Release 9.6

Release 9.6

[Bastien]

Hi all,

Org 9.6 a major release, is out, and will be available from GNU ELPA
in a few hours.

The release notes are published on https://orgmode.org/Changes.html.

On top of the many small and big enhancements against 9.5, here are
some changes you might want to be aware of:

- Org tests are now regularily run:
  https://git.sr.ht/~bzg/org-mode-tests/

  Thanks to Christian Köstlin for helping us setting this!  We are
  looking for someone to maintain this repository.

- Because Org is more than Emacs org-mode, and in order to help people
  working on Org parsers/tools, we made an effort to expose the syntax
  for .org files: https://orgmode.org/worg/org-syntax.html

  Thanks a lot to Tim (TEC) for his work on this!

- Worg is a good place to advertize useful Org packages: if you want
  write access to Worg, please create an account on https://sr.ht and
  send me your username.  (Patches can be sent with no account.)

  To help track orphan Org projects, there is now a dedicated page:
  https://orgmode.org/worg/org-orphanage.html

  Let's work from there to provide more useful information on the Org
  ecosystem in general.

- If you enjoy using Org, please consider supporting contributors via
  https://liberapay.com/org-mode/

  If you contribute to the Org ecosystem in any way (not just code),
  you are welcome to join the list of people who share donations, do
  let us know (don't be shy).

Thanks to *everyone* who continues to make this list a nice and useful
place for Orgers of the World.  Last but not least: thanks to Ihor his
truly amazing work and for being the de facto maintainer.

Enjoy!

-- 
 Bastien

Re: Release 9.6

[Max Nikulin]

On 29/11/2022 13:58, Bastien wrote:
> 
> Last but not least: thanks to Ihor his
> truly amazing work and for being the de facto maintainer.

I think, Ihor's role in this release is crucial. He spent a lot of time 
fixing bugs and reviewing patches, not to mention the org-fold framework 
to overcome performance limitation of overlays (the latter has been 
recently addressed in Emacs though).

Re: Release 9.6

[Thomas S. Dye]

Max Nikulin <manikulin@gmail.com> writes:

> On 29/11/2022 13:58, Bastien wrote:
>> Last but not least: thanks to Ihor his
>> truly amazing work and for being the de facto maintainer.
>
> I think, Ihor's role in this release is crucial. He spent a lot 
> of time fixing
> bugs and reviewing patches, not to mention the org-fold 
> framework to overcome
> performance limitation of overlays (the latter has been recently 
> addressed in
> Emacs though).

+1

-- 
Thomas S. Dye
https://tsdye.online/tsdye

Re: Release 9.6

[Tim Cross]

Max Nikulin <manikulin@gmail.com> writes:

> On 29/11/2022 13:58, Bastien wrote:
>> Last but not least: thanks to Ihor his
>> truly amazing work and for being the de facto maintainer.
>
> I think, Ihor's role in this release is crucial. He spent a lot of time fixing bugs and
> reviewing patches, not to mention the org-fold framework to overcome performance
> limitation of overlays (the latter has been recently addressed in Emacs though).

I agree. The amount of time and effort Ihor has put into org mode is
both crucial and hugely appreciated.

One question I do have is with respect to the new folding code, the
changes in Emacs to improve overlay performance and the correct way
forward.

On one hand, it was an immense amount of work for Ihor to implement a
better performing solution and something I'm sure those with large org
files will appreciate. However, on the other hand, I guess it also puts
a greater burden from a maintenance perspective on org maintainers as
org is now using its own unique approach to hiding/showing content
(folding).

Has anyone done any comparisons between the new overlay implementation
in Emacs 29 and the new folding approach in org 9.6? Is there still
sufficient performance benefit with org's approach over using Emacs
overlays or do we need to seriously consider changing the default back
to native Emacs overlays?

Re: Release 9.6

[Charles Millar]

On 12/1/22 18:10, Tim Cross wrote:
> 
> Max Nikulin <manikulin@gmail.com> writes:
> 
>> On 29/11/2022 13:58, Bastien wrote:
>>> Last but not least: thanks to Ihor his
>>> truly amazing work and for being the de facto maintainer.
>>
>> I think, Ihor's role in this release is crucial. He spent a lot of time fixing bugs and
>> reviewing patches, not to mention the org-fold framework to overcome performance
>> limitation of overlays (the latter has been recently addressed in Emacs though).
> 
> I agree. The amount of time and effort Ihor has put into org mode is
> both crucial and hugely appreciated.
> 
> One question I do have is with respect to the new folding code, the
> changes in Emacs to improve overlay performance and the correct way
> forward.
> 
> On one hand, it was an immense amount of work for Ihor to implement a
> better performing solution and something I'm sure those with large org
> files will appreciate. However, on the other hand, I guess it also puts
> a greater burden from a maintenance perspective on org maintainers as
> org is now using its own unique approach to hiding/showing content
> (folding).
> 
> Has anyone done any comparisons between the new overlay implementation
> in Emacs 29 and the new folding approach in org 9.6? Is there still
> sufficient performance benefit with org's approach over using Emacs
> overlays or do we need to seriously consider changing the default back
> to native Emacs overlays?
> 
Is this the cause for the behavior I described a few days ago in

Bug org-cycke fails after "replace-string"

Charlie Millar

P.S. Sorry for the typo

Re: Release 9.6

[Ihor Radchenko]

Tim Cross <theophilusx@gmail.com> writes:

> One question I do have is with respect to the new folding code, the
> changes in Emacs to improve overlay performance and the correct way
> forward.
>
> On one hand, it was an immense amount of work for Ihor to implement a
> better performing solution and something I'm sure those with large org
> files will appreciate. However, on the other hand, I guess it also puts
> a greater burden from a maintenance perspective on org maintainers as
> org is now using its own unique approach to hiding/showing content
> (folding).

Note that org-fold text property implementation will remain relevant
until we drop the support of Emacs 28.

> Has anyone done any comparisons between the new overlay implementation
> in Emacs 29 and the new folding approach in org 9.6? Is there still
> sufficient performance benefit with org's approach over using Emacs
> overlays or do we need to seriously consider changing the default back
> to native Emacs overlays?

I did.
https://orgmode.org/list/87359z8ssb.fsf@163.com

    Emacs master:
     - main Org branch using text properties for folding: Elapsed time: 2.185786s (0.639190s in 4 GCs)
     - main Org branch using overlays for folding: Elapsed time: 27.244284s (0.731581s in 5 GCs)
    
    Emacs feature/noverlay:
     - main Org branch using text properties for folding: Elapsed time: 1.586936s (0.476606s in 3 GCs)
     - main Org branch using overlays for folding: Elapsed time: 2.039803s (0.724576s in 5 GCs)
    
    The improvement is very significant. GCs are mostly caused by Org
    internals staff as you can see by comparing built-in and main Org
    versions. Without GCs, overlays are on par with text properties.

overlays vs. text properties are now mostly on par modulo some more GCs
when using overlays.

In terms of maintenance, it is not that bad. org-fold supports both
overlays and text properties. I just did not implement certain features
when overlays are used.

Eventually, we can drop text property support and go back to overlays.
We do not have to abandon org-fold for that. org-fold is mostly an API
for folding. It provides more than just a replacement of overlays. Other
places in Org code now use API calls instead of directly working with
overlays.

-- 
Ihor Radchenko // yantar92,
Org mode contributor,
Learn more about Org mode at <https://orgmode.org/>.
Support Org development at <https://liberapay.com/org-mode>,
or support my work at <https://liberapay.com/yantar92>