Ihor Radchenko <yanta...@posteo.net> writes:
> Evgenii Klimov <eugene....@lipklim.org> writes: > >> I see in the comment there the reason for that, but still would prefer >> to use unique IDs in tangled files: the probability to overwrite wrong >> source block under nonunique header overweights a side-effect of >> creating new ID in Org file during tangling. > > Did you actually encounter such situation? > `org-babel-detangle' is far from ideal in general. > For example, detangling noweb references is pretty much guaranteed to > create mess. Yes, I did. I keep my whole emacs configuration in an Org file and tangle it after adjustments at Emacs startup (~org-babel-load-file~). Recently I decided to split the tangled file and learned about detangling. And I have a lot of similar headers since my Org file looks like this: #+begin_example * package 1 ** basic ** custom functions ** custom bindings * package 2 ** basic ** custom functions ** custom bindings #+end_example >> Can we do something with this? Add another option or provide an even >> eager value in =org-id-link-to-org-use-id=. > > I do not like the idea of creating side effects in the original Org > document during tangling. > We can, however, use (org-id-link-to-org-use-id 'use-existing) to use > the already present IDs. Understandable, with (org-id-link-to-org-use-id 'use-existing) I would at least be able to manually add them where needed. > Ideally, we should improve the detangling to understand the order of > blocks better instead of searching the containing heading independently. Maybe recent feature suggestion [1] will be of use here. [1] https://list.orgmode.org/118435e8-0b20-46fd-af6a-88de8e19f...@app.fastmail.com
