Here Ill outline what i want to accomplish in the coming months. Ill explain a little of my situation, the status of my project and their most likely fate.
Progress & Future
Battlehub has made progress since the last build was handed out to the private testers. I have numerous plans, and Ive made a handful of changes to the UI, btu of ntoe are the sweeping architectural changes I’ve amde and will make.
The old battlehub used a more rigid messaging system and relied on the main lobby object, whereas the new battlehub relies more on asynchronous messages. This increases flexibility and can improve performance in various areas. The idea of message channels also allows for further performance gains, aswell as a means by which to implement multiple server support. I will need to modify the architecture further to allow these things to work while uncoupling the old further so that I can replace it.
I moved a lot of basic code out into an external library which I have since been able to make a bot out of. The ‘Battlehub framework’ is currently in the spring svn, although I have some modifications to commit to bring it into line with the refactors I mentioned above.
The GUI is also moving towards a more modular widget style interface. Support for the python scripting API when completed will allow users to mix and match python UI widgets as they please, and to add and remove those they like. One could take out the map selector and replace ti with a better one if they wished.
This whole widget based system would be tied into the ‘profiles’ idea in that the initial page shown would eb a widget, thus allowing content developers free reign over how the lobby appears. I intend to integrate a full browser into the lobby which would be accesible by the scripting, thus content developers can write entire campaigns and GUI systems if they wished, and they could do so using the tools best suited. they could create flash GUIs and have movies and animation, and do work in photoshop and After Effects.
I also want to further decouple battlehub from spring. I will move engine support into python scripts so that support for engines can be added by users and developers of other engines and then pushed onto the system to be distributed accordingly.
I expect the architectural changes will be done before christmas but could be nearer to february of next year. The GUI changes will likely be compelted by easter.
So far Ive made bug fixes and a handful of optimizations. I started working with some university students who were using it as part of a university coursework, and Ive made a few refactors of the task system code to be more object oriented and provide more flexibility.
I intend to make a few more minor refactors regarding resource systems. This should make resource representation more generic and make widening support for newer games much easier.
I have no timeframe for this. The changes I proposed are not wide sweeping and could easily be done in a few hours.
Since the file site started I have seen it rise to second position in popularity behind JJs site. It apepars most of the bandwidth mirror usage on JJs site is channeled into my webspace, although my main filesite is not coupled with the mirror system JJ and I set up at jj.darkstars.co.uk.
I have also noticed that the file wiki I set up is not being used by anyone save a minority. Considering the overhead of loading the iframe it might be worth removing this feature, could anyone comment on this?
With the recent return of Unknown Files however I have a problem. I do not have the server resources to rival JJ and iamacup. JJ does not wish to share his codebase and let me work on spring files, and the meer notion of working with iamacup is impossible. My own file site is simple in comparison.
So, I am goign to rebuild the file site. It will be much more flexible and it will not be confine to the file site idea. It should be a lot nicer than anything thats come before it, and should be able to win over the traffic from JJ and Iamacups sites. However it will take time.
I estimate that I will have the new file site up during next year, possibly june, july, but I am worried that may be optimistic.
As a part of the new filesite and the battlehub improvements, I aim to consolidate the two into each other, as much as possible. I want to have Application store/extension site style system for the scripts, aswell as for downloading. This will take a lot of recurring requests and deliver them all at once. I want battlehub to be distanced from spring and take up a more generic position. If this is done correctly the integration should provide a strongpoint for the projects, by making things easier for the end user, aswell as making them easier to maintain.
My ETA times may be optimistic or too conservative, Im not sure. I have started a new job as a part fo my degree course, and I now work at Sony Liverpool. As such I am not able to spend nearly as much time as I’d like developing code for my personal projects and I must be realistic. This was one of the reasons I stopped working on the new spring site. The introduction of fnordias work was both demotivating and changed the value of continued work such that it became a waste of time and effort, thus it was cut out.
With the new python scripting I expect it will be much easier for other people to contribute to the battlehub project, and scripts written for tasclient and springlobby will be usable too since they share the same basic API. A number of things will also now be possible thanks to the necessary changes and the codebase is much easier to take in for potential contributors.
If people have any ideas or requests, I’d like to hear them. Thankyou for reading, and expect updates in a months time!