Posts made by PhiNotPi

I decided to send a creep into another player's room, but I have not been able to. I have checked that there is a clear path between the two rooms.

The following code is what works correctly for the room W3S1, which is NOT owned by anyone.

    if(creep.memory.roles[role] == 'voyage'){
        voyageCount++;
        var dest = 'W3S1';
        if(creep.room.name != dest){
            console.log(creep.room.findExitTo(dest));
            creep.moveTo(creep.pos.findClosest(creep.room.findExitTo(dest)));
        }
        else{
            creep.moveTo(creep.room.controller);
        }
    }

As soon as I change the destination to W2S1, the code gives the -2 "path not found" error code, despite the fact that there is an open connection between the rooms. The room W2S1 is another player's "starting" room.

Last week sometime, I saw a variety of messages appear in my console as well. This seems to be the same bug. It's definitely the logs of other players.

I'm sure you will receive an official answer sometime soon, but here is what I know. In order to start playing, you must receive the Early Preview unlock code by email. I assume you did not sponsor the Indiegogo project, so you must complete the application form in order to be selected.

Last week sometime, I sent the staff an email asking about the expected wait time for Early Preview acceptance. The response was that they will decide how many to accept based on server capabilities, and that they may start inviting new people 2 - 3 days after the initial Early Preview launch.

After you submit the application, you must wait to be accepted and receive the activation code to play in the MMO World mode.

From this unofficial wiki page, I know that there are several actions that conflict with each other. Sometimes I have creeps that attempt several of these actions at once, and I would like to know how certain actions receive priority.

Here are several questions about specific cases.

1) If my creep attempts two actions with the same body part, which is performed? Let's say that it attempts to attack two separate targets. Does the result depend on the most recent (successful) call, or what?

2) If my creep attempts two different types of actions with the same body part, what happens and what does it depend on? Examples are heal+rangedHeal or rangedAttack+rangedMassAttack.

3) If my creep attempts two conflicting actions with different body parts, what happens? Does the order of body parts matter? Examples are rangedAttack+rangedHeal and build+attack.

I think it would be useful to have a getAge() function that returns the number of ticks that have passed since that room was created. This can apply both to survival rooms (created every time a new game of survival starts) and to rooms in the MMO.

I think this would allow for scripts to easily recognize when they are in a new room, and thus are starting a new game. Since a script may not be running in any particular room when that room is created, the script cannot keep track of the room's proper age on its own. This is why I believe a builtin is required.

I think one type of body part that hasn't been covered yet is a DEFEND part, which would represent a creep's ability to dodge/block attacks. This part would actually decrease the amount of damage inflicted upon the creep. A regular TOUGH part, on the other hand, increases the creep's hits but doesn't decrease the amount of damage the creep receives. The DEFEND part would thus add a lot more protection than a TOUGH part, but at a higher cost.

The advantage of this part would be the creation of creeps that are able to withstand more attacks, but which are able to move faster than a creep weighed down by a bunch of TOUGH parts. A creep with a DEFEND part would also require less healing, as it would have less damage to be healed.

I envision creating a healer bot that is quick enough to move to where it's needed, but also able to dodge some enemy fire while healing other creeps. There would also be more protection for transport bots, which can't afford to be weighed down with a lot of TOUGH parts if they already have a lot of CARRY parts.

The DEFEND part would probably require an extension. I don't know how much they would cost.

Note: I am using the following pages for information sources:
http://support.screeps.com/hc/en-us/articles/203016342-ConstructionSite
http://support.screeps.com/hc/en-us/articles/203086021-Territory-control

Ramparts are currently planned to cost 2000 energy and have from 1000 to 7000 hits, depending on game mode / RCL. I believe that ramparts are too expensive for the limited protection that they provide, and that their high cost will discourage their use.

My proposal is to substantially decrease the cost of ramparts and substantially increase their hits at lower RCLs. Since most of my experience has been in survival mode, it would probably be wise to watch how Early Preview unfolds for a little while before making any changes. In Survival Mode at least, I do not think ramparts are currently a viable strategy at all.

I think a useful way to look at efficiency is in terms of hits/energy. A single rampart could add 1000-7000 hits of protection at a cost of 2000 energy, giving a hits/energy efficiency between 0.5 and 3.5. A regular TOUGH part, however, gives 100 hits at a cost of 20 energy, giving an efficiency of 5. Furthermore, ATTACK parts have an efficiency of 1.25 (higher than some levels of ramparts), in addition to their valuable offensive properties.

The 2000 energy to build a single rampart could instead be used to create 100 TOUGH parts (10000 hits) or 25 ATTACK parts (2500 hits with up to 750 hits/tick of offense). To surround a spawn or other object, it takes at least 8 ramparts, costing 16000 energy. Basically, it is a lot cheaper to build an army to defend a spawn than surround it with ramparts.

A more direct comparison is with regular constructed walls, with 6000 hits for 500 energy and an efficiency of 12. Walls are many times more efficient than ramparts for creating a barrier. Also, if you want to have a "gate" in the wall, it is cheaper to have a hole in the wall with some guards than use a rampart as a gate.

The point I'm trying to make is that other forms of defense are far cheaper and stronger than ramparts. In order to turn ramparts into a viable defense strategy, their cost effectiveness should be increased.

I think Blocky would help expand the game's audience if it were an option, but I don't think it's a good idea to require it as the only option.

A good reason to include Blocky as an option is that many potential players don't know JavaScript well enough to write an AI in it. Every programming competition has something called a "barrier to entry" a.k.a."How much work does it take to write something that's functional?". In the context of Screeps, a minimally competitive entry would need to be capable of mining energy, transporting it to the spawn, and fighting enemies. The lower the entry barrier, the more people are willing to try it. I think adding support for Blocky would make it easier for someone, without knowing much about JavaScript, to write a basic AI that works.

I don't think it would be a good idea to limit people to using only Blocky. From my perspective, as someone who develops their AI directly on the website, it's a lot easier to type JavaScript than drag-and-drop programs. Blocky slows down copy-pasters, but it slows down everyone else too, which can be a problem in a "real-time" game. Personally, I would never use Blocky to write my code.