[referencer] usability regressions

[Henry Gomersall]

Having used Referencer for some time now, I have noticed a couple of
usability regressions appear in recent versions.

The first is the default edit mode, where one can edit metadata in the
list view. This disallows, what I would argue, is the more common usage
of using Referencer as a means of organising commonly accessed papers.
Ie, edit once, view many. For this reason the emphasis should be placed
on easy opening, not on easy editing. Triple click is not a discoverable
way of opening documents (i stumbled upon it accidentally!). Allowing
editing in list view is also inconsistent with disallowing it in icon
view. Further to this, the open action should follow the desktop default
if possible; e.g. look up the /apps/nautilus/preferences/click_policy
key to get whether single click opening is required. This is a good
example of where single click is simply unused at the moment (which is
crazy, as its the most basic mouse entry method!).

The second regression is the loss of the tag check-box pane. Now adding
tags to an entry in list view requires 2 extra clicks per tag (and
discoverability is broken). The power of Referencer is in its tags and
so hiding the tag attachment is a loss. Perhaps this could be somehow
combined with the visible tag list (in F-Spot, tags can be dragged onto
the item to be tagged). The little mouse-over icons in icon view are
neat (but I think should not disappear after selecting a tag - allowing
multiple tagging at once).

Finally, instant apply would be a very useful feature. I have often
added papers and exited only to open Referencer again and realised they
are no longer present in my list.

I would be happy to get involved in development efforts.

Other than the above gripes, I think referencer is a great piece of
software!

Thanks,

Henry