Krita shared pointers usage

[Boudewijn Rempt]

On Sat, 20 Dec 2014, Stefano Bonicatti wrote:
> 
> I do know what you're talking about and still believe that is somewhat wrong to come to a point where you have a so large codebase with the architecture
> undocumented.. at least on an high level.

Oh sure, and we've tried to start documentation a lot of times. The big 
problem is that architecture changes, and often quite quickly, and then 
the documentation becomes outdated. Of course, that's not peculiar to 
Krita, I've _never_ seen an adequately documented project.

It all tends to heap up, change after change. In our case, there are 
several places where wehave layers of design heapd on tpo of each other, 
every developer forgetting to check what has happened before.

And then, there's the problem that most developers are good at coming up 
with designs but extremely bad at understanding other people;s work, 
leading to a tendency to excessive refactoring. The "I had an idea how it 
should have been done, now I'm looking a the code and it isn't done the 
way I imagined it would be done, so it must be redone."

But the upshot is that it is quite probably not logically possible to have 
good documentation for Krita's code and  architecture outside of the code 
itself.

But if you want to give it a try, you're welcome :-)

Boud