I don’t know if anyone here has ever tried cross compiling Qt4 apps for Windows on OSX for PPC but my attempts at persuading Google to divulge the vital information failed, so I suspect not. I recently announced that there were now official nightly builds of Psi for OSX available at http://kismith.co.uk/files/psi/mac/nightlies. As these were due to a donation from a user of some RAM to fix up my mini, I felt compelled to try and maximise the return on investment and try to get win32 builds going. It turns out that running a virtual host for Windows and compiling them natively wasn’t going to happen so I had a go at cross compiling them; I had something of a fight with my Mac Mini on the issue but, eventually, the mini lost the will to resist and we now have nightly builds of the Psi mainline for win32 available at http://kismith.co.uk/files/psi/windows/nightlies. Of course, as with all builds of mainline there’s no guarantees and it’s likely (certain at the moment) that there are an assortment of crashes just waiting to happen but there seems to be a lot of demand for crashes so we’re happy to oblige .

As for the progress of my SoC project, I’m pleased to say that I’ve been making progress, albeit slowly, on the Python scripting plugin and the general infrastructure of plugins. Fritzy’s even been able to write a short script which parses incoming messages to extract the elements. It’s been suggested that it’d be ‘cool’ to have Psi flash the keyboard toggle/status lights when there’s an incoming message – if anyone can post how to do this in Python, I’ll even throw up a quick script for Psi to do this and commit it to mainline. I’m very near the stage now where ‘all’ that’s left is the biggest task; the plugin API and the associated glue. My next step is some sort of outgoing event hooks and probably writing an ‘echo’ plugin to test them. At that stage it’s possible to write some (pretty basic) useful plugins and I hope filling in the rest of the (large) blanks will result in an SoC project everyone can consider a success.

Oh, and my thesis? The deadlines are getting tight, very tight indeed.

Categories: Jabber, Psi, Research
4 Comments »

The second Summer of Code is upon us and it’s a good year for Jabber projects and for Psi in particular (this year’s projects are listed here: JSF SoC 2006 Projects). Of these 10 projects, 4 are aimed at directly at Psi, with a 5th hanging in the balance for the client to be used.

My own project is among those chosen (for which I’m obviously delighted). This project is to add plugin and (Python) scripting support to Psi, something which should be a great benefit to Psi users and also to people wishing to implement new protocol features, who will be able to build upon the Psi codebase to quickly prototype them in Python.

Anyone interested in tracking my progress should keep an eye on my previously mentioned project page on the JSF wiki (unless Peter Saint-Andre comes up with some other tracking system).

Just to prove I haven’t been idle, here’s a quick preview of the early plugins work (no, it’s not faked, these are real (although non-functional) plugins being found).

Oh, and my PhD/thesis? Things are finally looking up again after a stressful period of fear, uncertainty and doubt. I’m constantly amazed by the emotional effects being near the end of a PhD seems to have and I’m extremely grateful to have Cath around to help me stay sane.

Categories: Coding, Jabber, Psi, Research
3 Comments »