Re: [Gnumed-devel] Re: [GPCG_TALK] GPs split on software ads

[Tim Churches]

Karsten Hilbert wrote:
> On Sat, Mar 12, 2005 at 09:38:08AM +1100, Tim Churches wrote:
> 
> 
>>Horst wrote:
>>
>>>If you check the available code, you'll see that gnumed doesn't really need
>>>that much work any more to become fully functional.
> 
> functional != usable by Joe User, MD.

I agree. That's why we have indicated to Horst that we really need to
see his GnuMed-mini to assess its value-proposition versus funding work
on Gnumed-proper (or alternatives such as Vista Office, OSCAR etc).

>>I have been monitoring the GnuMed email list for the last 18 months, and
>> the level of activity on GnuMed development is quite amazing
> 
> ...
> 
>>(If anything, I feel that GnuMed is in danger
>>of becoming over-designed and over-engineered, and that the absolutely
>>excellent has become the enemy of the good enough, but that's just my
>>opinion).
> 
> This is perhaps correct.
> 
> 
>>However, there is still a HUGE amount of work to be done before GnuMed
>>is ready for production use.
> 
> ...
> 
>>... because moving from
>>Version 0.x to version 1.0 involves enournous attention to zillions of
>>details ...
> 
> 
> Tim, I always much enjoy reading your sobering messages. I
> think you are very realistic about a lot of things. I am
> really happy to have you monitor us and give advice (eg reign
> in the madness ;-)

Thanks Karsten. I was sincere in what I said about the quality of the
GnuMed work. However, I am an obsessive doctor too, and looking at the
GnuMed project I recognise the same (Quixotic) quest for perfection and
generality as I suffer from in my own projects. However, I am perhaps
fortunate to be limited by deadlines and a fixed budget, as well as
working with a truly excellent analysts/programmer who constantly points
out "If we add feature A then there is no way we will have time to add
B, C or D". On the other hand, I have to stop him when he proposes "If
we just refactor module X, then everying else will be so much
easier/more elegant/robust." - when module X has already been reworked
several times and is Good Enough.

I think it is a matter of finding the "sweet spot" between having enough
time to do a good job, and having a fixed deadline and some defined
deliverables in order to drive the project towards some significant
milestone.

Tim C