On Nov 21, 2010, at 11:26 AM, Bryan Hoyt | Brush Technology wrote:
Be disappointing to turn away something that works just because of potential problems (though I can understand there might be a need, depending on the issues). I've been really looking forward to some more interactive python stuff. The DBus API was a tantalizing start.
Yes, the DBus API has some interesting potential. However, it was not actually designed for this UI purpose, and does need some major rework to be able to support it. The student adding it in as a Google Summer of Code project had to make certain choices, and focused on getting some base DBus functionality solid and did not have time to look at the higher-level calls that might be needed for certain things in the long run.
What we need to focus on is getting the replacement for the extensions API done. If we have people working on the extension side of things, this can be done much more quickly. The key here, I think, is to focus on getting good interactive python stuff going... but in a manner so that the hooks python use can be used by any other approach to scripting (C++, other scripting, D-Bus, etc). Then all can benefit from that single implementation instead of needing to do a one-off development effort for Python, a separate one-off development effort for Perl, etc.