Yes, the Salon (and the wardrobe to a lesser extent) is in a sad state at the moment ?. There are more long blog posts coming, but this is my priority.

I wouldn't call it a wardrobe 4.0, because it's using a lot of the same components as the 3.0 for the front end. What's really being shifted is the back end, which can be explained as:

• The /wapi.php API that was built on top of the old avatar system. Slow loading of items, loads them to a mongoDB for caching, just incredibly resource intensive.
• /slib/ which is the old avatar system, and builds the avatars themselves. Thousands of lines of code that are now just powering the avatar system
• The database structure from 7+ years ago, that has serialized variables (not the right practice anymore) that have to be turned to strings, then turned to JSON for the wapi.php controllers.

What I'm doing now is rebuilding those pieces. I'm replacing our home-made framework in /slib/ with Laravel, a battle-tested php framework. This will also set the base for more amazing API calls in the future to make the site faster (which I'm really excited for).

Secondly, I'm replacing the thing that renders the avatar, which is the /slib/avatar.php system. It's over 750 lines of code, brings in tons of classes, and means that when or I want to add a new feature like shared avatars, it's a heavy lift that takes a week. I was able to implement my avatar in a new system in 70 lines of code, with ease of adding new features (hair isn't included).

I have all of the tasks for "feature parity" on these worked out, and there is a real timeline that has been kept up so far internally.

I know there is MASSIVE FATIGUE for posts like this, with "working on an entirely new system!" especially after the last wardrobe rewrite (which was actually technically pushed on us faster than we'd have liked because of the XML changes) but this time it's be doing right. Tests are being written, timelines have been decided and being stuck to.

It's always a real tossup. When developing the v3 wardrobe it seemed like the best past forward to use the older codebase as the framework, and it just ended up not really working well or being great planning. It got something out of the door quicker than otherwise, but just wasn't good planning :|

The front end is also being changed, in tandem. There needs to be changes to the way that it grabs information - which is what is going to resolve a lot of what you're asking for in that thread.

The majority of those items on your list seem like front end problems, but they're all tied to the way that the avatars are stored on the backend and the sort of hacky fixes I had to make to make them display in the new front end. This is resolved with a backend designed for a new system, and an API that makes it easy to change the items. Things like the shared items, way that items are listed, etc, are all back end storage issues.

The hex value isn't as simple as adding an input, and that was actually there at some stage of the wardrobe testing. It interfered with the color picker and the choice I made then is that more users will get benefit from the color picker vs just an input to put in a hex code. Like I said, we were running out of time before the wardrobe became unusable in updated versions of modern browser. Instead of going back and making more hacky changes to that front end, I'm spending my time on something a lot more valuable and useful going forward.