After hacking the old to just barely use plString, I received some UTF-8
asserts. So, that seemed like a good excuse to convert the whole mess to
plString. Note that the output is now saved in VaultClient.log. This
should help with getting some of the chaos out of OLD_ASYNC_LOG.log
In this case it now also works for non-owners of the age. Previously it only worked for owners because the ageInfoStruct does not contain the vault node ID, so it needed to be looked up somewhere, and that was in the AgesIOwnFolder.
This reverts commit b393947cd5.
The commit was seemingly harmless, but it created or otherwise uncovered
stack corruption deep inside the vault voodoo. While I would like to know
what was up, I'm tired of dealing with it. The crash addressed here was
fixed in a subsequent moul-scripts commit, so let's just toss this one.
This code called back into the python vault operation thingy, which
deletes itself. So, anytime we got dupe CBs, there was a use-after-free
error. Nasty crashes. I tightened up the CB management, threw away some
weird faux-management mess, and squashed some potential silent bugs.
fix plVault LogMsg calls + remove useless wide char usage
fix usage of plAvatarMgr::FindAvatar() and plLinkEffectsTriggerMsg::SetLinkKey() with temporary variables
plNetServerSessionInfo.h need std::string
fix plProgressMgr.h: NumLoadingFrames() have an useless class context information
fix plSocket usage of non-standard NULL
A typo introduced in 49dfb4e5 caused camera stacks in AgeLink nodes to
overwrite spawn point names. This broke shell and journey cloth links in
the Relto bookshelf.
Function and macro names for endianness were previously vague,
and on big-endian systems entirely misleading. The names are
now properly descriptive of what they actually do.
