[Gnumed-devel] Usability ?a matter of taste

[Richard Terry]

Yes and No. Please read all the text before looking at the pngs for they will 
not be self explanatory.

A good analogy is anyone driving an old ford may think its the bees knees. 
After all it gets you from A to B. Give them a porche, and let them compare.

Good software design is both about backend engineering and the ergonomics and 
functionality of the work flow offered by the front-end.

As someone else mentioned in a previous post, we have had 10 years + 
experience in clinical software in this country, and  I have had the 
opportunity over many years, initially in my role as IT manager for the HUDGP 
+ project software writer for divisional projects + implementation in my 
office, to see the huge difference in functionality of different products.

When I wrote my initial script writing software - it was just that - script 
writing. We used it in a dozen practices as part of a project, and I 
continued to support it for several years later until I had to forcibly 
wrench it out of the computers as I no longer wanted to run around and do the 
drug updates etc.

To this day, those users, most of whom use medical director or MedTech (NZ) 
bemoan the fact that they had to stop using it. Fellow GP's who have 
subsequently sat and played with my software can immediately see how good it 
is - mainly because of the speed and method of the work flow.

This is a concept that I think many people on this development team still do 
not understand  - ie that one has to define functionality/user need. It is no 
good person A coming up with what they think will be a good design for 
something if it a) doesn't work in practice and b) dosn't fit into an overall 
design philosophy. Countless hours are then put into something which may not 
be easy to change.

The SOAP editor is yet another example of a good concept which currently is 
being implemented in an unworkable fashion - yet most of you will never ever 
know this, even after you start to use it on your desktop, because you will 
not have been able to experience an integrated alternative. 

Lest you forget, the SOAP editor was written by Ian with my design specs, at 
my request. Carlos or Karsten (Carlos I think) came up with the fabulous idea 
of having multiple SOAP controls in the workspace. Now instead of the group 
running with that idea and optimizing it before its development became 
intrenched - that is were it sat, and is now the preferred production model 
using an essentially unworkable, unergonomic, multi-sash control that will do 
nothing but confuse the user, and should be replaced by the much more elegant 
notebook tab control which is easy to use/add/delete tabs.

I've taken Carlos's code and started substituting a tab control there, just to 
see what it looks like, and I enclose some pngs, but I don't have the ability 
to make it work.

To try and explain why the multi-sash concept becomes unworkable, I've entered 
a few consultations, which are not  untypical of what one sees in general 
practice  - where many many of our consults are not one liners.  I've kept 
the size of the multiple soap at something reasonable given that it will 
probably shouldn't occupy the whole of the screen.  

Note particularly that one very rapidly loses multiple lines of the 
consultation notes from view because of the confines of space. Where the 
scrollbars are, was about 4-5 lines of history notes (probably the most 
important part of the consult) and that it is no longer visible.

Next I've put up a png with Carlos's code + a tab instead of a multi-sash 
(havn't figured out how to pass the SOAP to the tab). Adding a new tab BTW is 
a simple one liner  tab.AddPage(control). Notice in this png, that the design 
includes both the active history list, recent consultation list, and summary 
of the current consultation list. 

Lastly I've cut and pasted the amalgam of the multi-tabbed soap + list of 
recent encounters + active problem list, into wxPython25 design to show how 
much more reable/useable the multi-tabbed SOAP design would be when one can 
allow larger screen area than is possible with multi-sash  to the soap 
control , and also how the whole concepts fits into the large picture of the 
patient record.



Regards

Richard


On Sun, 13 Mar 2005 12:02 pm, Karsten Hilbert wrote:
> On Sun, Mar 13, 2005 at 10:34:13AM +1100, Syan Tan wrote:
> > Q. is usability a matter of taste?
>
> I would think yes.
>
> Karsten

carlos_multisash_vs_tabbed.png
Description: PNG image

carlos_tabbed_incontext.png
Description: PNG image