[Interest] Determanistic Org IDs

[Interest] Determanistic Org IDs

[TEC]

:PROPERTIES:
:ID:       1161308a-ef0c-4490-bb36-13634f8de727
:END:
User-agent: mu4e 1.4.10; emacs 26.3
Message-ID: <87r1t7mu52.fsf@gmail.com>
MIME-Version: 1.0
Content-Type: text/plain; format=flowed
Received-SPF: pass client-ip=2607:f8b0:4864:20::102b; envelope-from=tecosaur@gmail.com; helo=mail-pj1-x102b.google.com
X-detected-operating-system: by eggs.gnu.org: No matching host in p0f cache. That's all we know.
X-Spam_score_int: -7
X-Spam_score: -0.8
X-Spam_bar: /
X-Spam_report: (-0.8 / 5.0 requ) BAYES_00=-1.9,DKIM_ADSP_CUSTOM_MED=0.001,DKIM_INVALID=0.1,DKIM_SIGNED=0.1,FREEMAIL_FROM=0.001,NML_ADSP_CUSTOM_MED=0.9,RCVD_IN_DNSWL_NONE=-0.0001,SPF_HELO_NONE=0.001,SPF_PASS=-0.001 autolearn=no autolearn_force=no
X-Spam_action: no action

Hello everyone,

I have an html export of an org file checked into VCS. To try to 
reduce the 'noise' created by commiting the random org IDs 
changing each time, I took @alphapapa's unpackaged.el 
org-export-html-with-useful-ids-mode, tweaked it, and extended it 
to cover other elements.

The end result it that the resulting html export is fully 
determanistic. While I'm quite pleased with the result, it occured 
to me that others might be interested in having their exports 
behave like this.

From my usage, these are the percieved pros/cons.

Pros:
 - Reduced 'noise' if exported files are commited - (With HTML) 
 links to particular elements in a file keep working across 
   multiple versions, most of the time 
 - (With HTML) links become more descriptive 
 
Cons:
- Inceased chance of ID collisions across files Longer IDs  I 
- think this could make a nice option for export settings. I'm 
- aware of the 9.5 feature freeze, but thought I'd mention the 
- idea to see if there's any interest in it.

If you want to see what I've got happening, see 
https://tecosaur.github.io/emacs-config/config.html#nicer-generated-heading.

Please let me know if this idea seems promising!

Timothy.

Re: [Interest] Determanistic Org IDs

[TEC]

 
Hi All,

Noticed some autoformatting happened on send. This is how the cons 
paragraph /should/ read (with a minor correction as well).

*In last email*
 
> Cons: - Inceased chance of ID collisions across files Longer IDs 
> I  - think this could make a nice option for export settings. 
> I'm  - aware of the 9.5 feature freeze, but thought I'd mention 
> the  - idea to see if there's any interest in it. 

*Should be*

Cons: 
 - Inceased chance of ID collisions across files
 - Longer IDs 
 
 I think this could make a nice option for export settings. I'm  
aware of the 9.4 feature freeze, but thought I'd mention the  idea 
to see if there's any interest in it.


Hope that makes things clearer.

Please do let me know what you think, this would be my second 
contribution, so I still very much feel like I'm finding my feet.

All the best,

Timothy.

Re: [Interest] Determanistic Org IDs

[Alexander Adolf]

Hello Timothy,

TEC <tecosaur@gmail.com> writes:

> [...]
> Pros:
>  - Reduced 'noise' if exported files are commited - (With HTML) 
>  links to particular elements in a file keep working across 
>    multiple versions, most of the time 
>  - (With HTML) links become more descriptive 
>  
> Cons:
> - Inceased chance of ID collisions across files Longer IDs  I 
> - think this could make a nice option for export settings. I'm 
> - aware of the 9.5 feature freeze, but thought I'd mention the 
> - idea to see if there's any interest in it.
>
> If you want to see what I've got happening, see 
> https://tecosaur.github.io/emacs-config/config.html#nicer-generated-heading.
> [...]

Interesting. Just a thought: have you considered computing a SHA-2 hash
over either the totality of the heading and all its children (if you
want the links to change when the content changes), or only the heading
itself (keeping the id invariant for as long as the heading remains
unchanged, but content below it can change)?

This would have the advantage of solving both, your "noise issue" and at
the same time removing the threat of id collisions; but at the cost of
not having "speaking" links.


Hoping to have helped,

  --alexander

Re: [Interest] Determanistic Org IDs

[TEC]

 
Alexander Adolf <alexander.adolf@condition-alpha.com> writes: 
 
> Hello Timothy, 

Hi :)
 
> Interesting. Just a thought: have you considered computing a 
> SHA-2 hash... 
I have not. Simply put, I fail to see any advantage.
 
> This would have the advantage of solving both, your "noise 
> issue" and at the same time removing the threat of id 
> collisions; but at the cost of not having "speaking" links. 

As it stands there is no issue with ID collisions within the same 
file, just in-case you wanted to look across files.

This is ensured in the (while (--any (equal ref (car it)) cache) 
... block.

All the best,

Timothy.