Re: master 24a8cc5: Define revert-buffer-function for *Memory Report*

Re: master 24a8cc5: Define revert-buffer-function for *Memory Report*

[Stefan Monnier]

> +  (setq-local revert-buffer-function (lambda (_ignore-auto _noconfirm)
> +                                       (memory-report)))

I hope someone will find it in them to tackle the problem of common
"reference" format which can be shared between `revert-buffer-function`,
`bookmark`, `help-xref`, ...


        Stefan

Re: master 24a8cc5: Define revert-buffer-function for *Memory Report*

[Lars Ingebrigtsen]

Stefan Monnier <monnier@iro.umontreal.ca> writes:

>> +  (setq-local revert-buffer-function (lambda (_ignore-auto _noconfirm)
>> +                                       (memory-report)))
>
> I hope someone will find it in them to tackle the problem of common
> "reference" format which can be shared between `revert-buffer-function`,
> `bookmark`, `help-xref`, ...

What problem is that?

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

Re: master 24a8cc5: Define revert-buffer-function for *Memory Report*

[Stefan Monnier]

Lars Ingebrigtsen [2021-07-17 15:46:14] wrote:
> Stefan Monnier <monnier@iro.umontreal.ca> writes:
>>> +  (setq-local revert-buffer-function (lambda (_ignore-auto _noconfirm)
>>> +                                       (memory-report)))
>>
>> I hope someone will find it in them to tackle the problem of common
>> "reference" format which can be shared between `revert-buffer-function`,
>> `bookmark`, `help-xref`, ...
>
> What problem is that?

The problem that those should all rely on a single piece of information
provided by the mode and it should work for help-xref, for bookmark, for
revert-buffer instead of having to set those separately, each one using
a slightly different representation and calling convention.


        Stefan

Re: master 24a8cc5: Define revert-buffer-function for *Memory Report*

[Lars Ingebrigtsen]

Stefan Monnier <monnier@iro.umontreal.ca> writes:

> The problem that those should all rely on a single piece of information
> provided by the mode and it should work for help-xref, for bookmark, for
> revert-buffer instead of having to set those separately, each one using
> a slightly different representation and calling convention.

Ah, yeah, that's true -- the use cases do overlap somewhat.  And it'd be
nice to go able to bookmark more special-mode buffers.

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

Re: master 24a8cc5: Define revert-buffer-function for *Memory Report*

[Juri Linkov]

>> The problem that those should all rely on a single piece of information
>> provided by the mode and it should work for help-xref, for bookmark, for
>> revert-buffer instead of having to set those separately, each one using
>> a slightly different representation and calling convention.
>
> Ah, yeah, that's true -- the use cases do overlap somewhat.  And it'd be
> nice to go able to bookmark more special-mode buffers.

Using revert-buffer-function for bookmarking then is similar to
saving revert-buffer-function to the desktop file
to be able to revisit it after restarting Emacs.
This will remove the need of implementing both hooks:
bookmark-make-record-function and desktop-buffer-mode-handlers.