Sorry for having lit this thread.

Anyway, I am not going the change the defaults for saving and reverting
file-visiting buffers.  It might be unemacsy, but I think the current
behavior does make sense from a Git perspective.  It's also what Magit
has been doing for a long time, so changing the behavior would break the
expectations of many users.  Magit is an *Emacs* mode for *Git*, and
both perspectives implied by that description are equally important.

- Many users, when they press "g" in the status buffer, expect that that
  buffer then becomes completely up-to-date, especially with regards to
  showing all pending changes not have not made it into a commit.
  That's only possible if the files are saved.

- Many users, additionally have grown accustomed to not even having to
  press "g" for the status buffer to be up-to-date most of the time.
  Which is why the automatic saving cannot be restricted to just the
  `magit-refresh' command, but has to happen must more frequently.

- Likewise during a rebase it would be very inconvenient, if Magit did
  not automatically revert the buffers.  For example, one can directly
  jump to a conflicting in the file-visiting buffer, from the status
  buffer; but of course that only works if the conflict actually is in
  the buffer, which requires that it has been reverted to include the
  conflict markers inserted by Git.

In many other cases such automatic saving and reverting is also
desirable, but because not everyone agrees with these things happening
automatically (at least not every time Git is run for side-effects or to
refresh a Magit buffer) there are options that allow disabling this
behavior.

You are arguing that the features are fine, but should not be enabled by
default, one argument being, that it breaks the expectations formed in
long years of Emacs usage.  And I disagree because Magit has been around
for a few years too now, and its users also have grown accustomed to its
way of doing things.

Given that I strongly believe that reverting/saving is the right thing
to do, and that I have reason to believe that users agree, and that even
more users have grown accustomed to it (with or without having formed an
opinion on it); but (in the case of reverting) also agree that there is
an (IMO small) risk of data-loss; doing these things verbosely, is the
only available compromise.

-- 
You received this message because you are subscribed to the Google Groups 
"magit" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to [email protected].
For more options, visit https://groups.google.com/d/optout.

Reply via email to