Re: [Ring] redaction feature

[Julian Foad]

bill-auger wrote:
> [...] a terrible feature that discourages mindful communication and
> facilitates evasion of the responsibility for one's words
> [...] think carefully before you press "send" or simply stop
> worrying [...]

I feel embarrassed when I make a mistake in front of all my correspondents.

I want to correct my mistake.

I have plenty of ways I can "speak" my correction, certainly:
   "I meant 'Bob', of course."
   "Sorry, that should be 'Bob'."
   "s/boob/Bob/"
   "*Bob"

This is a modern computer system so why on earth can't it let me make my 
correction directly, like in a WYSIWYG word-processor? Why should I have 
to write another message in which I indirectly, fuzzily, refer to a 
mistake which my correspondents have to interpret as probably being in 
the last or one of the few last messages I wrote, or perhaps I was 
pointing out a mistake in someone else's recent message, when the 
computer could perfectly well enable me to point precisely at the place 
I mean and display a clear indication of what I want to change it to? 
Like when "show changes" is enabled on a word processor, and the 
computer shows the old text with strike-through and the new text with 
underlining and/or different colours.

When I "speak" my correction, the other participants can judge whether I 
am behaving inappropriately or trying to pretend I didn't really mean 
something I said before. Similarly, if I use a computer-assisted 
correction mechanism, it should also enable the participants to see so 
they can judge me, and the social pressure of the group is what keeps 
the participants honest.

Here is a real example of the "honesty" principle. I use and participate 
in the development of the Subversion version-control system, and one of 
its features is you can edit the log message of an old committed change. 
One could potentially misuse this to pretend that the change was made 
for a different reason, or by a different person, or had been tested 
when in fact it had not been tested. But the social conventions of a 
team mean that the feature is in practice not abused -- because if 
someone abused it then they would not be welcome to stay in the team.

What I am trying to say is don't underestimate the importance of giving 
users a good way to achieve their basic need to be seen as a pleasant or 
careful correspondent, or whatever particular characteristic is 
important in their situation.

This doesn't mean the system should allow arbitrary un-tracked editing 
of any and all messages, certainly not. A system that pretends it has 
absolutely finally redacted or replaced a message is lying, in 
principle, so that is a bad user interface, and just annotating the new 
version with "(edited)" is hardly any better if it does not let you see 
the old version. But then going the whole way the other way and showing 
a whole (branched?) version history is ridiculous too.

But there are some lovely shades of grey in between black and white.

- Julian