Monthly Archives: May 2015


Found three unused EDU slots in M2TW! These were occupied by the unused units: Bohemond of Taranto, King Guy, and Qilich Arslan.  These seems to be meant for the unfinished pre-Kingdoms Crusades campaign (the latter two made their appearance in the Kingdoms expansion as minor characters).  The trouble would be to decide which factions to allocate those slots to – it would probably help out the game balance by giving the less polished fringe “boss” factions a look right now (Mongols, Timurids, and Aztecs).

On the side of Unity development, Unity 5 has done away with several legacy shader systems, also making it possible to display shadows directly.


Notice the distortion in the shadow due to terrain geometry variations.


Has it really been a week since the last update? Really need to get back in the field, else time will slip by very quickly.

Currently thinking of using a decal system in Unity3D to test out the Ochre project.

On the modding side, importing custom battle maps from the Kingdom campaigns to the base game is relatively straight-forward.  Though there is a catch of figuring out the correct climates based on the coordinates in the base folder.  Still, when done properly, the results can be quite pretty:

kingdoms 2015-05-28 10-26-50-55

Battle selection screen.

kingdoms 2015-05-28 10-29-05-05

Battle map.

kingdoms 2015-05-28 10-31-17-52

And horse kebab.

Keeping up Appearance

You would think that the abundance of crossbows in Medieval 2: Total War would mean that the artists took a little more time to model the weapons correctly, that was not the case.  I am seeking to rectify that by adding the windlass and cranequin loading mechanism to several units.

Aventuriers_Steel_CB Granadine_CB_Top

LODs are really the bane of my existence…

End of Disk One (Thesis)

Summer is ahead. Where are the cicadas?

Before that, a short report:

Post-Mortem/After Action Report


February 16, 2015, judgment was passed at the end of the freakishly warm winter, only five of the original ten group prototypes made it through the industry filters. The people of the Fifth Cohort hastily coagulated under their chosen banners, again forming balanced teams. This second exercise was conceived to create the Thesis Games for the Fifth Cohort – a year-long project that would hopefully blossom into publishable titles for all of its members. Alpha was officially declared on April 28, 2015, before the public showing of the games in the EAE Festival.


Personal Contributions

Ochre, previously known as Plato’s Cave, was composed of two producers, seven engineers, one artist, and three technical artists (including myself). With the hearty helping of engineering experts, I followed the familiar routine as before: primarily functioning as an artist with some expertise in engineering.

The compressed time frame of production demanded the rapid production of placeholder art assets; it was a role that I comfortably fitted in. Both before and during my tenure in Ochre, I have created assets ranging from textures, 3D models, animated 2D sprites, and their previews in .gif. As a result of the white-boxed nature of the alpha, a goodly portions of my creation, notably the controllers both in-game and out-of-game was featured in the showing.

Besides the creation of art assets, I also provided the research for the art team based on archaeological samples, as well as accompanied the producers to court the on-campus Native American liaisons and keep them updated on the progresses made by the art team – it would seem that I have other uses outside of development.


What Went Wrong

I left the Crash City team because I cannot rightfully assert ownership to assets that were made by others (that would spell my failure as an artist and crush what is left of my self-confidence), and through some odd twist of events, I found myself a member of Plato’s Cave, a promising novel project with a number of talented teammates… the following will be the kinks that have insinuated themselves onto this project’s unfortunate existence:

We have wasted the first month of the development cycle in one technology, only to switch to a completely different one – OpenCV was its name. The time spent was with splitting up the large engineering team to work on various mechanics instead of using that numerical advantage to test out various different types of engines. We dropped all of our codes during the transition. Switching to Metaio allowed for us to accomplish what we had during EAE Fest, though at what cost? The remaining issues sadly have little to do with technical limitations; rather, they were team-dynamic related.

Ochre experienced more inter-personal drama in its short existence of two months than the six months of prototyping that preceded it. The most ridiculous one to look at in hindsight was the argument over the future of the game – starting out with a casual comment to port to mobile, and degenerated into behind-the-back machinations in a factionalised team to oust certain members… fortunately we came to our senses in the final month of April and put aside our differences, but the damage to team solidarity was done.

The next issue was with absent team members, it was somewhat disheartening to work at 50% capacity most of the time in the art team, not to mention a general lack of initiative for participating members to cover for those that were missing. We were able to overcome the majority of the issues through sheer perseverance and perhaps a bit of luck, though it begs for the question of how much more we could have done if we did resolve the issues more quickly.


What Went Right

Looking on the bright side, our team did have an impressive track record of excellent task division supported by adequate technologies, for one, we established a file-sharing/version control system early on. Although it has its shortcomings as a bottleneck when we required multiple team members to work on the same scene at the same time, the GIT Hub system has done its work for now, and summer is the time to look for an alternative. The task division of splitting up the group into small, two persons teams where the two could cover each other’s was also effective when we were working on multiple scenes.

Feature creep was made controllable with our change request/user story function in that everyone would have access to the request, and interested group members could attempt to implement them in their spare times.



A majority of the issues can be seen as growing pains – the use of novel technologies raised a plethora of design questions that of which were exactly the ones that we will answer with the game. Experiences will make us more able to ignore, or better yet, resolve many issues that were witnessed with this experiment. Better to remain hopeful, and proactive, for the best is yet to come.