[Top][All Lists]

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Gnumed-devel] the id_name debate

From: Richard Terry
Subject: Re: [Gnumed-devel] the id_name debate
Date: Fri, 3 Sep 2004 18:55:10 +1000
User-agent: KMail/1.5.4

On Fri, 3 Sep 2004 06:10 pm, Karsten Hilbert wrote:
> > I remember now why I got the shits with the names just being ID:
> >
> > It is because, when you are using say a visual designer where you can
> > drag and drop the joins (like pgaccess) then you can't tell what foreign
> > key is likely (meant to) be linked to what table!
> Ah, that of course makes your rationale clear.
> However, a visual database designer that doesn't make it
> painfully clear which objects one is operating on is IMO

They do - see the png attatched - it is the IMHO crappy naming scheme which 
makes it painfully unclear!

I don't expect you to relent on this but the attitude is typical of what makes 
linux a pain to use in many situations, and windoze a breeze.

Take a look at the notablehead.png (I've wiped out the table headings)
As you are looking at it you've not much clue as to which ID links to which 
have you (of course not there are not headings), but you would immediately if 
the primary key was named id_address, id_street etc. You wouldn't need the 

Now, look at the second png with the headings. Of course, you can easily 
identify which id in which table links to which names external key in which 
know which id links to which you have to look up to the top of the table 
heading and them back down to where you want to link to - This is slow, 
confusing and a total pain - you wouldnt have to do this if they were 
properly named.

This extends to every bit of design/gui design. The current gnuMed gui design 
is a total pain as well - because whovever stuck in the tabbed system at the 
bottom just dosn't realise that. I've always made my designs so one needs as 
minimal as possible eye movement to get around. This is the same sort of 
argument I put forth when I modified the login screen some time ago (and it 
was rejected). I showed conclusively that though in that case, I had actually 
added another layer of textboxes up front, that the number of key presses 
needed to go the basics was often much less.  I guess it is a matter of 
philosophy, but I can tell  you from my on-the-ground tutoring of GP's in 
their offices, back in the days in the late 1990's when we had projects 
running for them to get them into computers, they get easily confused by  

Also screen design placements make navigating with a mouse which one often has 
to do easy or hard. IN the current gnumed - one types in a name at the top, 
and then has to mouse travel all the way to the bottom of the screen to 
change tabs, then back to the top etc. 

Anyway - hope this makes some impression and not necessarily a bad one.



> broken in the first place. And, frankly, I don't care one wit
> about broken visual schema design software. By definition such
> software would operate under the suspicion of creating faulty
> schema output.
> Karsten

Attachment: pgacess_visual_query_designer.png
Description: PNG image

Attachment: pgacess_visual_query_notablehead.png
Description: PNG image

reply via email to

[Prev in Thread] Current Thread [Next in Thread]