11.09.2011 05:31, alan@??? пишет:
> That raised several issues with gtkme. An example is that gtkme made
> python segfault while using a GtkBuilder file containing an HBox. I had
> to workaround this issue using a GtkTable. But that's not the key
> question.
Seems odd indeed.
> gtkme seems buggy and undocumented. I have no idea which problem it's
> supposed to solve, but I'm under the impression that keep maintaining
> a tails-greeter that is using it will be a pain, while some of us have
> skills in normal pygtk programming. So I'm wondering wether we should
> try to get rid of it or not.
For the sake of reference: gtkme's project page is here -
https://code.launchpad.net/~doctormo/doctormo-random/gtkme
> Max, do you know how strongly gtkme is tighten with tails-greeter code?
Essentially gtkme is kind of a wrapper which tries to abstract pygtk boilerplate.
It shouldn't be too hard to move that boilerplate code into tails-greeter.
> Max and Tails developpers, do you think it's worth putting energy in
> porting tails-greeter to plain gtk or it's better to try to maintain it
> using gtkme?
I'm not sure. Boilerplate abstraction is a good idea but I haven't looked into gtkme
code to judge on the quality and potential maintenance burden.
I'll try to take a look in upcoming weeks.
cheers,
Max.
P. S> add gtkme's author to CC :)