Bay 12 Games Forum

Regarding the stockflow plugin, perhaps it would be easier to make it work more like workflow that's tied to a stockpile (having read a separate thread about the difficulty of having stockpile generate jobs in connected workshops)?
Instead of using job manager, stockpile will suspend/resume relevant job in the "Take from" workshops. Then each stockpile can "borrow" the workflow's interface to set up a set number/stacks of items to maintain in the stockpile.

Instead of having stockpile "create" the job, require user to create a job on repeat in the linked workshop. The stockflow UI can indicate if it can detect a relevant job it can control in the linked workshops.

My couple of cents.
If you already have a programming language you're comfortable in, use that. Unless you're trying for something on the scale of DF, whatever performance penalty in your language of choice would likely not matter.
Personally, I'm partial to C/C++ since that's the language I'm deeply familiar with. If you have no coding experiences, I suggest Python as starting point. If you need a bit more performance, try learning Java. Both languages should be more than sufficient for most roguelike game (again, unless you're trying to code DF clone or something similar).

I would argue that if you REALLY want to learn as much as you can about programming, try C/C++ at least once (whether it being the first language or Java as a stepping stone). If nothing else, it makes you appreciate features in Java like garbage collection.

Hm... I guess this depends on the fact that it takes time for items to heat up based on the specific heat.
Might be a problem for people using DFHack to reduce FPS load due temperature updates by make items heat-up and cool-down faster.

I wonder if magma misting would work equally as well (set the goblin and cloth on fire, wait till they all burn to death).

I generally have my fortress at about 90% duty level (120 dwarfs, about 10 idle at any time). Basically I have nearly ALL labor enabled for a couple dozen "Peons" (like Girlinhat), excluding labors such as miner/woodcutter (conflicting tool setup), hunter (don't want hunters), planters (don't want a lot of small stack plant returns).
Since I'm slightly OCD in keeping my fort clean of loose boulders. So I have a mine-cart based quantum stockpile, which has a feeder stockpile that takes directly from 3 other feeder stone stockpile (this way, I have 9 haulers w/ wheelbarrow hauling stone at all time). I also have a set of siege training + pump operator training room for all peons. If there's a need to haul, I just deactivate the siege engines and pumps to free them up for hauling. Currently working on getting a auto-swimming training system setup http://www.bay12forums.com/smf/index.php?topic=124729.0 for those sweet, sweet agility attribute gain.

Have you tried siege engines?
Setup a siege engine (preferably ballista, but catapult might work too) about 30+ tiles away (or however far away enough that civilians don't run, also for catapult, you have about 30 tiles minimum range required) with LOS through fortification. Start pounding the FB with rocks or ballista bolts.

This way, you get some siege operator practice too!

I'm trying to imagine what it would look like in real life if 5 minecarts, all loaded with green glass spikes (with sand, infinite spikes), launches the spikes in parabolic arch at you.
Basically, what would 1500+ green glass spikes coming at you look like...
And how much "skewering" would they do to a goblin siege.

Is it possible for the Workflow plugin to trigger a job based on input count?
Basically have an option for a "push" based workflow to compliment the current "pull" based workflow.
For example:
A Mill Rocknut task will auto-resume if the number of rocknuts (input) is > 120, and auto-suspend if its < 100.
A Dye Thread task will auto-resume if there's > 10 powder of dye, and auto-suspend if its =0.
This would be helpful in automating, say, soap production with plant goods in the first case, and dyed thread/cloth production in the second case.

I'm surprised that this wasn't suggested before.
My idea here is that you can designate a minecart route as passenger route.
When dwarf tries to reach its destination through pathfinding (assuming A* pathfinding algorithm). Once it encounter a stop that part of a desginate Minecart route, it will take that route into consideration based on an assumed speed on the route (say, assume the cart always travels at 0.25 tiles/tick).
For multiple dwarfs using the same stop, each successful path using the stop will increase that stop's pathing cost by a certain amount (so you don't get all your fortress dwarf waiting at the stop because the route using that stop is the "shortest" for them all).

So, you can setup a kind of minecart network that allow you to shuttle dwarfs over long distance automatically (they'll take a cart to get there if it's a shorter route, or the only route).