Bucket inconsistently copies trigger formatting

[dgw]

This is kind of a follow-on to #87, but should be discussed separately, as it is not a bug per se.

Bucket has some interesting behavior when the trigger line is formatted, and the resulting factoid uses any verb other than <action> or <reply>. If the whole line is bold (or italic, underlined, or strikethrough; i.e. the line starts with a control character that has no terminating counterpart), Bucket outputs the entire factoid with that formatting. If only part of the line is formatted, Bucket outputs the trigger portion verbatim, but the rest of the factoid is unaffected (because the formatting is terminated properly). Presumably, these quirks will also apply to colors when #87 is fixed.

The thing is, I don't see why Bucket should regurgitate formatting at all. At the very least, I think it should be consistent—either all factoid types (including <action>s and <reply>s) should adopt the trigger's formatting in the same way, or no factoid types should adopt the trigger's formatting.

I'm partial to ignoring the trigger's formatting, but I'm mostly looking for thoughts on which direction it would be better to go in. Maybe there's an alternative that I haven't thought of. My only strong opinion here is that the current behavior is frustratingly (to me) inconsistent and should be changed.

[zigdon]

I think I'd be for stripping control chars everywhere. If we want to be
fancy, we can limit that stripping from the trigger side, and still allow
it on the tidbit side.

On Sun, Oct 2, 2016 at 1:30 AM dgw [email protected] wrote:

This is kind of a follow-on to #87
#87, but should be
discussed separately, as it is not a bug per se.

Bucket has some interesting behavior when the trigger line is formatted,
and the resulting factoid uses any verb other than or .
If the whole line is bold (or italic, underlined, or strikethrough; i.e.
the line starts with a control character that has no terminating
counterpart), Bucket outputs the entire factoid with that formatting. If
only part of the line is formatted, Bucket outputs the trigger portion
verbatim, but the rest of the factoid is unaffected (because the formatting
is terminated properly). Presumably, these quirks will also apply to colors
when #87 #87 is fixed.

The thing is, I don't see why Bucket should regurgitate formatting at all.
At the very least, I think it should be consistent—either all factoid types
(including s and s) should adopt the trigger's formatting
in the same way, or no factoid types should adopt the trigger's formatting.

I'm partial to ignoring the trigger's formatting, but I'm mostly looking
for thoughts on which direction it would be better to go in. Maybe there's
an alternative that I haven't thought of. My only strong opinion here is
that the current behavior is frustratingly (to me) inconsistent and should
be changed.

—
You are receiving this because you are subscribed to this thread.
Reply to this email directly, view it on GitHub
#88, or mute the thread
https://github.com/notifications/unsubscribe-auth/AACtVIaJrQ_t6P8vFXnXdckcKY035ytsks5qv2uOgaJpZM4KL9qA
.