Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[Martin]

Yeah, I've taken to saving around the 10th of the last month. Usually the crash occurs between the 12th and the 17th, so I'm reasonably sure it's invaders, but it's quite difficult to get past the crash once it occurs.

[parlor_tricks]

Wow thank God/Armok that you posted. I was toying with the idea of bumping this again, but having someone putting the cenobite necroposter image up again has been holding me back. .


Long live morul.

Incidentally, when you did morul, did you have to do shearing?

I'm trying the challenge for 2012 and good god, some skills are a huge PITA.

Shearing, and soap making and so on.

[Martin]

Long live morul.

Incidentally, when you did morul, did you have to do shearing?

I'm trying the challenge for 2012 and good god, some skills are a huge PITA.

Shearing, and soap making and so on.

No, there was no shearing then. Shearing wouldn't have been that hard though. The hardest stretch really was trapping and milking. Trapping because there was no way to modify alerts to prevent them from pause/centering the game, so every hit, every stolen bait, was a pause/center alert. 600 successful traps were needed for legendary, and roughly ⅓ were stolen bait, so probably 1000 pause/centers. The milking was brutal because you could only milk maggots and Morul liked eating the maggots, so you had to be extremely proactive with forbidding, etc. There were barely enough maggots to get through the milking and you could only milk then once per season, and I'd inevitably lose half a dozen or more to him eating them because I wasn't fast enough at forbidding them. So, the hardest stuff for me were the things that I couldn't put on repeat and then go eat dinner and come back to have them done, half done, etc. Aside from Moruls's tasks and building out the fortress (which I did just to keep it interesting) the rest of the fortress was effectively automated. There was decades of food and drink, clothing wasn't an issue in that version, so if it ran for a few years unattended while Morul fished, no big whoop.


So yeah, lots of skills took a long time (like shearing would), but they were relatively hands-off, which didn't bother me so much. Part of the challenge is working out an efficient way to do the skill. How do you arrange pastures and workshops and stockpiles in such a way that Morul can do the job as quickly as possible - knowing that you have a fortress of haulers and helpers? I didn't know how long all of the skills would take in total and was worried that Morul would die of old age before I finished. Even in the current version with lots of additional skills, I don't think that's a serious concern. I'd be a lot more relaxed about making sure he's skilling every moment of every day.

[parlor_tricks]

Hmmm.

In this version I think I snagged milking early - remembering your experience. Once that was done I could butcher all the cows for increased FPS relief.

Haven't started trapping, lets see how it goes.

Shearing and beekeeping are a few deviations away from maggot milking, without the eating of llamas in this case.

They are on a seasonal basis, and take forever to level up, so you can't execute a forced march for those skills. And if you interleave it, then he will run across the map to pick up some unfortunate creature and shear it, or get stuck in a find creature->lose creature->booze! cycle.

Attribute gain is also pretty slow. I have some 10-15? legendary skills on him, and he isn't super strong or super agile even yet.


The tedium of watching him build soap forever gets to you. As a result the fort has slowly ended up terraforming the entire region. And every dwarf is also being forced through regimens of attribute improvement.

[Martin]

They are on a seasonal basis, and take forever to level up, so you can't execute a forced march for those skills. And if you interleave it, then he will run across the map to pick up some unfortunate creature and shear it, or get stuck in a find creature->lose creature->booze! cycle.

Go for the animal that can be penned in the smallest space without starving (sheep?) and put the workshop dead center. On the edge, put a hut with bed, booze, food, and your beehives next to that. Set a burrow and assign him. Put a double door entering (so you need to pass 2 doors to enter) to keep the animals from escaping. Keeps him from roaming, and minimizes walking distance for skills. If he still idles, stick a butcher shop in there as well and have him butcher all of the males but one and butcher the offspring. That should skill him up reasonably quickly as the max population - 1 females will all get pregnant simultaneously and shoot out a zillion offspring since the animal max won't get hit until the first set is born, and by then all the others will be pregnant. That should keep him plenty busy. Oh, and shear the offspring before you make gyros out of them.

Attribute gain is also pretty slow. I have some 10-15? legendary skills on him, and he isn't super strong or super agile even yet.

Yeah, that's changed a lot since Morul's version since you don't get the across the board automatic skill gains from making socks and taking inventory like you used to. But the attribute gains really didn't come in big until I was throwing 100 steel clad supersized orcs at him so he'd be launching dozens of counterattacks every tick. Getting that balance was really key. They needed to be strong enough that he wouldn't just bisect them all in one shot, but not so strong that he'd exhaust himself and fall unconscious.

[parlor_tricks]

Hmmm.

Hmmmmm.


I may have to build a seperate zoo enclosure for this. Or not - I have 2 floor wide tree farms I clear cut recently. I could convert that.

Ah crap, the bees.

I'll have to make it above ground. I guess I could do it, by building a large walled enclosure, and setting up a grand home for him.

I'd lose my doctor though (I have him switch out for opportunisitic heals). Still, I guess getting rid of the time consuming skills at one go could count as a fair trade off. I might be able to get wax works and tanning too.

Vis-A-Vis animals, I thought alpacas were the most suitable. (argh that reminds me I need to get cheese making up ><)

After checking - yes its sheep.

Mass butchery coming up. (and hopefully some more fps)

Hopefully no flying critters will come through.

I am using a few things to make it manageable - Speed 0 and reveal. Along with autodump to handle moving stone post .11.


-----

For weapon skills and combat skills, I am considering how I am I will manage teacher and student skills. I think if I got one of my dwarves to super dwarf status and made him a weapon lord, he may be able to provide 1 on 1 instruction to "The One".

If I can't figure that out though, I will mod in super orcs/goblins.

[Martin]

Autumn 597


We hadn't seen a caravan in years due to the constant sieges, so we decided to clear the enemy from around the fortress. Hundreds of bodies and body parts had been raised and were surrounding the fortress. They grew so numerous that a year ago we counted 11 goblin ambushes in one season, all of which were cut down by the evil creatures - who then later came back to life themselves. Fortunately this attention on us has provided much relief to other mountainhomes, but we will need to remain diligent and remove these body parts promptly after attacks.


Thankfully, a caravan has broken through from the mountainhomes. Along with this message we've sent instructions to begin sending migrants. We're not prepared for full habitation, but it's time to get started. Getting basic production up and running will free us to finish the remainder of the fortress. We've got 28 industries ready to go and will start preparing the rest now. We believe we've built a safe system for receiving migrants and we'll put it to the test.

Spoiler (click to show/hide)

The map is up now. I'll add more points of interest to define each bunker. Morul is outside just below the approach to the core. He just cleaned out some zombies.


Level 1 is mostly done. It's food and plant related industries - leather, cloth, dying, bone carving, etc. The pottery and glassmaking bunkers are incomplete as I was planning on having them span both levels and will need a collapse from the surface to get sand/clay down to them. The founders/guards have no proper living area yet (yeah - they've been flopping in the bunkers for now) so that will need to be dug out and prepared and that'll come next. The surface defenses need a lot of work, but that can wait. Level 2 will have the wood and stone industries. The miner will have an obsidian farm. I need to lay those out yet, but it'll look much like level 1. The caverns will have wood, fishing, plant gathering, and a few others. I'll also finish up the cistern and get water into the level 1 well system and get those pumps built. One of the rail lines is set up and tested. Yes, you can put 60+ stops on a rail line. I'll add the next 3 soon. That may get expanded further as I may have to put multiple carts on each line to keep the throughput high enough. I'm worried that I can get food/booze distributed fast enough to reach the folks at the end of the line. Each additional cart requires defining a new hauling route, which is a bit annoying, but also provides enough granularity to the system to make sure you can load/unload the right stuff at each stop. Collisions on the line are a concern. I don't want my guys getting splattered by carts.

[Mapleguy555]

PTW

This is interesting. Very interesting.

[parlor_tricks]

Autumn 597


Level 1 is mostly done. It's food and plant related industries - leather, cloth, dying, bone carving, etc. The pottery and glassmaking bunkers are incomplete as I was planning on having them span both levels and will need a collapse from the surface to get sand/clay down to them. The founders/guards have no proper living area yet (yeah - they've been flopping in the bunkers for now) so that will need to be dug out and prepared and that'll come next. The surface defenses need a lot of work, but that can wait. Level 2 will have the wood and stone industries. The miner will have an obsidian farm. I need to lay those out yet, but it'll look much like level 1. The caverns will have wood, fishing, plant gathering, and a few others. I'll also finish up the cistern and get water into the level 1 well system and get those pumps built. One of the rail lines is set up and tested. Yes, you can put 60+ stops on a rail line. I'll add the next 3 soon. That may get expanded further as I may have to put multiple carts on each line to keep the throughput high enough. I'm worried that I can get food/booze distributed fast enough to reach the folks at the end of the line. Each additional cart requires defining a new hauling route, which is a bit annoying, but also provides enough granularity to the system to make sure you can load/unload the right stuff at each stop. Collisions on the line are a concern. I don't want my guys getting splattered by carts.[/spoiler]

Hmm, this is magnitudes of complexity higher than morul for sure. Probably approaching a dorf computing project.

Taking a look at the map now.

edit: saw the map.

[Martin]

Hmm, this is magnitudes of complexity higher than morul for sure. Probably approaching a dorf computing project.

Taking a look at the map now.

Indeed it is much more complex. Safe to say that it's more complex than I had anticipated at the start.


I don't know that it's more complex than a computing project - it's messier, though. Essentially it is a computing project but one where the dwarves are part of the machine, and we just don't know how that's going to work with this kind of isolation. I think it's going to work quite well, because pathing mostly becomes a non-issue and job assignment is quite deterministic. The challenge is all setup, so it's kind of the inverse of everything we're accustomed to doing.


It's also really odd for me to build this much of a fortress and I haven't even pierced the first cavern, and I'm doing all these plans for magma and I have no clue where I'm going to find that. I know it's down there somewhere, but I hope I don't need to pump it up from the bottom - not that I haven't done it before. I built enough power to cover that eventuality at least.

[parlor_tricks]

Edit: Thats... truly a staggering number of mechanisms and switches. Have you considered opening up a book on old railway route design?

I attempted following the route of a theoretical cart operating during full activation of the project, and (unsurprisingly), got lost.

Am I correct in assuming that final minecart operations are to be automatic, and not involve a station manager?

Also you seem to have a dead g (goat/gobbo?) on level 76. Its right of the central column, in the Goat herder chambers, above the first mechanism/gear box. Sorry for the lack of better directions. Hmm. seems like a dead S, to the north on that level as well. Its past the first set of chambers north of the central chamber, and to the right of the horizontal axle.

Is it just me, or are the colonies/silos/(insert appropriate sci-fi name) not sealed/air tight/isolated/quarantined/(insert appropriate term) at the moment? It looked like that someone could cross over.

I'll look further and see if I can figure out anything else.

(If its not too much, is it possible to see it without a graphics pack?)

[parlor_tricks]

Indeed it is much more complex. Safe to say that it's more complex than I had anticipated at the start.


I don't know that it's more complex than a computing project - it's messier, though. Essentially it is a computing project but one where the dwarves are part of the machine, and we just don't know how that's going to work with this kind of isolation. I think it's going to work quite well, because pathing mostly becomes a non-issue and job assignment is quite deterministic. The challenge is all setup, so it's kind of the inverse of everything we're accustomed to doing.


It's also really odd for me to build this much of a fortress and I haven't even pierced the first cavern, and I'm doing all these plans for magma and I have no clue where I'm going to find that. I know it's down there somewhere, but I hope I don't need to pump it up from the bottom - not that I haven't done it before. I built enough power to cover that eventuality at least.

I see it as similar - this is essentially a computation project, and in a way I assume its a finite state machine. Inputs for one state get used as inputs for another and everything finally goes to the depot.

That said your major issue, as I see it, is uncertainty.

When an orderly/deterministic system interfaces with chaotic/random/non-deterministic inputs leads to all sorts of interesting edge cases/outliers/break downs.

Thats what I foresee as your major bottle neck. I was going to post this above statement, but thought it would be to presumptive/arrogant to write it out. .

So my question is, do you have any redundancy planned?

What if your colonies get over populated? Do you have an alternative way to get materiel transported through the system? Perhaps a fail safe quantum factory unit?

[Martin]

Edit: Thats... truly a staggering number of mechanisms and switches. Have you considered opening up a book on old railway route design?

Yeah, 2400 mechanisms so far. Maybe 100-200 in stockpiles. Lots of legendary mechanics…


I'm pretty familiar with railway design - did a lot of model railroading as a kid with my grandfather.

Am I correct in assuming that final minecart operations are to be automatic, and not involve a station manager?

The general route design is quite basic. The main loops are just that - loops. Each stop brings the cart to a stop, there will be two depart conditions - one to unload whatever is desired until the appropriate stockpile(s) are full, and a second based on time so that the carts aren't too delayed. The dwarf rides the cart to the 2nd stop at the bunker and then has similar rules to load the cart and push it on its way.


Unfortunately the system doesn't work unless every bunker is occupied, which is why so much planning was needed. I considered (and am considering) a fallback to this, which would switch the stops off for unoccupied bunkers and add a powered roller to ensure the cart always has enough velocity to clear the pits.

Also you seem to have a dead g (goat/gobbo?) on level 76. Its right of the central column, in the Goat herder chambers, above the first mechanism/gear box. Sorry for the lack of better directions. Hmm. seems like a dead S, to the north on that level as well. Its past the first set of chambers north of the central chamber, and to the right of the horizontal axle.

Yeah, lots of dead livestock around. This version doesn't automatically assign offspring to pastures and they sometimes get out of the pen and starve.

Is it just me, or are the colonies/silos/(insert appropriate sci-fi name) not sealed/air tight/isolated/quarantined/(insert appropriate term) at the moment? It looked like that someone could cross over.

Bunkers. Some are sealed up, others aren't. I'm filling in mood stocks still. Cloth and gem bins hold WAY more than they used to, and at least half of the bunkers don't have any of one or the other - particularly silk cloth. Once immigrants arrive, I'll seal everything up so they can't interact with the guards. They can get the mood items later off the trade line (hopefully). Bones is another item I haven't distributed. I suspect that may go badly once moods hit. We'll see.

(If its not too much, is it possible to see it without a graphics pack?)

When I update DF next time I'll try and remember to put a a plan map up.

[Martin]

So my question is, do you have any redundancy planned?

What if your colonies get over populated? Do you have an alternative way to get materiel transported through the system? Perhaps a fail safe quantum factory unit?

No real redundancy planned. The central core and the trade depot provide some degree of redundancy. I plan on building up massive stockpiles of roasts/booze so the founders can dump that onto the line if the workshops fall behind. The farm is probably the biggest potential bottleneck (other than the time it takes to push the stuff around given limited cart capacities), but fortunately I can have the kids help harvest. Two brewers and cooks should be able to keep up with 200 dwarves with nothing else really to do. Each bunker will have a well as a fallback. But yeah, there's certainly the possibility everyone will wind up starving to death. That's all part of the experiment.


Overpopulation really shouldn't be a big issue. The baby popcount is pretty low. If they get well distributed around the fortress, we'll be fine. If it's a handful of bunkers turning out all the kids, well… And seeing if randomly matched pairs of dwarves marry is another experiment here. They'll never interact with anyone else, so it should happen. I think we can force that in dfhack now, though.


My hope isn't to have to micromanage too much. There will be a manager with nothing else to do but manage and eat. For many things I can just queue up a gazillion jobs. For others (where one workshop can do multiple jobs on multiple skills - like the farmers workshop) I can't because there's no way to make sure the milkers workshop gets the milking jobs and the threshers gets the threshing jobs. So, well see.

[Terror_Incognito]

I am really glad this thread is still going. I am keeping a close eye on Martin's *Science* (I don't know how to do Masterwork symbols) for my own Minecarting.