I don't have a problem that needs fixing, but I had a similar experience.

I had a graffiti that just did not want to transfer to steam: G-01090:MO:A. It would disappear from my inventory, but next time I refreshed the pixels page it would be back. I finally succeeded in transferring it a few days ago.

The 15 day hold is only for steamid not using steamguard for the last 7 days (at least that is how I read the hold notifications). So in about a week I should be able to trade on steam instantaneously.

Continual boosting has pretty tight margins. An example: saving a single CARRY part usually saves you 75 energy. But it costs 20 energy to boost. means you need to acquire 30 KH for less than the cost of 55 energy. Not impossible but tight.

Boosting your Mineral harvesters is usually a good investment, (unless you're mining Oxygen )

It think the math works out for PowerBank and Deposit farming since they're high value and require large creeps.

Otherwise most non-military boosting becomes a play to reduce intents buy being slightly less credit efficient. Which opens up a whole new layer calculations to play with (more remotes using boosts vs fewer efficient remotes, etc).

I think your math on Oxygen is correct. You've found one of the unexpected gotcha's of Screeps market. The energy price + the energy cost of dealing on the market can create some odd situations (You can make more credits by not mining Oxygen and instead selling more energy on the market).

I think currently the easiest way to make credits is increase your remote mining + energy income. If you've got the GCL pick up an X room since X seems to keep a consistently high price.

Is a terminal worth it; Yes. It's a fun aspect of the game. Many peoples' market code is rudimentary (stupid). So it's sorta like the stock market on easy mode.

Picking your creeps decoration from Memory isn't a good idea, if other players can see it. Besides the performance nightmare of accessing Memory every tick of every player drawn on the screen, you also have the issue that now there is pathway in the server that allows other players to peek at your memory, don't cross the streams Ray.

The core idea isn't bad though, something akin to Creep.say that you can set to the id of the decoration you want applied.

I think you can pull a creep from behind! (AKA push).

The _add-fatigue routine loops to the head of the pulled creeps and adds the fatigue there, as expected.

Slightly unexpected the MOVE part decrease of fatigue via the same _add-fatigue routine with a negative value.

That means a creep can "pull" a train from any position.

A couple caveats:

• The head of the train must have a MOVE part.
• It's more intent efficient to have the all the MOVE parts at the head of the train if the train has too much fatigue to move every tick. You need to issue all the pull/move intents every tick to keep the rest of the train helping with the heads fatigue.
• Edge traversal.

@Smokeman good point about the steam marketplace. I rarely use it and completely forgot about it.

I'm less sure about steam economics. I'd rather pay the devs $10 for some decoration than pay another player $10 on market place and hope some of that makes it back to the devs. But it's really up to the devs to make sure the steam marketplace transactions pay off.

I think the cost of targeted pixelization is too high.

A targeted pixelization is 5000 pixels for a 100% chance of getting a particular theme.

Grinding 5000 pixels gives you 46 regular pixelizations. Grinding for a particular theme and you have (6/7)^46 chance of getting the theme you want. That's 99.927% of the time!

I did a bit of grinding to get a creep decoration to play with on PTR. I found the process overly grindy. I could make it little less grindy by pixelizating a large batch, then converting back. But for the more rare decoration types it's grueling.

I'd much prefer a way pay a premium to pick the decoration type like we can with the theme.

Expanding from the current formula. The price appears to be 500/exclusivity * 1.43.

So randomly pixelizating a single theme costs 500/(1/7) * 1.43 = 5000.

Or pixelizating any theme except one: 500/(6/7) * 1.43 = 834.

Assuming creep decorations are 1 in a 100, pixelizating a creep decoration: 500/(1/100) * 1.43 = 71500.

Or a fire themed creep decoration: 500/(1/7 * 1/100) * 1.43 = 500500.

The advice for beginners is to keep your cost values low.

For advanced pathing anything "works". For efficiency you should set your heuristicWeight to at least the minimum tile weight. The default heruristic of 1.2 is assuming roads 1, plains 2. But heuristic 6, road 5, plain 10, swamp 50, is just as efficient and returns the same path. The low values problem is because PathFinder caps the heuristic at 9 so heuristic 24, road 20, plain 40, swamp 200, should be just as efficient but it costs more.