Re: [Gnumed-bugs] Bug (?) trying to sign multiple results?

[Karsten Hilbert]

On Tue, Oct 04, 2011 at 09:15:46AM -0700, Jim Busser wrote:

> >> In 0.9.9 ...
> >> 
> >> ... if I select a cell in which reside three (multiple)
> >> results, then even when I have selected all three, the
> >> middle one is unactioned (unsigned) despite that all had
> >> been selected before I reached, and checkboxed "sign"
> >> within, the subsequent signing dialog.
> > 
> > In this case something maybe not quite expected but
> > certainly very logical happens: One is applying "doing
> > nothing" to the previously selected results.
> 
> Well then we should understand an improvement to be
> desirable here, because the user is here required to take
> multiple steps for something to happen when a single step
> should be possible.
> 
> I am ok with the concept of having to select something
> like assigning normalcy or abnormality or entering a comment
> etc but it would not then make sense to require to checkbox
> that you are signing it, because that should be understood
> i.e. 'implicit' should here be acceptable).

This whole confusion stems from the fact that someone
introduced the term "sign" (over simply review) into the GUI
labels.

GNUmed simply does NOT OFFER any way to *sign* a result
without reviewing it. Of course, going back to front, having
reviewed a result does imply that it is signed off. But this
does NOT imply that a result can be signed without a review.

On paper this may work - results are expected to be flagged
*already* and thus can simply be signed off by -- gasp -- a
second (!) reviewer.

In that sense it is only logical that applying "nothing"
will not "sign" -- because simply nothing exists to document
signage without review.

Quite logically, the term "sign" must come under scrutiny:

        - review
        - evaluate
        - flag
        - ...

?

Karsten
-- 
GPG key ID E4071346 @ gpg-keyserver.de
E167 67FD A291 2BEA 73BD  4537 78B9 A9F9 E407 1346