Closed axxroytovu closed 1 year ago
If you have any links or them in comments, but sure, we can do our best approximation
Also I'd use -theme- not -thm-, thm looks like a set code
That's 87 more booster types. This is great, but at some point I'll need to really optimize this, or add some precomputed booster cache, as the server is taking longer and longer to boot up.
I already made e: queries a lot faster (previously it was checking all sets as you can do something like e:"Commander"; I short circuited it for set codes), and that made a huge difference, but it's still kinda slow.
But that's just a note for me, feel free to drop as many booster types as you want.
I have to disable 6 of the boosters, as they were getting into infinite loop while trying to generate random packs.
It would be nice to get better debugging tools for it.
Which boosters? I thought I caught that.
It's afr-theme-dungeons.yaml iko-theme-b.yaml iko-theme-g.yaml iko-theme-r.yaml iko-theme-u.yaml iko-theme-w.yaml which all want more cards than they can get.
This was supposed to raise exception, but I didn't get correct code for it:
Error building iko/uncommon: No duplicates support for any
Now that I fixed error check, it at least crashes instead of going into infinite loop.
I guess I'll need to implement duplicates: true support for any.
Enabled again.
Weird. I wonder why those crashed and not the others. Glad it got fixed.
There's 11 possible uncommons, and pack definition wants 12 of them, so it's pretty obvious. IKO has a lot of multicolored uncommons, so it was probably reducing the number to below the limit.
There are almost no resources on the distributions for theme boosters, but this is the best I was able to determine from watching some pack openings. Is this OK to support even with limited knowledge?