Are Sub Attributes possible?

[gjh42]

For starters, you can work down in subcategories a good bit farther than just "gold rings". It seems reasonable to have a number of styles of gold rings as the products; this will let you display pictures of typical examples, and then the customer can go into the product to select stones, endpieces, etc.
You may consider the balance between a strict logical heirarchy of features and a more level spread of feature combinations. Having to pick between crystal and swarovski pearl before picking through a few colors doesn't strike me as easier than picking between six distinct stone options.

If you have one product that can vary radically in appearance depending on options chosen, how will the customer know which she wants to explore? Don't make the customer try all the combinations before seeing what they look like. Many of them should be visible side by side in a listing.

[ktomas]

If I am understanding you correctly, I agree. Given the example I gave, it would be plausible to have a number of products that account for the most popular combinations of options and attributes. Unfortunately, this was just a simple example. the actual options go much, MUCH deeper than 3 selections (for instance, "Crystal" has 80 color options, of which have 2 shapes associated with each color, and 2 sizes within each shape). Without dependent attributes, it doesn't seem like there is any way to tweak the current capabilities to accomplish this. Am I understanding you correctly?

[gjh42]

With that many option values, picking combinations directly does become less feasible. But I question whether you need to have the shape dependent on the color, and the size dependent on the shape. Square 6mm, square 8mm, round 6mm and round 8mm seems like a simple choice to make at one go. Do they vary in price based on crystal color, size or shape?

The coding required to really get close to the heirarchy you want would be involved enough that you would probably be better off waiting on v2.0 to try for that.

[ktomas]

I agree about the size and color relationship. In actuality, the color is the one dependent on the size, as not every color is available in every size and shape (another variable that I haven't mentioned yet that further complicates things).

When is v2 set to launch?

[gjh42]

It is supposed to be released as a public beta this month, with a few intermediate versions before final v2.0 release in the spring. As soon as the beta is out, there will be people swarming over it to try out the new possibilities for modification.

The color availability issue is one that will definitely be addressed in v2.0 - see the announcement:

Stock/sku support

* stock can be set for attribute combinations
* ajax suppport for stock display
* sku can be set for attribute combinations
* ajax support for sku display
* individual reorder levels can be set per attribute/sku combination, and a global setting can optionally override

[kobra]

Buttin in here:

Crystal" has 80 color options, of which have 2 shapes associated with each color, and 2 sizes within each shape

Size = 2 options
Shape = 2 options
color = 80 options

Or is that not what you said??

[ktomas]

That is what I said. What I've decided to do is to have 11 options, and simply group them together with the design rather than trying to have 4 options with dependent variables (obviously, because this is impossible). For instance, I'll have:

Center Stone
_Crystal
_Pearl

Stone Color
_Brass
_Green
_Bronze
_etc

Stone Shape
_Round
_Square

And so on.

The only problem with this, is that the customer may not want to have a center stone, and would chose "none," but they are still able to chose stone color and shape...etc, which become void since they said "no" to a stone in the first place. I've seen this done before, it just doesn't seem as professional to me.

Is there a way to "turn off" subsequent options if a preceding one is not checked? They could still be visible, just not active? I'm pretty sure I've seen this before.