sure I haven't screwed
> anything else up, then fix it.

Yup, looks like that was it.

Previous Next

GNU bug tracking system
Copyright (C) 1999 Darren O. Benham, 1997,2003 nCipher Corporation Ltd, 1994-97 Ian Jackson.