Re: [Gnumed-bugs] GNUmed does not allow same-day substance discontinuations

[Busser, Jim]

On 2013-07-11, at 2:33 AM, Karsten Hilbert <address@hidden> wrote:

> ...
> 
> Citing from the CHANGELOG:
> 
>       1.3.2
> 
>       FIX: exception on substance-discontinued in the future [thanks 
> S.Hilbert]

what caused you to decide to look in the above changelog? Was it the result of 
your checking back across the changelogs, to visually identify (or to identify 
on the basis of the word 'substance' whatever was the most recent manipulation 
of substance-related code? Or is it because you had a relatively clear 
recollection of having then done some patching of discontinuation-related code?


> 
> The corresponding patch:
> ...
> @@ -1493,17 +1493,21 @@ class 
> cSubstanceIntakeEAPnl(wxgCurrentMedicationEAPnl.wxgCurrentMedicationEAPnl,
>                       else:
>                               self._PRW_duration.display_as_valid(True)
> 
> ...
> The code thus reads:
> 
>       # end must be "< now()" AND "> start" if at all
>       end = self._DP_discontinued.GetData()
>       if end is not None:
>               # not in the future:
>               if end > gmDateTime.pydt_now_here():
>                       self._DP_discontinued.display_as_valid(False)
>                       validity = False
>               else:
>                       # and not before it started
>                       if start > end:
>                               self._DP_started.display_as_valid(False)
>                               self._DP_discontinued.display_as_valid(False)
>                               validity = False
>                       else:
>                               self._DP_started.display_as_valid(True)
>                               self._DP_discontinued.display_as_valid(True)
> 
> So, this really should work. Going checking... OK, I can
> reproduce it. So let's instrument the code a bit... GNUmed
> seems to think that end is > than now(). Ah, there's the
> rub. The started PRW is pre-filled with now()

In the code, what is it that you have chosen or designated the expressions 
(abbreviations)

        PRW
and
        DP

to represent?


> The user chooses (or types the
> date for) today. This will be without the time. GNUmed
> assumes 11:11:11.111am. ….

Can you identify for us whether GNUmed ever saves a GUI-displayed date with 
"time precision" even when the time was not displayed?

Also, I find it cumbersome to enter date-times (other than now) into GNUmed, in 
as much as

- the time of the interaction with GNUmed is often not real-time, such that I 
would prefer the time entity to be "rounded off" or "truncated" (for example, 
when a patient informs me that they stopped their medication late last week, 
there is uncertainty whether it was Thursday or even Saturday

- are there therefore easier ways to enter a date, such as anything like

        today at 10:15 am --> 10:15
        two days ago at noon --> -2d 12
        tomorrow at 1700h --> 1d 1700 (or +1d 17:00)

??

-- Jim