I want to change the catalog viewing system. When browsing (in addition to the main filters and sorting options), I want it to be possible to switch between three main views - Sets, Miniatures and Other - for each company, game or faction.
Sets. These are primarily boxes of miniatures - the way they’re sold, what people buy in shops. Each set may be re-released many times with different packaging designs, but the contents remain the same, containing the same miniatures made from the same or different materials.
Miniatures. These are individual miniatures from sets, which may be included in different sets. If there are 10 of them in a set, then 10 miniatures are shown separately, rather than the set as a whole.
Everything else. These are books, cards, merch, dice, terrain and so on - not miniatures or models. But they are also sets, not individual items. This is so that these sets aren’t confused with the miniatures and don’t get in the way of viewing them. GW has a lot of supplementary items, which isn’t always necessary or important for viewing on miniature's collector's catalog.
When a users visits the company’s page and opens its product catalogue, they will see ‘Sets’, with the option to switch between viewing as main image the contents of the set or the front of the packaging. They can add the entire set to their collection at once – the miniatures from that set will also be added to their collection. After selecting all the sets, they can switch to the individual miniatures within the sets and select those miniatures that they already own separately, without the sets.
Overall, I think the system works well and will be useful to many people. Dividing all the sets into individual miniatures is time-consuming, but not difficult. Marking images for different miniatures or cropping different individual miniatures from a single image is entirely possible (with some additional automation), although it does take time. There are a few tricky issues that need to be resolved.
The first issue. What should be done about the miniatures’ IDs? After all, for the website, a single miniature will technically also be a material "set", just with a separate checkbox indicating that it is part of certain sets. For the ID of a single miniature, it’s perfectly feasible to use set_id_xxx, for example gw-90321243123-001. But if this miniature has been released in several sets, which of the set IDs should be used? After all, an earlier set might turn up later. Using the actual GW component number is not an option; it is not always known and adds unnecessary complexity. We’ll also have to add the miniature from sets containing just one miniature.
Second point. What should we do with multi-part miniatures? What if a set offers 10 different options for arms, legs and heads for each of the 15 miniatures? A set from which you can assemble 2, 3 or 4 specific, exact miniatures isn’t a problem - we simply create the variants. But if everyone can assemble whatever they like and the results will always differ - that’s where it gets tricky.
Does anyone have any thoughts on this? Please do share them.
The idea of somehow describing models separately from sets is a very good one.
However, I think these shouldn't be two completely independent systems (sets and models), but rather a new, expanded description block for sets. The idea is for this block to contain the necessary taxonomy to describe the set's contents.
On the one hand, this would be something similar to what they tried to do in the Lost Minis Wiki:
https://www.miniatures-workshop.com/lostminiswiki/index.php?title=File:C...
(see the Taxonomy section at the bottom of the page)
That is, we have a race, gender, weapon type, and clothing type. Here, we can use the existing division into universes on the site as a starting point and expand on it a bit. This would also solve the important problem of searching for similar sets and displaying them. By describing this set as Roman legionaries from the late Republic or orcs with swords, I can immediately identify all similar sets from other manufacturers.
On the other hand, this will help describe sets in terms of their composition and variability, i.e., not just as we do now, stating that there are 5 or 10 models, but that you can assemble one of these models or one of that model, or that there's only one option, or that we have 7 simple warriors and 3 that can be assembled as warriors or as a command group. In this section, you can then indicate, for example, that you don't have a complete set, or that you assembled two copies of the same set of 10 miniatures as a squad of 20 soldiers with a command group.
The result is a description through logical operations (AND, OR) of the contents of a set from a hierarchical taxonomy, for example: 7 (historical - person - ancient Rome - legionnaire - spear) AND 3 (historical - person - ancient Rome - legionnaire - spear) OR 3 historical - person - ancient Rome - legionnaire - command group).
I think this aspect of providing detailed descriptions of miniatures is slightly different, but also useful.
At the moment, the terms (tags) on the website are already linked to specific tag types – race, weapon, clothing, items, pose, etc. I haven’t shown this anywhere, but it’s very easy to convert into a table like this. Many years ago, I categorised around 30,000 tags into these categories. Synchronising miniatures with sets isn’t a big problem. It’s true that it makes more sense to display the terms on the specific miniature rather than on the set.
For me, the most important question right now remains: how to account for random miniatures from a pile of different parts and how to record the ID correctly. With the old GW miniatures, it’s simple – Goblin 1, Goblin 2, and so on. Our sets have their own IDs. We could simply assign a sequential or random number, but that wouldn’t look very neat in the URL....
Guys, what do you think about it?
For GW, it's enough to add two new set types: Bit and Sprue. They even have separate codes.
It's worth noting that in the old days of monopose miniatures, the bit could match the miniature, but then individual parts began appearing, so that, for example, a single blister pack could contain three bits: two miniatures and one arm for one of them.
For plastic sprues, detailing them down to individual parts is overkill, in my opinion.
This all a good idea, but also very ambitious.
When it comes to miniatures that map to extra sets, I think you already have a good system in place, you have some things where you can add extra fields. So what we could do is when making a minis only entry, give that the option to "also found in:" and we can put in a second product code.
Nowadays GW tends to package things by sprues, those would have far more reliable gw codes, but collecting sprue codes might be a bit overbearing. And then the very old GW material did not have a sprue code printed on it, only the more modern stuff does.
The best idea I can think of for trying to parse out variants is simply to use the instruction booklet. GW books, and many other manufacturers, list alternate builds as a separat-ish entry. So if the first miniature has two possible builts, it could be listed as 1a and 1b in the instruction book. Some of the Kill Team models can be assembled three different ways. This would take a LOT of work to do mind you. This information is not always readily available, but it would make for an incredibly organized system where people could keep track of what they built. You could almost get to the point where you could syphon some of this off and make a force builder system with that level of detail.
A lot of the larger vehicles kits often have multiple possible loadouts too. The Baneblade has like 4 different builds in one bax, Warlord Games' Panzer III I think has up to 3 variations in a single box. Gathering good pics of all the possible alternate builds will also be challenging. These are not always available on the manufacturer websites and it will be left up to the users to do this in many cases.
Some of the major advantages of this is that you can get rid of the product codes you list with a -1 when a set builds two different minis. This way you can have two minis map to one production box, which will be useful for collectors to figure out what's in a box. Just a few quick thoughts I had on this.
Good ideas, thank you!
If we treat miniatures as a separate material type (rather than a ‘miniature’ set), that will make things much easier for us. We’ll be able to link sets and miniatures and have them inherit various properties from one another. Many of the properties of sets aren’t needed or important for miniatures – for example, weight, price and many others; even a description isn’t really necessary, but we should leave the option open. On the other hand, it will be easier to distinguish between sets and miniatures.
We can leave the IDs of miniatures as random or sequential, or even link them to the sequential numbering of materials on the website. That way, we’ll avoid the hassle of having to come up with numbers for them. The URL will be something like /miniatures/783297, and that’s it. As for Goblin 1, Goblin 2 and so on, we’ll include them in the title.
Enabling the display of tags by category (weapons, race, armour, etc.) isn’t difficult. And it will be much simpler to do this for a specific miniature rather than for a set containing many different miniatures. As there will be a direct link between the set and the miniatures, it will be easy to pull tags from the miniatures into the sets and vice versa.
For miniatures with lots of different assembly options (5–10 variants of weapons and everything else), we can simply add a note to the description stating that different configurations are possible and that the image does not reflect reality. We could add a ‘multi-variant’ tick box to highlight this.
And yes, we’ll be able to move away from the current hybrid system, where sets and miniatures are combined into a mixed category and two identical sets are stored as different miniatures/models.
As for ambition. All in all, it was quite ambitious to launch this whole project 13 years ago, investing so much of each of our time and effort. So we’ll cope perfectly well with yet another complication! We’ve manually sorted through tens of thousands of sets and hundreds of thousands of images, so we’ll manage this too. We just need to think everything through so that it all makes sense, and then we’ll be able to proceed according to plan. And automation will help.