On Feb 14, 8:00 am, Stefan Seifert <[EMAIL PROTECTED]> wrote: > If I may ask: why even bother with API and - even harder - ABI > compatibility? > > How many plugins from how many different sources are there gonna be? > Will they be so many, that it's impossible to manage them in a single > distribution?
The main idea is to have a plugin repository. The API is different than the current API, making it more usefull and easy to make small and simple plugins. Ideally, there will by Python/Ruby/etc bindings to make it even easier. For this to work, we need a stable ABI, otherwise people will be put off from making plugins and maintaining them. I like the new (not new anymore?) interface to SuperKaramba and how it can download modules and install them from within the application. Something like that is very user friendly, and if possible, it would be nice to see in Licq. > This is open source. There's nothing preventing one from compiling the > plugins with the core on the same system. So it's feasible to just ship > all known plugins with the core and make just the loading optional. This > model even works for the whole Linux kernel with thousands of drivers > written by at least hundreds of different contributors. I can hardly see > licq growing bigger than that. Is it easy for a newcomer to Linux to configure and compile the kernel? Not at all. But the point is that we are not gonna provide a huge source release, just the basics. I'm hoping for many user contributed plugins that make it quite useful. Of course, we will have to make the basic ones (UI, Protocols, new features) by ourselves at first. > Providing backwards compatible APIs and ABIs ain't fun. It's difficult > and tedious work which steals developer's time that may be better spent It isn't that bad. Besides, there is a nice guideline available from KDE: http://developernew.kde.org/Policies/Binary_Compatibility_Issues_With_C%2B%2B Jon
