I don't want default notifications. A road is under attack? I don't care. A creep is under attack? Also don't care. Controller about to level down? Don't care. Controller levelled up? You guessed it, don't care!

What I do care about is custom notifications. I can reimplement any default notification using these. The trouble is they get drowned out by the default notifications. This basically makes the notification system fairly useless. I can disable a lot of the default notifications but not all - eg. I can't disable "road under attack".

I kinda also care about error notifications, so a separate checkbox for them would be nice.

Yeah, that's my bot, sorry!

The reason this area is so open and empty is because my bot attempts to (and usually succeeds in) killing anyone who tries to claim rooms here.

As for what you can do? Probably nothing. The nuke is probably overkill but I don't yet trust my code to distinguish between a high level player trying to grab my nice open space and a new player like yourself. It'll send attackers once your safe mode is up and probably kill you before it lands.

Fortunately the rest of the world tends to be much more peaceful so probably best bet is to respawn elsewhere further away.

@artch Yes - that'd be great!

Someone on slack pointed out the console.addVisual function, which would help a bit if it was "official" as it would allow doing as you suggest: you'd be able to save the object on heap and just pay serialize cost. Would be even nicer if we could just pass a serialized version ourselves

Without that you've got to recreate the object every tick. That's going to be expensive using the API calls for extensive visuals.

It would be nice if we could pass a persistence option to visuals. Rather than having to draw stuff every tick, an option that says "draw the same thing for the next x ticks" would be great. It'd allow much more complex visuals without overhead.

For example: I could draw a status of every room on the map based on scouting info. It'd be great for me to see what's going on at a glance. I'm probably not going to do it, and certainly not always on, because it's just too much CPU to run every tick. On the other hand, the data does not go stale quickly. Updating it once every 100 ticks would be more than sufficient for me.

I'm not sure how visuals are implemented. The server tells the client something. I don't think it'd be too hard for the client to track persistence - wouldn't add server load and the 100kb limit could still be per-tick (ie. persisting visuals don't count).

@zpike said in Discussion: long-range logistics revamp:

It is already very hard to quickly come back after a wipe and transfering all my energy through terminals to a focus room is required to speed that up at all. Using a leg force of transports is a huge CPU and Spawner hit for a new player and is thus not much of a solution.

You can do RCL 7->RCL8 in under 50k ticks with without power creeps or external shipments. Seems fast enough to me. Remember batteries are a thing and while factory cooldown means you can only get 50 energy/tick from them, that only uses 5/50 of the terminal throughput, which gives you 95 energy/tick coming through total. Combined with upgrade boosts (going to use some terminal capacity, but 2x on energy) and local energy harvesting and 50k ticks is very acheivable.

You can move a lot of energy with creeps without using much CPU. I think it's a fine solution. If you're within about 5 rooms (call it 250 tiles) unboosted haulerers can move >15 energy/tick/CPU. Seems reasonable to me. I agree, you can be limited by spawn time.

Also, warp containers.

Also, I think most respawning players will have many rooms to level concurrently and there's no real need to push one at a time. Unless you're spending hundreds of credits/tick on energy you're not going to be hitting these limits.

I'm not trying to say it won't be a restriction, but I am saying that it isn't exactly slow to level rooms.

Any updates?

I guess for current tokens they'd have to away everyone Math.ceil(currentTokens / 10) "new" tokens as people have paid for these.

+1. Can't really see any drawback.

While we're on the topic of labs, they really should optimize the darn engine so that it short-circuits in checkStructureAgainstController for RCL 8 rooms. Lab intents can cost 0.4 CPU rather than the standard 0.2 because of this.

@zortea said in On the defensive power of strongholds:

@tigga said in On the defensive power of strongholds:

They always did. I don't see anything in the souce to suggest this has changed recently.

@tigga said in Right now it is very easy to take down a L5 stronghold with two nukes and loot 3/4 containers.:

Then I don't understand how you can clear a L5 stronghold with 2 nukes like you said a few months ago. Right now the repairer will fully repair its entire damage. Do you mean just use the nuke as a distraction for the repairer?

~9 months ago there was a change where repairers went from bredth-first unproritized repairs to depth-first prioritzed repairs. That was an improvement, but really just means you need more nukes. Nukes are cheap.