I've got good news, bad news, and what I can only call subjective news. I'll start with this last category first, since it will help explain the rest.
The Subjective News
The mage patch has been delayed. The Archmage crunch, the various PRC reworks, epic PRCs, crafting changes... there is no current timeline for its completion, and there is no timeline for its release. It may come next week, it may come next year, or it may never come. I don't know, and guessing its final fate is well beyond the scope of this announcement. Rhifox is the only one on staff who can complete it in full, and it is up to her to decide what to do with it and why. (I've actually written this post without consulting her, because I wanted to announce the patch I was going to run tomorrow, and explain why it might take a bit longer than normal, so... sort of improvised everything on the fly.
)TL;DR: No official release date for:
- Archmage crunch
- Master Alchemist revamp
- Shadow Adept revamp
- Epic PRCs of any kind
- Gish class changes
- A number of other changes, whether publicly announced or not
The Good News
We've overhauled our development processes, most notably migrating off SVN to Git and introducing a barebones branching model to our versioning scheme. This means that large or controversial (or both) projects can be safely committed to version control without forcing us to stop releasing updates until the project is both fully implemented and approved for release.
I've been looking forward to this for a long time. A lot of my ideas tend to land in this category (sometimes preemptively on my part), and I end up not doing any work on them at all: I expect my raw productivity to soar, even if we don't decide to use any of the things I make. First, though, I need to give the new plumbing a good trial run, so there'll be a patch tomorrow. This should fix some of the issues introduced in the August update, particularly those caused by a script-plugin mismatch from our failed plugin upgrade.
TL;DR:
- A more regular release cycle dictated by the availability of things to release, not the completion of all unreleased work.
- Better support for those big projects that would have clogged up the release pipeline... if anyone still remains who would undertake such things.
- More things making it out of the brainstorming stage and into actual, deployable code! (Which hopefully means more things going live. Surely not all of these changes will fail to receive approval.)
- Incoming update tomorrow morning, with fixes for various newly-introduced bugs.
The Bad News
We sort of rushed into this. We didn't exactly mean to, at first, but by my estimates, a lot of our remaining staff don't yet have even the basic Git proficiency needed to emulate their SVN-based workflow. (Not entirely for lack of effort, I've been trying to jump this shark for the last six months or so.
) I don't expect there to be much trouble, as such, but it might take a while for everyone to return to their current level of productivity.Worse: The immediate purpose of the Git transition was to excise all content related to the mage patch (and malfunctioning plugins) from the SVN trunk (or its Git equivalent, the master branch). This entailed a good deal of surgical editing to the freshly-converted Git history, the consequences of which I have only a loose and theoretical understanding of: I think I've safely stripped out the mage patch, but I won't know for sure until we see the server running smoothly without issues.
There's also practical limitations on how much of our release infrastructure I can realistically upgrade before bringing the servers down. Even if all goes perfectly, tomorrow's patch will take longer than normal as I make the final adjustments to our release automation.
TL;DR:
- Smaller updates over the next few weeks/months until the whole team's up and running again.
- No patch notes for this update. Should be roughly on par with today's state, plus some bugfixes and some extra content by our builders, but I have only a general idea of what I'm releasing.
- Extended server downtime for this update; ideally still well under 24 hours, but I can't make any promises. Sorry.