- Staff
- #1
- Oct 20, 2016
- 1,966
- 2,412
There isn't much new from the last update other than that we tracked down and resolved the root cause(s) for item stat generation for normal gear and dragon gear. We are currently weighing the pros and cons of overhauling the way that the Dragon's Prophet client accesses files and how the data system works; this is something that needs to be done before public testing can begin as it would require a whole new client entirely as the packed data itself would become obsolete.
The con is obviously that the source update will take longer to come out as a significant amount of development and testing will need to be invested into this change. It would also require changes to our custom data tool, the patching system, and an entirely new launcher. These are systems that we have already made some changes to or are things we plan to do more to in the future, though not to the extent of completely overhauling and optimizing the systems.
The pros are that we would be able to optimize the client further than we already have; we have so far doubled to tripled performance for some hardware configurations; newer hardware should be able to run around 60FPS, with some getting up to 200-250FPS. This is both GPU and CPU dependent, with load times being affected by SSD and OS configuration. We estimate that optimizing the file and data systems will bring us an additional 20-40% improvement, as they are currently accounting for 60% of the bottleneck that Dragon's Prophet currently experiences. While I also listed it under cons, the pros of overhauling the system means that we can make significant improvements to the patching system so that patches can be bi-directional, allowing us to properly rollback patches rather than just patching older patches on top. It would also mean that we can do a major overhaul of the launcher and introduce things such as a "first launch" where game data is downloaded through on first launch, allowing for a smaller installer.
Other than the consideration and research we are putting into overhauling the file and data systems, we are also looking at implementing another UI engine. The current engine would still be used as is, but we would have the option to implement new UI (such as the loot pet filter that has been requested) as well as rewriting the chat system so that we can put in support for modern emoji. This is currently being researched while we're brainstorming what is needed for the client overhaul (and whether we will do it or not). The likelihood of us converting fully from Scaleform to the new engine is very slim due to the amount of work involved, and that it would likely once again require a completely new client.
The news background is done by @Arborath (our resident masochist artist).
The con is obviously that the source update will take longer to come out as a significant amount of development and testing will need to be invested into this change. It would also require changes to our custom data tool, the patching system, and an entirely new launcher. These are systems that we have already made some changes to or are things we plan to do more to in the future, though not to the extent of completely overhauling and optimizing the systems.
The pros are that we would be able to optimize the client further than we already have; we have so far doubled to tripled performance for some hardware configurations; newer hardware should be able to run around 60FPS, with some getting up to 200-250FPS. This is both GPU and CPU dependent, with load times being affected by SSD and OS configuration. We estimate that optimizing the file and data systems will bring us an additional 20-40% improvement, as they are currently accounting for 60% of the bottleneck that Dragon's Prophet currently experiences. While I also listed it under cons, the pros of overhauling the system means that we can make significant improvements to the patching system so that patches can be bi-directional, allowing us to properly rollback patches rather than just patching older patches on top. It would also mean that we can do a major overhaul of the launcher and introduce things such as a "first launch" where game data is downloaded through on first launch, allowing for a smaller installer.
Other than the consideration and research we are putting into overhauling the file and data systems, we are also looking at implementing another UI engine. The current engine would still be used as is, but we would have the option to implement new UI (such as the loot pet filter that has been requested) as well as rewriting the chat system so that we can put in support for modern emoji. This is currently being researched while we're brainstorming what is needed for the client overhaul (and whether we will do it or not). The likelihood of us converting fully from Scaleform to the new engine is very slim due to the amount of work involved, and that it would likely once again require a completely new client.
The news background is done by @Arborath (our resident masochist artist).
