As far as thoughts on the release go, I can definitely say that we consider this release to be a very successful one, and are very happy to see how much you, the players, seem to be enjoying the changes and additions we’ve made to the game. The Building Ownership Revision in particular is something we have spent a massive amount of time and effort to first implement and then tweak and polish, and it’s very satisfying to get to read all the feedback and discussions around it, now that you get to try it out for yourselves.
When we announced the delay of 1.7/SoI, we did so because we wanted to use the extra time to focus on the quality of the release, and this is precisely what we did. We were able to use the extra time to polish and balance the new features, improve the AI, improve performance and of course fix a lot of bugs. In hindsight I can absolutely say that this was the correct decision and that I am very proud of what the team managed to accomplish in the extra time we were given.
The 1.7 performance improvements in particular is an area where I know a bunch of you have expressed surprise that we didn’t really talk about it much before release. The reason we didn’t is that the ‘real’ impact of such improvements are actually really tricky to measure until they ‘make contact’ with the playerbase, so to speak. We’ve had internal metrics which showed us that 1.7 was substantially faster than 1.6, but those metrics were collected on a limited number of hardware setups, and we wouldn’t really know how it would shake out on the thousands upon thousands of different hardware configurations that are out there until you actually got to try it.
Fortunately, as far we can tell, the overwhelming majority of players are in fact experiencing considerably improved performance in 1.7, which honestly is near the very top of things that I am personally most happy about regarding this release. Getting there was by no means a straightforward process, as 1.7 introduced numerous new performance challenges, not the least was the addition of AI construction calculations for foreign investment. So how did we do it? Well, the modifier rework we mentioned in Dev Diary #120 was probably the single biggest individual contributor, but it was actually the result of dozens upon dozens of improvements coming from across the team. A few examples, in no particular order:
- The AI spending system was rewritten to be much smarter about which data was updated and when those updates were actually needed
- The design team made numerous changes to events and other parts of the script which were running slowly
- Improvements were made to market updates to avoid unnecessarily frequent updates of pricing data
- Employment was made much more performant by eliminating ‘rounding errors’ in the hiring logic that resulted in numerous insignificant employment changes
- The programmers expanded the use of smart caching and multithreading pretty much across the entire game. Military graphics and other map graphics were also heavily optimized.
However, with all that said, a release of this size and complexity will always bring with it some bugs and balance issues that we weren’t able to discover and fix in time and which are now our top priority to address. When you are reading this, hotfix 1.7.1 should already have dropped as of a couple hours ago, and we are planning to follow it up with at least one more hotfix, tentatively planned to release early next week, and I’ll wrap up this dev diary by listing a few select fixes that 1.7.2 will contain:
- Substantial improvements (through AI improvements and balance tweaks) to the AI’s ability to execute on historical and historically plausible outcomes, particularly in nation-forming and ability to pursue Journal Entries such as the Meiji Restoration and Tanzimat Reforms
- Setup improvements to relations and AI attitudes to more closely match history
- More aggressive colonial AI, particularly for late-game land grabbing in Africa
- Making it harder to get reparations by occupying insignificant colonies
- Fix for poor building browser performance when using scrollbar
- Improved leverage UX
- Allowing Unrecognized Major Powers to form Power Blocs
- Balancing the Great Game to be less biased towards Britain
The above is of course NOT an exhaustive list, and we are not ruling out additional hotfixes after 1.7.2 if needed - while I think we released 1.7 in a good shape, we still want to ensure that any significant new bugs and balance issues are dealt with as speedily as possible. For this reason a part of the team (myself included) will continue working for a few weeks into July (taking our vacations later in the year instead), so that we maintain the capacity to release fixes as needed.
That’s all for today, but you’ll be seeing me again next week as I make the customary roadmap update and tell you about some of what we have in store for 1.8, 1.9 and beyond. See you then!
