Try not to calibrate 'list traffic' to 'work'. :-)
Seriously, I'd agree that there is a non-trivial amount of *discussion*
going into this feature, but the amount of work is pretty darn small.
I agree with you on the user base issue. From a marketing perspective,
I could probably think of a less strategic feature than 'vi keybindings',
but I'd have to use my imagination. Your estimate of "less than 10%"
defines a potential user base size which includes zero at the lower
end, so I'll have to agree with it. However, the upper end of your
range probably is a wild overstatement of the potential impact of this
feature. I measure the eventual user base of AbiWord in the tens of
millions of users, and at that scale, the number of people who use
vi keybindings will be far, far less than 1%.
However, I'm glad to see the feature coded anyway. It's not really
consuming very much actual project time. The presence of vi
keybindings is not going to damage the product in any way.
Finally, the feature just has a lot of geek appeal. Anything which
attracts more attention and involvement to AbiWord is good.
We coded a key binding mechanism because we wanted the mechanism
for other reasons. After we had the mechanism itself,
we joked about having vi keybindings, but we were never serious.
Then, in a sobering thought, we realized that since we were
committed to making AbiWord an open, community-developed,
Bazaar-like project, someone might actually come along and code
the feature.
And they did.
Vi keybindings are to AbiWord as <blink> is to Netscape Navigator.
The app's user base doesn't really *need* the feature, but the
app itself needs it, because it would have no soul without it.
-- Eric W. Sink, Software Craftsman eric@abisource.com