2014 Summary-Technical article

2014 My number of journals in Csdn fell to 8, which seems to be busier than ever this year. In fact, whether in technology or technology, I have gained more than 2013 years in 2014. it feels different to work in a start-up team. The changes in the environment are fast and unexpected and require me to learn more and more quickly.

One of the biggest technical surprises of the year was spending more than half of the time on the client side, spent on unity--, especially Ngui. This is very different from the hope that I had to focus on unity in the early year when I focused on server Linux and Lua. At the beginning of the one months of the login process, it is inexplicable to be caught to do the UI, but at that time I was basically ignorant of unity. But other colleagues did not know much about unity or Ngui at the time. So under the pressure of demand, I had to devote all available time to the study of Ngui, caught the opportunity to study Ngui source code (good source) and unity documents, and often hands-on practice. There was a weekend trip to the seaside, and I printed out the UIPanel code and took it to the seaside to read.

Fortunately such a day after the big six months, especially after several large-scale optimization of UI performance, I basically read Ngui code, mastered its internal principles, but also found a lot of UI production skills. Since Ngui actually walked the process of Unity's 3D model rendering, essentially no different from the model rendered in the scene, I also learned some knowledge of graphics and rendering-such as Drawcall, cropping, rendering order, and Z-cache. At the same time, the UI framework, UI specifications, and UI tools for the project itself have been built and perfected from scratch, and the project is starting at the beginning of the year when only one or two programmers groped for the UI, and now most programmers are able to do the UI, even the art and planning stages of UI tuning. So, while the main focus of the year was on UI technology unexpectedly, and hopefully in the future, I'm very happy and satisfied with the UI gains this year. I like this to master the original knowledge of the familiar.

At the end of the year, the main course took care of my personal interest in the work arrangement and arranged the development task for me. So I can finally return to the back end of a good experience. Although the time is not long, but I still take the opportunity to the service end of the various modules between the organizational logic, the various server threads, as well as the network module to understand a bit more, and the use of vim as hoped at the beginning of the same as a bit more slip. At this stage I also learned the basic use of git. But I just returned to the back end of this time because too excited, wayward a little, spent more time to play the technology and learning from the project far from the knowledge (such as Git), slightly delayed some progress, a little sorry. Fortunately, the leaders are more tolerant of me. However, at present, the project is still in the forefront of the performance of the stage, art or often toss the program, so I still occasionally be caught back to engage in UI. Hopefully, the project will be tested on-line this year, as soon as possible into the backend performance and stability stage, then the backend will have real thing experience opportunities.

Actually, I'm not the only one in the studio. A programmer unexpectedly spends a lot of energy on the client in 2014. It should also be related to the big environment of the game industry. In the past two years, the game industry has entered the era of hand-tour on a large scale. The performance of the mobile side is far less than that of the PC, but the expectations of the market and designers for the performance of the picture are not so much reduced-especially since many developers are moving from end-trip projects. While the picture is good and the experience is smooth, at the same time the development field of the 3D tour does not seem to have enough mature engines, toolchain and practices for this demand, and the unity engine itself is trying to improve itself to meet market demands. In this case, project programmers can only spend more time groping the front-end approach, trying to meet the art on the mobile side of the "perverted" design needs. In contrast, the service side of the hand-travel era will not be more than the end-travel era has higher technical requirements, even because of the decline of MMO in the hands of the era, the hand-travel era of service pressure may be lower than the end of the era of travel.

In addition to the need for programmers to spend a lot of time on the front-end, it also leads to more game and even conflict between programmers and art. This problem is not just about our projects, but in the communication, I find that the team of other game companies also have this problem. The following log summarizes this matter in detail.

2014 Summary-Technical article