KABoissonneault
commented
9 months ago Items in "DFU next" (whether we end up releasing it as 1.0.1 or 1.1) so far:
Jagget:
- Localize %dng text macro
- Fix SecondaryDisplayName localization. This is used by enchantment effects like CastWhenHeld, this localizes the actual spell it casts
- Expose text labels in the Banking and Character sheet windows to mods. This lets localization move those labels around
- Fixed building name localization in some contexts
- Fixed localization in %fcn text macro
- Allow modded books in packaged mods
- Fallback for questlog localization from older saves
- God name localization in Donation service
Podleron:
DunnyOfPenwick:
kaboissonneault:
- Fix error during Lycanthropy transformation and other magic effects
- Fixed Advanced Climbing ignoring skill checks after first climb, added "tpc" (Toggle Perfect Climbing) cheat command
- Fixed WorldData partially loading disabled mods
- Fixed error when watching an Ancient Lich from behind
carademono:
numidium:
Starting an issue since I can't ping people from a milestone.
So, despite DFU having been feature complete as a "DF experience" for qujte some years, and DFU 1.0 being pretty stable as a complete game release, users still have requests, expectations for the future of DFU.
Issues generally come in the nature of:
I don't see us addressing any other sort of issue (ex: new distributions), not the way things are.
The way I foresee things going, one of the big themes of post-1.0 development is going to be decentralization - away from Interkarma, yes, but also away from a new replacement "project leader". None of us have the intention of working on this full time, and I don't think the process should be held back by the availability of a few key maintainers.
Of course, there's a transition involved. Nothing would be happening right now if I wasn't leading a charge for a continuation. But in this transition, we need to replace "personal knowledge" with documentation that future maintainers could refer to.
To address issues in post-1.0, we're limited in these areas:
Contributions seem good so far. At least it's not the slowest part of the process.
Documentation is not great. A lot of project knowledge is still in the mind of long-term contributors who have an availability ranging from none to "can answer questions every few days". We need to evaluate what needs to be put in writing first: for example, "how to deploy a DFU build" is crucial for the future of new releases, but the exact intended mechanics of a specific UI window can probably just be figured out from the code itself and its Git history. Reviews are one of the most difficult parts right now. Interkarma justly introduced a "two maintainers" system to merge PRs, and this is something I want to keep. However, the people who count as "maintainers" for this process are:
These people have to press "Approve" on the review for anything to get done. Without it, there is no future DFU version.
In my development plan, I think we can get away with 3-4 releases a year. This means in that in the time span of 3-4 months, at least 3 of the people in this list need to get involved and clear some PRs at one point (I can't accept my own PRs after all). I foresee that even this pace is not terribly sustainable in the long term. For decentralization, we'll need some new trusted maintainers.
To any reader not mentioned above, here's a suggestion from me: review PRs, even if your review does not "count". It helps us evaluate who can be depended on for this sort of task, and having existing reviews helps other reviewers.
Later on, we can discuss adding more people to the list.
To conclude this, I will annotate a few issues and PRs as "milestone 1.0.1", which means they should be addressed before I would feel it relevant to make a new DFU release. I would like if the people mentioned above could take either a 15 minutes once in a while to read through one of these PRs, or a bit more time to actually test them. I'd like if at least one of the approving reviews actually ran the thing on their machine, even if ultimately we'll have to test the release anyway.
The date March 31st on the Github milestone tag is arbitrary, and could be pushed either way depending on what gets done. I don't think we should add any more new work except for https://github.com/Interkarma/daggerfall-unity/issues/2355, the rest can wait.
Comments and questions are welcome.