On Sun, Apr 17, 2011 at 02:28:04AM +0200, g...@crawl.develz.org wrote:
> commit 6da8dc81e50375633b768d3bbf1e1e8b07b9ebb7
> Author: Jude Brown <bookofj...@users.sourceforge.net>
> Date: Sun Apr 17 10:21:12 2011 +1000
>
> Resolve Lair ending loot issues by placing one or multiple vaults.
[...]
> It also removes all PLACE: Lair:$ tags and instead places Lair ending
> vaults in _builder_normal -- mimicking how PLACE: Lair:$ would be
> accomplished but instead of placing a single vault, placing two vaults
> dependent on tags.
Is it still possible for map generation to not go through _builder_normal()?
I think that with Zaba's changes this way is safe, but it's better to
confirm.
> + if (one_chance_in(3))
> + // Place a single large Lair vault and pretend it was from PLACE!
> + else
> + // Try to place two small ending vaults instead!
I'd say it the large vaults are much better, so what about bumping them from
33%? I'd say 66% would both give them proper recognition and ensure they
see a decent rotation. For the latter, perfect fairness would currently
require 60%.
Looking at the new version, it still gives 2x the weight to small endings.
--
1KB // Microsoft corollary to Hanlon's razor:
// Never attribute to stupidity what can be
// adequately explained by malice.
------------------------------------------------------------------------------
Benefiting from Server Virtualization: Beyond Initial Workload
Consolidation -- Increasing the use of server virtualization is a top
priority.Virtualization can reduce costs, simplify management, and improve
application availability and disaster protection. Learn more about boosting
the value of server virtualization. http://p.sf.net/sfu/vmware-sfdev2dev
_______________________________________________
Crawl-ref-discuss mailing list
Crawl-ref-discuss@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/crawl-ref-discuss
