openmw.org

And premature optimization is bad m'kay, hence my suggestion it wouldn't happen soon

Or, better yet, the "blend" value is an integer (still measuring "units").

< 0 = static blends into terrain
0 = no blend at all
> 0 = terrain blends into static

It could default to 0 to maintain current behavior.

Other content could be created with a larger ECLD in mind, and thus not have the issues seen with vanilla. I think improving performance in this area would be a great idea. However, as it would only benefit non-vanilla content, this is probably a post-1.0, or at the earliest in the final stages of p...

The checkout process sounds like a nightmare. The 80's called and they want your devops back :p

If .omwaddon files are not diff-able, the format needs to change.

I don't know if it's supposed to, but this topic doesn't show on the home page as a news item.

Well written and maintained unit tests are always valuable, not just in TDD.

My C++ is admittedly rusty (mostly do C# these days, with a little bit of js and java) and I'm not familiar with googletest, so it'll be awhile before I'm ready to start contributing tests.

raevol wrote: ↑08 Nov 2017, 22:42 This is way more entertaining than it should be hahaha.

Agreed

Zini wrote: ↑22 Sep 2017, 18:23 Little update: I am starting to write down stuff about the general de-hardcoding (cell/worldspace related de-hardcoding got its own section). At 21 pages now.

What are the odds we could see that "soon"-ish?

Hello! Long time lurker, rarely a poster. In meatspace I'm a Software Engineer in Test, which means I write automated testing as my day job. I was wondering if my experience could be utilized here? Having good unit tests is always of help in a project, especially as it nears "completion", ...

Woot! Great job everybody!