> I do hope to someday come up with a sane cross-platform build pipeline that doesn't involve a massive amount of irrelevant junk (e.g. nodejs, the nearest virtual sledgehammer sitting next to me when I set it up). Make is always a solid bet, but I need to tinker with it a bunch to make it work properly on Windows, unless anyone's got any ready-made suggestions there.
Make is also brittle and has extremely poor portability (because it runs external shell commands with no mechanism for installing any that are missing). You're better off keeping Node, IMO.
> Heh, I ought to do that. The topic just came up yesterday in another context, and I notice it's pretty easy to modify existing Git tags and upload "binaries" on GitHub. Best if I do that since the source code zips GH auto-provides aren't directly loadable into *ZDoom since the pk3 structure is in a subdirectory.
Note that modern *ZDoom does support loading archives that contain a single subdirectory (see relevant thread:
viewtopic.php?f=18&t=53522). If a module doesn't need to be built at all, then you should be able to play the source code zip directly.
> And yeah, now that I've started using Git for modding, I can't imagine how I lived without it. I just recently had a mishap with another mod (Eriguns) where folks ran into a pretty severe bug in the latest public release which I'd fixed in my local copy (but never uploaded). I spent ages running around in circles trying to reproduce the issue because there were no changelogs or commit history to fall back to. Naturally, the project uses Git now, for everyone's mental health.
I've been using Mercurial for modding for some years now. It's particularly been a lifesaver for dealing with Brutal Doom and its bugs, hacks, and messy code, and for keeping my changes merged with new official releases. Lots of bookmarking (Mercurial's equivalent to Git branches), merging, cherry-picking, etc. In case you're curious, the repo is here:
https://bitbucket.org/argv/brutal-doom