Niels Sascha Reedijk wrote: > On Fri, Feb 24, 2012 at 3:55 PM, Ingo Weinhold <ingo_weinhold@xxxxxx> wrote: > > Jürgen Wall wrote: > >> Just to throw in an idea, wouldn't it make sense to have someone analyze > >> the requirements and efforts of such an API transition in a CSoC task? > >> This task would enclose analysis of the Haiku API and the QT API > >> delivering a rough plan how to map/adapt/replace parts. > >> It'd be kind of an applicability research task. > > > > AFAIK such a task wouldn't be allowed as a GSoC project. While a project > > may include complementary tasks like requirements analysis, documentation, > > testing, etc. the main task is supposed to be coding. A practical approach > > would be fine, though, like: > > > > * Complete the Qt port (disclaimer: I don't have any idea what's the > > current state). > > * Add extensions to/on top of the Qt API to provide Haiku specific > > functionality for attributes, resources, inter-app communication, > > translators, etc. > > * As a proof of concept, port StyledEdit to the extended Qt API. > > Wouldn't it make the whole project more compelling it step 2 would be > to write a Haiku backend for Qt that uses the libbe <-> app_server API > to interact with the app_server? So a deeper integration as it were... While that would be a useful optimization, in the short term I'd be much more interested in the feasibility aspect. Like how well the Haiku specific extensions can be made, how much work porting a Haiku application to the Qt API is and how well the result works, etc. I think it will already be plenty of work to get results in that respect. CU, Ingo