I'm glad about the amount of traffic on the mailing list. Such discussions are either the end of a project or the beginning of a new process. Let's hope it's the first step towards the next PySide generation. The times of describing a problem and waiting for the solution of the development team are definitely over, so if we want this project to continue, we have to help ourselves. I think, before deciding implementation details, we have to stand one step back and answer the following questions one after the other:
- What are our goals for PySide? - Which development team would be available? - How much money is needed/available? With this information, we can decide which of the defined goals are realistic to implement... To get, document and archive this information, I created a little survey. Please do not answer the questions now already, because I would like to get the feedback of some of you about missing/superfuous questions. Could some of you visit the following link and just click "Continue" until the end, read everything, but not submit or answer anything and give me some feedback: https://docs.google.com/spreadsheet/viewform?fromEmail=true&formkey=dHVNa1d3dVpvdFdJV1U3THBxVVk2Tnc6MQ Thanks. Then we can start the official survey tomorrow or so. Cheers Aaron Am 14.01.2013 19:16, schrieb John Ehresman: > On 1/14/13 12:18 PM, Fabien Castan wrote: >> I'm not enthusiastic about a rewrite using swig; it seems to be a lot of >> work for questionable benefits. >> >> The main benefit could be to get a bigger community and concentrate >> efforts on the binding rules, instead of working on a binding tool. > I don't think the binding tool as needing a lot of effort. We do need > more people fixing bugs and improving the binding rules, but that is the > case with either tool. If we were starting from scratch, I'd think swig > would be something to look at, but we aren't starting from scratch. > >> I think much of the work with PySide is >> writing a Python binding given the specifics of how Qt works so it's >> less about using a semi-generic tool such as swiq or shiboken and more >> about how Qt object lifetime works. >> >> Yes, but users also need to bind their own widgets... And your widgets >> use your core objects... so you need to use the same binding >> tool everywhere. >> A generic binding tool could help for that. > You are correct that users need to either use one binding tool for all > qt related interfaces or do extra work to use something else. This > would be true with swig or shiboken or anything else. > > Cheers, > > John > _______________________________________________ > PySide mailing list > PySide@qt-project.org > http://lists.qt-project.org/mailman/listinfo/pyside _______________________________________________ PySide mailing list PySide@qt-project.org http://lists.qt-project.org/mailman/listinfo/pyside
