staikos at kde.org
Fri Aug 8 22:21:33 CEST 2003
On Friday 08 August 2003 16:08, C. Barth Netterfield wrote:
> Just curious:
> What are the advantages of map access vs list access in this case?
It was what we had talked about before regarding accessing the input and
output vectors and scalars in a more programmer-friendly way. The plugin
code was particularily messy. I had to come up with a name+order -> number
mapping and keep track of that internally. Now we can access the scalars and
vectors for data objects in a much more friendly manner:
The name there is the "programmer name", not the vector name, though there
is nothing stopping you from using the vector tagname as the key. In the
plugin code, I explicitly have to use the XML tag in order for it to be
It seems to be cleaning up the plugin code very nicely now anyways. I'm a
good portion through the vector code changeover, and I'll rerun the full
regression + some adhoc tests before it goes in.
After that I can more easily finish the delayed input loading and fix some
UI quirks of the plugin editor.
Using a list for this was an unfortunate mistake when I designed
KstDataObject, but this seems to fix all the problems I discovered.
KDE Developer http://www.kde.org/
Staikos Computing Services Inc. http://www.staikos.net/
More information about the Kst