Okay, so now that 1.0.21 is out of the way it’s time to start work on 1.0.22. If clothes were my main focus in the last patch (and largely they were, although not entirely intentionally – the scale of the task just ended up being much greater than I’d thought) then crafting is the focus for this one. I’ve long felt that our crafting system is so much better than the scale of schematics really demonstrate, and it’s time to fix that by throwing in a metric ton of new things to create. It really does feel good to spend time bulking out some of the sparser parts of the game.

We took crafting seriously in the beginning of Epitaph – it was supposed to be the primary way through which you built your equipment as you went through the game. Scavenging was supposed to be the unreliable, ‘maybe I’ll find what I’m looking for’ system – crafting was the ‘I know what I want, now I’m going to make it’ part. However, because of the relative paucity of items for scavenging, finding what you wanted was reasonably straightforward. It just took time. Due to the relative paucity of crafting schematics, you couldn’t really make whatever you wanted – there just weren’t enough moving parts. That’s going to change in the next patch.

To give you an example of what I’d always wanted from the crafting system, let’s look at the task of making something like a dress. In most crafting systems you get a set menu of what you want (pretty black dress or pretty green dress), you gather the components and then you make the thing which ends up looking pretty much like every other thing that other people have made. Some crafting systems ‘fake’ it by letting you simply write your own description of a thing you’ve crafted, which gives you infinite variety but I’ve always felt isn’t quite in the spirit of the thing. Our crafting system on the other hand is intended for items to be built up in parts. A dress then might mean assembling three different things together – a blouse, a skirt and a belt. Skirts and belts are items you just stitch together from fabric, but a blouse is in itself a combination of several things – sleeves, necklines and – whatever the rest of a blouse is called. The act of making a dress then requires you to craft several things – a skirt, a belt, a pair of sleeves, a neck panel and a base. In this, you get to choose which specific kind of thing you want (cap sleeves, or an accordion skirt) along with the type of material and the colour. So let’s say we make the following:

* A black felt accordion skirt
* A white leather skinny belt
* A pair of blue satin cap sleeves
* A purple lace queen anne neck panel
* A red velvet jabon base

Now, I’m not saying this is a thing you’d like to genuinely wear, but it shows just how flexible the system is. You’d combine the sleeves, neck panel and base into a blouse that had a description something like ‘This is a jabon blouse, expertly constructed from high quality red velvet. It has a pair of blue satin cap sleeves. The neck is made from purple lace and forms a Queen Anne neckline’. That blouse is yours – you designed it, pretty much, and it can be worn at that point. When you combine it into the dress, what you end up with is:

‘This is a long dress with a red velvet jabon bodice and a black felt accordion skirt. The blouse of the dress is expertly constructed from high quality red velvet. It has a pair of blue satin cap sleeves. The neck is made from purple lace and forms a Queen Anne neckline’

The writing here is a little rough, but that’ll improve as the quality of the descriptions increases and the language generation of the items is polished. I have plans for a system that would smooth out some of the clunkier auto generated text in that respect… at a simple level it might change ‘deer meat’ to ‘venison’ any time the occasion calls for it. On more complex levels it would work to combine things so that instead of saying ‘The left hand side is red. The right hand side is red’ it would parse it as ‘Both the left and right hand sides are red’. Nothing is ever *finished* on Epitaph, and that extends down to how we present text to players.

The key thing though is that with enough raw parts you can mix together you end up with exactly the eye-wateringly awful garment that you wanted, without us having to code it directly. It provides a completely closed and coded crafting system with the potential to permit genuine creativity. That’s sufficiently novel that it might actually be entirely unique – I certainly know of no other system that would offer the same thing, but I’d be interested to hear if people have seen something like it.

This configurability too carries through to other categories of crafting item, such as weapons. A sword might be made of:

* A blade
* A hilt
* A cross guard
* A pommel

And in combining specific objects of these types, you end up with exactly the kind of weapon of which you’ve always dreamed. It’s a lovely system I think, and one I spent a lot of time developing. As I say though, the relative paucity of schematics has meant that your ability to improvise within the system has been limited and I’m hankering to change that – and as I did with the last patch, I’m doing it through the use of generators. We currently have 200 new schematics ready to roll, but that will be at least an order of magnitude more by the time we’re done – I’ve done a chunk of hand-held weapons, a few clothes, but I haven’t even started on armours, jewellery, recipes and the like. Remember too that the number of schematics doesn’t mean ‘number of items you can create’. Within the 150 or so weapon schematics, you can make about 80 different kinds of hatchet alone – stone curved blades and metal handles, or metal beveled blades with wood handles and so on. This is only a rough sketch too – just proof of concept to make sure the whole crafting flow works. Each new kind of variation means many more options for getting exactly the thing that is currently fashionable.

That raises an issue of scale with regards to schematics, though – currently these are either randomly researched or scavenged. The worse of both worlds really, in that you’d need to scavenge up all the patterns before you could actually go to the trouble of making the things you want. Scavenge will remain a system by which schematics can be found, but I think we’re also going to look at a few other mechanisms for putting them in the game. The research command is likely going to be overhauled too – no longer will it just give you access to unique schematics, but it’ll let you direct your study towards any schematic you don’t currently know. You won’t have to find everything, you’ll be able to explore for them through the command. I’m also likely to add a new command this patch that lets you consume scavenged schematics that you don’t want in exchange for a ‘buff’ on your next crafting attempt. I’m also mulling over automatically making schematics available to people once they have hit a certain bonus of the appropriate manipulation/theory skills – we’ll see though. Part of the problem there is that the number of theory skills we have don’t let that happen as coherently as I’d like, and I’m not a big fan of adding in skills just for the hell of it, especially when they don’t ground into any other game systems. It’d be neat for some kind of ‘Eastern Culture’ skill to influence when you learned to make Katanas and Obis, but I’m not sure how that skill would be meaningful outside the context of crafting. How would you actually use it in the game world? Those are questions that need to be resolved before we could go down that path, although it is a path that appeals. It’d be nice if some people specialised in making Oriental swords because of the way their skills went, while others were more interested in European maces. It would allow for specialism, and I’m a big fan of that. But that’s not enough of a reason to introduce a whole pile of skills that are otherwise irrelevant.

It’s always a roll of the dice with regards to how much development time I get over Christmas so I don’t want to commit to a patch date yet since we’re almost guaranteed to miss it anyway. But this is just one portion of what’s on the agenda for 1.0.22.