Messages

Another benefit, actually, is that if a specific recipe starts causing serious problems, DMs may confiscate it and reshuffle the crafting system. Since everyone's recipes are linked directly to their corresponding items, no serious crafting progress is lost and DMs can perform shuffles more casually.

If recipes were randomised for each PC, this would not change the items used to access landscapes. It would only change the permutations used by characters to access particular items (ie. 'recipes' are no longer a thing).

By breaking down and randomizing landscape composition in a controlled way, this problem is also fixed by this remedy:

4.Landscapes are given a variable number of ingredient groups as parameters, each representing a different "sublandscape". For any individual PC, on a pseudorandom basis moderated by the PC's surrogate ID, an ingredient will appear in only one of the sublandscapes. This will virtually eliminate the ability of players to carry across landscape knowledge between PCs.

Basically a "goblin ear" might unlock a landscape for goblin candies on one PC, but this may not necessarily apply for the next PC, for whom instead goblin candies will appear in a landscape unlocked by "bottle of piss" instead. No more jumping to all of the correct landscapes every time you "start over".

Possible implementation alternative:

1. Two servers are up at all times
2. One server is marked active, and the other server automatically redirects to its active counterpart.
3. Reset occurs on the inactive server first, and only when complete is the active server marked inactive and players are transferred over to the newly-reset active server.
4. To reduce the annoyance of being server-jumped immediately on login, the EFU launcher would be equipped to determine which of the two servers is active and immediately log into the correct one.

Downtime due to resets is again eliminated, idempotency requirements are relaxed and quest-related complications are eliminated.

I think you misunderstand. The collaboration issue is fully fixed by this:

2.Recipe discovery produces a tiny, undroppable item that, when used on oneself, brings up a menu asking whether you want to craft the item again. The same DCs are rolled, all of the same risks are present, and the presence of an alchemy or herbalism kit is still necessary. This is achieved by placing all finalized hash calculations as local variables on the recipe item.
3.This tiny, undroppable item can be destroyed via its menu, or used on someone else to optionally share it with another crafter in exchange for a small XP and gold penalty for scribing.

You can still share recipes. As many recipes are composite and have several craft stages, you can still share parts of recipes. Collaboration is stimulated by taking away players' familiarity with crafting landscapes, but frivolous sharing is discouraged by the mechanism of XP/Gold penalties.

Resets are irritating, time-consuming for players and damaging to immersion, but I believe that there is a better way to handle them more seamlessly in a way that will be transparent to both DMs and players. An idle nwserver instance consumes very few resources. Therefore:

• EFU runs two instances of nwserver simultaneously, sharing the same vault and sharing the same server vault directory, making only some minor adaptations to ensure server idempotency against the database.
• When a DM triggers a reset, players are given a brief 1-minute warning of an upcoming instance switch, following which their locations are stored on their tokens and they are transferred to their corresponding positions on the sister server. New players logging into the now-empty server are transferred over to the sister server.
• The emptied server is reset with the updated module.
• When the process is complete and the new module is running, players are given a 1-minute warning again and are then transferred back over to the other server.

Downtime due to resets is eliminated, however there are a few considerations to be made:

• What is done with people who are mid-quest? Complicated issue, possibly best to introduce an update mechanism that triggers automatically when everyone has left quest areas.
• What is a PC's starting position on the updated server? Is it the position he or she was at previously, or does the PC return to their spawn point?

The crafting system is a cool and interesting aspect of EFU, but it has several well-documented flaws:

• Recipes are identical between any two characters, including ones played by the same person.
• Even if, hypothetically, recipes could be randomized for each individual character, items fall within predictable landscapes and finding new recipes for the same crafting paths that lead to "good" items is just a matter of grinding.
• If, hypothetically, recipes were randomized, collaboration would be destroyed and crafting would become even more unnecessarily isolationist.

I am convinced that there is no need to throw out the baby with the bath water, and that the system can be salvaged with the following steps:
[LIST=1]

• Randomization of crafting outcomes per-user by appending the individual PC ID to crafting hash calculations.
• Recipe discovery produces a tiny, undroppable item that, when used on oneself, brings up a menu asking whether you want to craft the item again. The same DCs are rolled, all of the same risks are present, and the presence of an alchemy or herbalism kit is still necessary. This is achieved by placing all finalized hash calculations as local variables on the recipe item.
• This tiny, undroppable item can be destroyed via its menu, or used on someone else to optionally share it with another crafter in exchange for a small XP and gold penalty for scribing.
  
• Landscapes are given a variable number of ingredient groups as parameters, each representing a different "sublandscape". For any individual PC, on a pseudorandom basis moderated by the PC's surrogate ID, an ingredient will appear in only one of the sublandscapes. This will virtually eliminate the ability of players to carry across landscape knowledge between PCs.
  

It may or may not also interest EFU DMs to create less deadly effects on failure and lower DCs, but correspondingly less "epic" higher-tier items. Increasing frequency of success and item discovery will also make the system less "grindy" and more accessible to normal human beings.