Wow, it's been over two years since a news post has been put up here. I don't really know if anyone is even following anymore! So I'll cut to the chase. For the past two years I've been heavily involved in the NeL project, which is the engine that drives Werewolf. There have been a great number of improvements, set backs and strides. One of the key challenges was the failure of both Nevrax and then the following year Gameforce France (whom bought Ryzom.) At this point the NeL community forked, resulting in this (opennel.org) web site. About a year following that a new company bought Ryzom and we began the process of merging the fork back into the new company, the result is dev.ryzom.com and it has been a fantastic success.
Some good improvements have been made and a couple are on the cusp of being "broken" which are crucial to the development of Werewolf. Chiefly among these is the ability to include a NeL graphics context within an existing window object. This functional is important so that we can begin creating cross-platform, non-MFC tools that function similar to the NeL Object Viewer.
Werewolf has been refactored slightly to remove the integrated login tools, functionality that, should we desire it, would be easy to put back in place as most of the actual code is replaced by the NLMISC::CLoginClient methods now (a huge savings for Werewolf developers.) I removed this in favor of a Qt-based cross-platform launcher. The reasons behind this are not entirely related to Werewolf - the intention is to have the login service (LS) be communal, allowing players to use a single launcher against a single service to play any NeL-based game online that they have installed. That project has a long way to go but since the basis of it was in place for my Snowballs development efforts I chose, for simplicity, to make the change in Werewolf. It eliminated a lot of code in CNetworkTask. A lot of fairly ugly code, I might add.
To get things running in Linux with the most recent GCC, which I have been forced to adopt as my preferred development environment due to hardware problems, I had to do away with my registrar code. This, as it turned out in retrospect, would ultimately have to have been done for Visual Studio 9 as well as it has also become a little more strict. While it was an incredible hack that made my life easier it was, nevertheless, just a hack. I opted for the course of simplicity. In WWCOMMON I added a new function called registerEvents() which you are obligated to call upon initialization of the service/game/tool you are coding should you want the polyptr's for the various events to work.
Additionally I ported the build to CMake. With all of my interesting post-build and pre-build scripts no longer necessary this task was fairly trivial. It allows me to work off of NeL in an installed directory on Windows and Linux. Previously, especially in Windows, we were sourcing our NeL-based Visual Studio projects directly from the NeL Subversion repository. For someone like me this is a problem as my NeL source tree tends to be "in development" perpetually. I also think that it better simulates what an outside contributor might be exposed to.
For the website I took back up the task of cleaning up the wiki that you see here. There's still a lot of work to do here. I believe Jason began the process of doing this so we had a number of duplicates such as Werewolf Server and Server - I jockied things around and moved them into the simplified destination that Jason had originally intended to use. I also limited access to the World Design sub-sections, specifically the General Timeline and the Encylopedia. You may notice that there are two in there - I did not know what state Henri, Elo and Jason had left this migration to the new encyclopedia format. It may have to be taken back up by them or potentially by new content writers if we succeed in attracting them. I've noticed that I'm linked to our old server for our art resources (see the media information for the Ironic Demo design documents.) I still need to figure out how to get all of those reference images, those are beyond value.
To prepare for the next "lunge" of the project I also created a new JIRA project. At the posting of this news I have 16 issues open and have added a view to the release portlet for the Werewolf project onto the wiki homepage. Most of the issues exist in the 0.1 version but you may note that most of them are also trivial. The ones that aren't I think are crucial to returning Werewolf to a usable state. The first 2 and probably 3 versions of Werewolf will likely not address advanced features or even nagging problems of any outstanding complexity. For example the failure of the client and server to keep in sync without introducing jitter via DR updates is beyond my capabilities. An issue will be opened to address it at some point but merely getting a usable framework is of the utmost importance.
Keep an eye here, I'll try and do weekly updates.