I like it. I will miss exploring the new world from shard0, but I think it makes good sense for players who are thinking about respawning as well as new players.
Thanks for being patient with your vocal, fiesty community and working toward a solution on the API question. I know you here a lot of complaining when new features like this come out, but I think for most people (definitely for me) the complaining is because we enjoy the game and care about the direction it takes.
My two cents on being able to control CPU distribution from game code:
One of the things I would love to do with my AI is implement nomadism; code is always looking for new homes, never keeping a room for too long, exploring and finding new neighbors to interact with, whether the result is friendly trade, unbridled conquest, or crushing defeat. It sure would be nice to be able to have my code opportunistically expand into a shard without having to check on it myself, carefully evaluate the CPU / GCL allocated to each shard I'm currently on, and manually redistribute based on how much I think my code might want there.
First of all: I'm impressed and super excited that you have a prototype of this up for testing. Great work!
I like the idea of reducing TTL on creeps that use an inter-shard portal. I am a little worried about how difficult it will be to claim a room in another shard with the 200 tick penalty. But that may just be motivation for me to claim rooms closer to highway crossroads :).
I think the aura idea is interesting, but I have to disagree on the commander being the weakest. Here's some ways I plan on using mine:
* RENEW, and bring a healer along with the power creeps. Paired with an executor with HARVEST_ENERGY, the healer can tag along indefinitely, and avoid the need to return home to heal power creeps.
* SUMMON isn't as obviously useful for harvesting as Executors, but it essentially lets you replace the CPU cost of moving a hauler creep 48 steps across a room with a single action. This would be a huge CPU savings Seems like it could also be useful in combat situations.
* Being chased down by a sniper / killbot Executor? Put SIGHT on your own Executor and kite / instakill them.
* Being chased by a group of boosted creeps? Keep hitting their healer(s) with EXHAUST and force the attackers to stay back or risk getting shot down.
* BERSERK + cloud of snipe executors.
I can see HARVEST_ENERGY being useful if you pair it with a Commander that has RENEW. Then you can bring a normal creep with HEAL parts along with your roving band of power creeps and keep it sustained, as long as you regularly find neutral rooms to harvest energy from.
Unfortunately, safe mode wasn't a feature in August 2016. I must admit this would have been a good use case for it; I didn't have the code to do a sustained patrol to keep him from rebuilding his other rooms at the time, and it would have given him time to launch a counterattack. Either of those situations would have made the fight Too Expensive To Continue for me.
It seems like one of the problems is that newer players don't get a feel for how dangerous things like boosts are until they get wiped by someone using them. I don't think power creeps will be quite that drastic, but I can see similar situations arising once they are released into the wild. So: what if the feature that lets players spawn NPC invaders in their claimed rooms was extended to include a "T3 boosted" tier, that had creeps with 10 tough, 30 of ranged attack / attack / work / heal, and 10 move, all tier 3 boosted? It gives players who are interested in testing their defenses a taste of fully boosted creeps, but they can be banished at any time, and it still leaves room to be surprised by how real players choose to use those creeps as opposed to the NPC invader AI's smash and grab.
Nice! You are a gentleman and a scholar.
https://screeps.com/a/#!/history/E5N61?t=17143606
Play it for a few dozen ticks. Tell me what's wrong with this picture.
If more than one players has creeps in a room at the same time, often you get the "NO DATA" message when looking at room history, rather than being able to see what happened. This is kind of a bummer, because room history when two players are interacting is usually the most interesting. I suspect this started weeks ago, but since room history works most of the time, it took a while to narrow down when the problem actually occurs.
Recent example:
https://screeps.com/a/#!/history/W44N47?t=15944033
https://screeps.com/a/#!/history/W43N47?t=15944033
Here you can see an errant bootstrapper of mine in W44N47 wandering toward FuddyDuddy's room at W43N47. He does have a creep in W44N47, so it's not 100% of the time that more than one player is present. Maybe just if there's combat? Or if the room is claimed?
https://screeps.com/a/#!/history/W43N47?t=15944079
Here's the last tick that room history data is available. My bootstrapper is two steps away from entering the room.
Ticks 15944080 - 15944099 are unavailable. I suspect the data gets saved serverside in 10-tick blocks, and something in those blocks ended up as invalid data and can't be retrieved.
https://screeps.com/a/#!/history/W43N47?t=15944100
Here room history has returned, and the only evidence we have that something happened is a pile of corpse energy by the west entrance and towers that aren't quite topped off anymore.
What happened in there? What did my creep see that had to be silenced!?
Seriously though, this happened a lot in room battles a week or two ago, and it made it hard to watch how well automated defense code did.
Bonzai: First of all, thanks for posting this. I appreciate your approach to these subjects.
The way I see it, there are three big problems that have led to the saltiness in the ever-growing Screeps World War:
1. Not everyone gets the same story about what's happening, even when everyone involved communicates as clearly and openly as they know how, let alone when we're in an adversarial (albeit friendly) game where hiding motives and goals is part of the play. I live in a country that recently elected a new leader. Presumably we all want a good one, but we sure don't agree on who the best choice is, and we all have wildly different interpretations of what looks like the same information. It's enough to take people who would otherwise like each other and make them stolid enemies. This is one of the things that makes being human interesting, but it also sucks when you want to get along with everyone.
2. We're playing a strategic game with self-organized groups of people, but many of us don't have the investment to enforce the level of organizational discipline and unity that would be absolutely necessary if this were a "real life" political landscape. In other words, many players just want to have fun and be themselves, don't want to do the hard parts of running an organization with people in it, but still want all the benefits of being part of a group.
There is perhaps no better in-game example of this than The Culture. We are a loosely organized lot of developers who like hanging out with each other. We don't really have a leader. Our alliance charter (found here under the Charter tab) is eight words long. Our logo is a rainbow butt! And I think most of Culture's members like it this way (maaaybe some don't like the rainbow butt). I am continually surprised by the spectrum of human diversity in our alliance, and I love being a part of it. I have made friends here that I hope will last beyond whenever this game stops being such a big part our lives, however wishful that thinking might be.
But it also means we often don't have a unified voice. I've had a lot fun in the last few days posting Culture News Reports in our public slack channel about the ever-changing "Mouth of the Alliance" and their near constant replacement in violent coups. That was mostly for giggles, but I also wanted to gently impress the idea on the community that we don't have a leader, and we are all grossly unqualified to be one, either from a lack of desire to lead or a lack of disposition. So when we declare war on an alliance that threatens one of our members, we might not have all our ducks in a row for the initial conflict. And our members get to be as vocal or silent about their perspective as they want. We each take whatever role we feel moved to take in the alliance. I know some other alliances in the game are quite a bit more regimented than this (SUN comes immediately to mind), but it seems like each has some level of freedom of expression for its members.
This makes me unsure of where to go with this, because we don't all feel the same about the current war within The Culture, and we definitely don't have a "this is the message we will communicate to the world" beyond our declaration of war to show solidarity for a member. If I step back from the situation, this doesn't seem fair to the other alliances, that our political position is the aggregate of what our most vocal members say in Slack and in in-game messages. I don't really know what we would do differently, though. Maybe this is my own personality and damage coming into play, but I have a lot of resistance to telling people what to do or what to say. I could also point fingers at the words and actions of players outside of our alliance, but I don't think that's necessary, nor do I think it would help (see point #1 - we all have experienced this conflict differently).
I do agree with Bonzai's assertion that the larger, more vocal players shape the culture of the entire community, and that a big part of the draw of the community (for me, at least) is that we're a generally friendly lot and want to have a good time together. I hope we can hold on to that.
3. The in-game military conflict has taken an emotional toll on some of the players. I think this looks different for each person involved, but I know of more than one person that has lost sleep over this and has let things that are important to them in real life slide to stay involved. I know I have, and I've barely seen any of the conflict. My own personal damage means that I take the responsibility for everyone's safety and happiness on my shoulders, so I really struggle when one of my allies gets attacked. I want to do what I see as my sacred duty and protect them, and if I am anything less than overwhelmingly successful, I feel terrible about it. The game stops being fun and feels like another job, a job that is taking time away from the job that pays me money and the people in my life. I can barely focus on anything else.
This reaction is a problem with me, not the game, but I imagine a lot of people have their own version of this. We get emotionally invested in this because we enjoy it, because it's an avenue of self-expression, because we care about our in-game buddies, and it's hard to let go of that if things don't go our way.
------
I guess none of the above suggests a clear solution to improving relations between the players while the wars rage on, but I think it's a big part of what's contributing to the current drama.
Again, thanks for opening a discussion here, Bonzai. I've been thinking about doing something similar, and it's good to see others wanting us to all enjoy this in the end. And I'll talk to you about your and my personal shortcomings outside of this thread 
In claimed rooms, most hostile structures can be removed through the UI or by calling destroy() on them. However, hostile spawns seem only removable through the UI. Calling destroy() returns 0, but doesn't remove the structure.
Similarly, hostile construction sites can be removed through the UI, but although a remove() call returns 0 on them, they don't actually get removed.
I tend to agree with qzar / dissi on this. Presumably, if you make a concerted attack against another player's room, your goal is to remove that room from their control. From a balance perspective, I originally thought that the stage where you're attacking a fully built room would be the most costly in terms of resources / time spent, but unless you wipe the player, the downgrade phase is far more costly, both in terms of resources / time and in terms of lost progress if there's a short lapse in your offensive push. That seems backwards.
To look at it another way: let's say you have a GCL 15 player you want to attack with no allies to help them. All of their rooms are pretty close together, reachable by you, and have comparable defenses. Will it cost you more resource / time / effort to attack one of their rooms and push it down to RCL 0, or to attack all 15 of his rooms and remove his creeps / spawns so he can't rebuild? Assuming the player decides they don't want to lose that one room, this is not an simple question to answer, and that seems really silly, that it might be less expensive to wipe a big player than to take one room from them.
+1 to this. It does slow down the process of switching to other rooms a bit (~2000 ticks to refill), but it seems like that's not a big increase in ticks compared to the time to get to RCL 6. Bonzai's energy sink rooms would still work with a slight variation.
Having said all that, I sympathize with the frustration in implementing code that uses specific mechanics to your advantage and then having to throw that out when the mechanics change.
Not being able to easily move the root controller around would definitely hurt players who are more distributed (like me), but I understand that most people don't choose to spread themselves out that much.Other than that, it sounds like an interesting idea! You'd still be able to pour lots of energy into GCL, but you have to choose how you balance your empire's energy carefully, and there's more than one viable strategy for using it.
"Only one room can be in safe mode at the same time."
Love it. Limits its usefulness for bigger players, still useful for smaller ones. I vote we see how the rest plays out.
Also loving the changes to nukes/observers.
I like the idea of having something that protects more vulnerable players from waking up to finding their fledgling empire was crushed overnight by someone who was bored or wanted to claim rooms in a symmetrical pattern.
I don't like the idea of giving veteran titans another tool to discourage anyone from chipping away at their continental empire.
I like the concept of limiting this in some way, either by RCL, total rooms owned by player, or a combination of the two. An activation countdown like n00bish suggested might be good; novice areas would be exempt from this limitation as well.
If this were limited, it would definitely need to be something that is quick to calculate server-side.
+1 to this.
One of the things I really appreciate is the approachability for new players and how CPU use is structured to support that. When you're just starting out on your first room, you really can't overuse CPU unless you're doing something horribly wrong. Then as the game progresses and your influence grows, you start (gently, at first) running into things that are taking more CPU time than they should, but even the most basic of optimizations gives you enough headroom to keep growing. Then you look at the top players, and CPU optimization is almost all they think about. They're coming up with creative ways to stretch every last bit of usefulness they can out of it.
I also appreciate the way the base library is organized. The obvious, first level stuff that you want to do with creeps is readily accessible. There's quite a bit of depth to the available information, but it also leaves what I assume are some intentional gaps, things that technically work, but might not provide things in exactly the way a developer would want them. This is, after all, a programming game. I'm sure many of us have added quite a bit to the core objects to support our goals.
Finally: a shout out to power banks. I know there's a discussion about whether power creeps make sense in the game balance, but mechanically the concept of power banks is brilliant from a game design perspective. It provides an energy sink for players with more energy than they can spend, and it provides an incentive for otherwise peaceful players to implement code that could be very easily converted to be useful in assaulting remote rooms. Many players would never develop code for this without such an incentive.
Your team and efforts are much appreciated. Looking forward to more!
Thanks! Swimming in more power banks than I can handle now.
+1 to this. I want to pick up a few more central rooms, but not just so that I can have access to power banks.