Re: Question collaborative editing.

[Eli Zaretskii]

> From: Qiantan Hong <qhong@mit.edu>
> CC: Ergus <spacibba@aol.com>, Fermin <fmfs@posteo.net>,
>         Jean Louis
>       <bugs@gnu.support>, Noam Postavsky <npostavs@gmail.com>,
>         Emacs developers
>       <emacs-devel@gnu.org>,
>         Karl Fogel <kfogel@red-bean.com>,
>         Stefan Monnier
>       <monnier@iro.umontreal.ca>
> Date: Wed, 30 Sep 2020 17:04:09 +0000
> 
> >> You don't need buffer-modification hooks, you can use other existing
> >> mechanisms.  Emacs knows that buffer text was changed even if the
> >> modification hooks were inhibited.
> > That’s a workaround, but I’m not sure if it can capture user intent
> > (i.e. it gives operation-wise information rather than just a diff between
> > versions of buffer). If it can’t then we will need to implement both
> > update based on modification-hooks and change state of the buffer,
> > which is an additional impl/maintenance burden.
> I see the difference as, if implemented as a C library that keep a buffer
> and synchronize with emacs, then this synchronization *must* be
> absolutely correct, otherwise it mess up any operation on the buffer.
> However if the CRDT is kept on the emacs buffer itself, it’s much
> more tolerant to untracked text in the buffer.

I'm sorry, I probably miss too much background information here,
because I don't really understand what is the issue you are describing
here.  E.g., why is user intent important, when all you need is to
keep buffer text synchronized between several sessions?