A quick short detour from the simulation module to support module
Graphics setting (WIP)
Implemented borderless mode
Implemented active monitor switching for multi-monitor user
Implemented overall quality preset
Additional 4 quality settings
Visual Effects
Texture
View distance
Foliage (future proofing; in the current build there is still no foliage asset)
Implemented auto-detect functionality
Various other improvements (minor layout changes; Primary GPU detection. Resolution change confirmation widget, etc)
Implemented a custom configuration file. Several user-specific preferences now saved correctly and are persistent across playthroughs. (e.g: Measurement unit preferences, autosave interval, etc)
Nativitized most of setting logic from blueprint to c++. Better performance and easier to extend/maintain in the future
Initial implementation of Mecha AI. Previously; a mecha is only controllable by a human player. Internal progress has been made to allow a mecha to be controllable by AI. This feature will be enabled in future build
General codebase maintainability. A controllable mecha with different locomotion type now shared the same base class. Less redundancy
Various refactorings. Preparation works to start integrating modifiable mecha in tycoon module with controllable mecha in the simulation module
Refactored graphic setting logic and UI (WIP).
Fixed checkbox for FullScreen and VSync not visible inside main game setting
Other movement types will be implemented in upcoming updates
Track (e.g: Tank / Panzer)
Flying (e.g: Propeller / Jet)
Initial Implementation phase
Still using placeholder mesh
Still detached from tycoon module-> unlocked component/attachment in research tree cannot be tested. Focusing on locomotion logic completion
The room itself is also still being worked on; the idea is to add more unlockable obstacle/facility as the development progress further
Mecha control hint is displayed on the inside wall for quick reference
A reset pad used for reloading initial test mecha position is implemented inside the test room for cases where the mecha is out of control (see above GIF)
Build Update Cycle Changes Biweekly build update introduced in Update #51 has been tested for several weeks, and I am reverting back to weekly build update. Even though there are several administrative-overhead every time a new build is uploaded; at the end of the week it is still more productive
Hello! A little background on the current focus. Last year focus was the [tycoon module]. It is still work in progress; with the last update; work on the [simulation module] has begun. The 2 modules shared a common [data module] which is also still WIP. It is the aim of starting work on simulation aspect to catalyst the data module completion progress; as it is relatively easier to be tested and improved/balanced upon within a testable environment
Build Update
Unreal Engine minor update from version 4.22.1 to 4.22.2
This is a minor update; current progress is not included in the newest build as it still needs a lot of tweaking (see GIF for problem example)
Movement Progress Update
(WIP-Not included in the current build):
Logic Implementation
Locomotion logic is now in c++ for better performance
Implemented more abstract solution for leg-type locomotion -> can be adapted to other leg types (quadruped, hexaped, etc) relatively quick
Internally using an experimental feature [CCDIK] (Cyclic Coordinate Descent Inverse Kinematic) from Unreal Engine to implement foot surface detection as opposed to simple [TwoBoneIK] in update#051. This has some several trade-offs; a beneficial one is a support for a leg with multiple joints
Various other WIP improvements such as hips offset; turn-in-place detection; etc
Improved default animation workflow.
Internal animation pipeline is now correctly creating a base skeleton with root bone
Improved default animation creation. For leg-type basic movement; a total of 9 default animations per leg-skeleton are needed (forward, backward, left, right, 4* diagonal and default stance). In the new workflow; only 3 are needed; the others can be generated from the metadata -> faster development time
Implemented bone naming convention -> future implementation of other controllable legs with surface detection will be faster
Leg Locomotion movement logic; atm using humanoid bipedal
Initial implementation of dynamic foot surface detection
New build will be uploaded next week. Currently experimenting with biweekly build update to further optimize current workflow (as opposed to weekly build update). Progress update is still being reported weekly.
There is no new build introduced with this update; most of the time was consumed in non-technical works, replacement of hardware failure and recovery of development data
_________________________________________________________ Short Summary:
Solo development started around February 2016. It entered Steam Early Access on 4th May 2018. As of today the game has been in development for around 3 years and 3 months; in which the last 1 year of it has been in steam EA.
The main purpose of this announcement is to update the game estimated launch deadline (version 1.0)
Originally in the store page description; it was stated as
“Mechsprofit will be in Early Access until approximately Q2 2019.”
It has now been changed into
“Mechsprofit will be in Early Access until approximately Q2 2021.”
An additional 2 years of development time is chosen carefully with currently known data and considered to be safe / worst case scenario. That being said; I would like to also apologize for not being able to fulfill the original timeline
_________________________________________________________ 1 Year; personal thoughts
To be as transparent as possible with the community regarding the development; without being too personal; I'll share 2 short paragraphs that have no place in the usual weekly update announcements
Reflecting back on the progress gave me a mixed feeling; in one hand there has been a lot of improvement; on the other hand; I expected to at least have reached version 0.6 by this time. I have thought a lot about all possible ways to tackle the progression problem. One of the more intuitive solution in accelerating development time is by adding additional workforce which is not commercially viable at the moment. In the current indie game development scene, this seems to be the norm.
It is known from the start; that this is a marathon; not a sprint. The workflow reflects that. Practically; I always worked on the most complicated/ "boring" problem first; i.e: World event generation; Market logic; debugging/refactoring etc as opposed to a task that is immediately rewarding/visible e,g: new mecha/ factory appearance, etc.
_________________________________________________________ For the next 2 years
There were 51 weeks between the first technical update (Update #001) and the last update (Update #049). Which translate to around 0.9 updates/week. The "missing" 3 updates happened on Update#027 due to Save Game problem bug which needed more time to be resolved. I am doing this full time and according to my log; there were 2,693 hours invested in 2017 and 2,904 in 2018; which translates to around 7.67 hours per day; 7 days a week; 365 days a year for 2 years. Time invested in this year is on track to be more intensive than in previous years.
Numbers above are being mentioned to assure current supporters that even though it might take a while, I am adamant in finishing the game in the best of my ability. I would also like to thank everyone that has supported the game so far. It really won't be possible without your support
I am excited for the next 2 years; here's to a more productive development!
Improved metadata for current 89 mech component placeholder objects
LOD: Level Of Detail (each component has 2 extra LODs) -> This will reduce performance cost when there are many visible mechas on the screen
Collision Data: Continuation from Update 0.5.27.-> Randomized attachment will now attach itself closer to the surface of the parent component. Component selection in assembly now behaves correctly
BoundingBox: Camera auto center + zoom is now more precise
Not all improvements have immediately noticeable benefit in the current build (i.e: LOD). The main objective from these improvements are streamlined component generation workflow which will aid in a faster transition from placeholder to the final version
Preliminary implementation of attachment reprojection when it's parent component changes. (WIP)
Numerous other minor improvements (Attachment installation sound effect, Improved assembly center camera logic, code refactorings, etc)
Fixed [context pop-up panel] not updating when randomizing Component/ Material/ Attachment