[RFC] Removal of [1]-like footnotes

[RFC] Removal of [1]-like footnotes

[Nicolas Goaziou]

Hello,

As discussed previously, I pushed changes about footnotes in a dedicated
branch, "wip-no-plain-fn", for testing.

In a nutshell, in this branch, Org no longer recognizes [1]-like
constructs as valid footnotes, an no longer spend time matching them.

As a consequence, "fn:" can now be unambiguously removed from label and
become part of the syntax. Thus, [fn:1] is labelled "1" and [fn:label]
is labelled "label".

In the process, I had to rewrite `org-footnote-normalize', which used to
turn every footnote in a document into [1]-like footnotes. Now, it turns
them into [fn:1] footnotes.

However, this function also contained a part dealing with non-Org
buffers, e.g. just before sending an email, this function could turn Org
footnotes into footnote.el-compatible footnotes. This is no longer
present currently in the branch. 

I don't really mind adding it back, but it ought to be a separate
function, with a different design. It is not really possible to treat
non-Org buffers as Org ones (cf. `orgstruct-mode' problems). However,
before spending time on it, I'd like to know if there is any incentive
to use it over, e.g. `footnote-mode' right from the start?

Feedback welcome.


Regards,

-- 
Nicolas Goaziou                                                0x80A93738

Re: [RFC] Removal of [1]-like footnotes

[Rasmus]

Hi,

Nicolas Goaziou <mail@nicolasgoaziou.fr> writes:

> As discussed previously, I pushed changes about footnotes in a dedicated
> branch, "wip-no-plain-fn", for testing.
>
> In a nutshell, in this branch, Org no longer recognizes [1]-like
> constructs as valid footnotes, an no longer spend time matching them.

Thanks a lot!

> As a consequence, "fn:" can now be unambiguously removed from label and
> become part of the syntax. Thus, [fn:1] is labelled "1" and [fn:label]
> is labelled "label".

Can I now write:

    X[fn:1]

    [1] foot

If so, why the need for the fanciness?  Why not just require label and
reference to be the same?  I doesn’t sound like something that would be
nice to have to explain to an Org newcomer.

> I don't really mind adding it back, but it ought to be a separate
> function, with a different design. It is not really possible to treat
> non-Org buffers as Org ones (cf. `orgstruct-mode' problems). However,
> before spending time on it, I'd like to know if there is any incentive
> to use it over, e.g. `footnote-mode' right from the start?

No there is not.  You can probably get the desired style with
(Footnote-set-style 'unicode).  We can eventually add keybindings for
footnote-mode in norgstruct that resemble Org, if necessary.

Rasmus

-- 
I feel emotional landscapes they puzzle me

Re: [RFC] Removal of [1]-like footnotes

[Nicolas Goaziou]

Hello,

Rasmus <rasmus@gmx.us> writes:

> Can I now write:
>
>     X[fn:1]
>
>     [1] foot

Uh? No.

> If so, why the need for the fanciness?  Why not just require label and
> reference to be the same?  I doesn’t sound like something that would be
> nice to have to explain to an Org newcomer.

I just mean that, internally, 

  (org-element-property :label reference) => "1"

not "fn:1" anymore. In the buffer, "fn:" is part of the syntax and, as
such, always present:

  X[fn:1]

  [fn:1] foot

Thanks for the feedback.


Regards,

-- 
Nicolas Goaziou

Re: [RFC] Removal of [1]-like footnotes

[Rasmus]

Nicolas Goaziou <mail@nicolasgoaziou.fr> writes:

> Hello,
>
> Rasmus <rasmus@gmx.us> writes:
>
>> Can I now write:
>>
>>     X[fn:1]
>>
>>     [1] foot
>
> Uh? No.
> [...]

Very good!  Thanks for clarifying.

Rasmus

-- 
A page of history is worth a volume of logic

Re: [RFC] Removal of [1]-like footnotes

[Thierry Banel]

[-- Attachment #1: Type: text/html, Size: 1573 bytes --]

Re: [RFC] Removal of [1]-like footnotes

[Simon Thum]

Same here (yes please!), for the same reason ;)

On 12/19/2015 04:27 PM, Thierry Banel wrote:
> I vote "yes".
> It was annoying to see bracketed numbers as links to nowhere.
>
> The Nobel prize for the detection of the neutrino _[1995]_ was
>    FrederickReines and the prizes for the discovery of neutrino
>    oscillations _[2015]_ were Takaaki Kajita & Arthur McDonald.
>
>    The resulting sets are [1,2,3], [1,2], _[1]_
>
>
> Le 17/12/2015 11:03, Nicolas Goaziou a écrit :
>> Hello,
>>
>> As discussed previously, I pushed changes about footnotes in a dedicated
>> branch, "wip-no-plain-fn", for testing.
>>
>> In a nutshell, in this branch, Org no longer recognizes [1]-like
>> constructs as valid footnotes, an no longer spend time matching them.
>>
>> As a consequence, "fn:" can now be unambiguously removed from label and
>> become part of the syntax. Thus, [fn:1] is labelled "1" and [fn:label]
>> is labelled "label".
>>
>>
>>
>

Re: [RFC] Removal of [1]-like footnotes

[Nicolas Goaziou]

Hello,

Simon Thum <simon.thum@gmx.de> writes:

> Same here (yes please!), for the same reason ;)
>
> On 12/19/2015 04:27 PM, Thierry Banel wrote:
>> I vote "yes".
>> It was annoying to see bracketed numbers as links to nowhere.
>>
>> The Nobel prize for the detection of the neutrino _[1995]_ was
>>    FrederickReines and the prizes for the discovery of neutrino
>>    oscillations _[2015]_ were Takaaki Kajita & Arthur McDonald.
>>
>>    The resulting sets are [1,2,3], [1,2], _[1]_
>>
>>
>> Le 17/12/2015 11:03, Nicolas Goaziou a écrit :
>>> Hello,
>>>
>>> As discussed previously, I pushed changes about footnotes in a dedicated
>>> branch, "wip-no-plain-fn", for testing.
>>>
>>> In a nutshell, in this branch, Org no longer recognizes [1]-like
>>> constructs as valid footnotes, an no longer spend time matching them.
>>>
>>> As a consequence, "fn:" can now be unambiguously removed from label and
>>> become part of the syntax. Thus, [fn:1] is labelled "1" and [fn:label]
>>> is labelled "label".

Pushed. Thanks to everyone for the feedback.


Regards,

-- 
Nicolas Goaziou