From: John Levon (email@example.com)
Date: Tue Jun 03 2003 - 21:05:40 EDT
On Tue, Jun 03, 2003 at 04:36:00PM -0700, Dom Lachowicz wrote:
> > http://mpt.phrasewise.com/2002/03/11#a146
> All I'll say about MPT is that he's
I wish people would stop implying he invented all this stuff :)
> save button (which you still have to do at some point
> anyways if you want to persist your document).
Why ? It's the computer's job to do that (or it should be).
> also not clear to me that the desired behavior should
> happen at the application level
Certainly any implementation would be best done at a "system" level
(inside the Gnome APIs somewhere).
> Often, saving is fundamentally equivalent to putting a
> "seal of approval" on a particular revision of the
> document. Once you write a letter, it stays written,
> right? Unless you cumple up the letter and toss it in
> the garbage - the equivalent of hitting the 'X' button
> on the window bar. Otherwise it was just 'scratch
And guess what - if I leave that scratch paper on my desk, it stays
where it is when I leave the desk. It doesn't just vanish into the
> Saving the file is fundamentally no different
> than putting it inside of a file cabinet, folder,
Correct. However, with most applications, when you *don't* save, you
lose it. Your example above illustrates the difference quite clearly: I
have to make a positive action to get rid of a document. Not saving is
the absence of an operation, not an operation. I disagree strongly that
closing the window is equivalent in most user's minds to shredding a
> This is also the de-facto behavior for any similar
> piece of software, such as MSWord or WordPerfect.
> That's not to say that we can't strive to do better,
> but it will be counter-intuitive and "non-obvious" to
> a user of MSWord how to save a document in AbiWord.
> This is a bad thing.
It would be a bad thing if you were to remove the Save function. So
don't do that.
> Finally, this sort of behavior should be well defined
> and implemented across the desktop (or better, OS and
> across OSes) as a whole.
This archive was generated by hypermail 2.1.4 : Tue Jun 03 2003 - 21:21:11 EDT