Re: How to make Emacs popular again.

[Eli Zaretskii]

> From: Richard Stallman <rms@gnu.org>
> Cc: philipk@posteo.net, eduardoochs@gmail.com, spacibba@aol.com,
>       emacs-devel@gnu.org, dgutov@yandex.ru, jamtlu@gmail.com
> Date: Sat, 03 Oct 2020 23:38:51 -0400
> 
> There are other reasons why it is bad to adopt "talk to a server" as
> your default way to do something, and for bad the community to adopt
> "talk to a server" as its default recommendation and practice.
> 
> Getting an answer over the network can take time.
> 
> It can cost you money, in some circumstances.
> 
> It can fail, if the network is having a problem.
> 
> It can result in tracking your movements, if it requires you to
> activate a network connection that identifies and tracks you.
> 
> It can be impossible, if you don't have a connection
> at that time or place.

I think we should let the user decide whether such risks are relevant
in each and every case, and whether these risks, if they exist, are
worth taking.  We explain these issues on the FSF site, so the
considerations and the risks are well known.  We can explain this
again in our documentation where relevant.  This way, we can consider
users informed and capable of making their own decisions.

That said, I agree that testing for the availability of a local server
first is a good idea that dictionary.el should implement, so I'm not
really sure what are we still arguing about.