Print

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

[Kipi]

For example, if you had six dwarves in squad and you had minimum number set to 10 you could see one dwarf arranging dodge demonstration and five dwarves waiting for the demonstration to begin, which never happened.

Actually, that part is a myth...

Is it?

Why then I remember several occasions back in 0.31.01 at least where, if the minimum number set by schedules wasn't filled the demonstration, never begun?

I admit, I may remember something wrong here.

[Garath]

I've had a squad do 2 different demonstrations, 3 attending each, 2 sparring, one on break and 1 watching the sparring, so it's uncommon that 1 demonstration takes up a whole 6 dwarf squad who are waiting for one to come off break

that could be the myth?

[Quietust]

I've just discovered a very interesting myth, and it's persisted for years on the wiki despite being totally false.
The myth is that "for every unit of material size an item has, 1/10th of a bar of that item's metal type will be recovered".

This has never been true.

I first discovered this while studying a disassembly of version 0.23.130.23a - I found the code for determining melt yields, examined it, and found that all of the melting yields were constant, based solely on item type and stack size. Curious to see if it was still the case, I found the same code in 0.34.11 and found that they did respect material size, but they returned 3/10 instead of 1/10 per size unit.

Then I started looking at older versions to see when it changed. What I found shocked me - aside from adding new item types, the code for determining item melt yields has not changed a single bit between version 0.23.130.23a and version 0.34.05, which is when it changed to the current behavior.

I could not track the exact origin of this information, but it was first added to the 40d page when a user imported it from the archive wiki where it pertained to 0.23.130.23a (where it was obviously false) and nobody ever verified it, and the same information has been blindly carried forward to every new version released since. This is absolutely shocking to me, and it means that a vast amount of information on the wiki can possibly no longer be trusted.

[Finn]

Well, to be fair to the wiki, it would be hard for it to be accurate since it's composed by users based on observed behavior.  In a game this complicated it can be difficult to tell cause from effect, or discern between causality and corolation. 

In other words, I think the wiki contributors have done a tremendous job working with what they have. 

That being said, nice catch!  I presume you changed it?

[Quietust]

Well, to be fair to the wiki, it would be hard for it to be accurate since it's composed by users based on observed behavior.  In a game this complicated it can be difficult to tell cause from effect, or discern between causality and corolation. 

Actually testing this would have been trivial - the information about melt yields being accumulated in 0.1 bar increments (at each individual smelter) was correct, so melting 10 of the same item would have shown the exact yields and revealed that none of them depended on material size.

The wiki has indeed been updated.

[Sutremaine]

I've just discovered a very interesting myth, and it's persisted for years on the wiki despite being totally false.
The myth is that "for every unit of material size an item has, 1/10th of a bar of that item's metal type will be recovered".

This has never been true.

...Huh. I think I encountered this once (trying to squeak an extra bar of steel out of the first autumn caravan when low on trade goods), but after some head-scratching put it down to a miscalculation on my part.

[slink]

Here is the man himself on the topic, in 2006.  It's the oldest mention I could find of melt yields, although one of the posts in this thread implies that there was an older post by him.

Yeah, it works in fractions.  There are no random calls within the procedure.  A goblet should produce 20% of a bar.  Each smelter stores the fractions separately by material.

As goblet size was said to be 2, this would have implied 1/10th bar per item-size to the readers.

http://www.bay12forums.com/smf/index.php?topic=8269.0

By 2008, the 1/10th bar per item-size was apparently already in the DFWiki.

http://www.bay12forums.com/smf/index.php?topic=23157.msg254280#msg254280

[Garath]

great catch

[Azure]

The myth about demonstrations not working is mostly from newbs flubbing it up.

If you have:

• At least 3+ teachers so that skilled dwarves are teaching.
• All the same weapon type

You'll get a couple skill ranks a year. Which is reasonably matched to running around stabbing animals/prisoners. As you get more teachers and they get more skilled you'll see quicker turnover until you're a training factory sending guards and supplies to a fellow brobold fortress in later versions.

[PainRack]

So, how does one develop teachers then?

[Tirion]

So, how does one develop teachers then?

Keeping them active, alive and out of action.

[PainRack]

Is there a way to train specific teachers then? Furthermore.... at a risk of being a noob again, aren't teachers limited to teaching skills up to their level?

[Sutremaine]

Is there a way to train specific teachers then? Furthermore.... at a risk of being a noob again, aren't teachers limited to teaching skills up to their level?

You train specific teachers by making those specific dwarves the ones who always stay in the training squad. Teachers will teach up to their level, and then switch to sparring, which allows them to gain more levels which they can then teach to the next set of newbies.

[gestahl]

Then I started looking at older versions to see when it changed. What I found shocked me - aside from adding new item types, the code for determining item melt yields has not changed a single bit between version 0.23.130.23a and version 0.34.05, which is when it changed to the current behavior.

I wonder what the exceptions were, since I remember that in 0.31.25 melt yields were roughly in line with materialsize / 10.

It's funny, I had checked the wiki before dusting off my plans for my bolt splitter since the 34.05 / 06 ment I was going to need quite a bit more metal, find someone seeing some strange behavior, and after a little testing turns out I don't need the bolt splitter after all  .
After all, chaining defenceless goblins to the wall and having my dwarves beat on them with wooden sticks for ages is a much better use for them then just getting shot at but never hit.

[miauw62]

You must laboriously assign every single dwarf a bedroom.

It DOES sastify OCD tough.