I think the big one is to save a scene (and reload it) during playback. It's clear the editor has this power in some fashion already, and it would be hugely useful for many types of games. 

But I know that's no small task. So, besides that, there's a ton that Unity could include, like: 

More data types for PlayerPrefs. The abllity to save arrays, for example. See EasySave on the asset store for a good third party solution for this.

Unique GameObject IDs that are static. When you save data associated with an object, you need to uniquely identify the object. GetInstanceID is a unique id for each object, but they can change every time you load the scene. I came up with a simple hack for powerups, but there should be a better way, especially since Unity's own way of saving data (PlayerPrefs) requires unique keys. 

An automated way to load and save a script's current properities. Right now, you need to write your own saver and loader for every single variable in your script. Add a variable? Add some code to saver. And don't forget your loader.  This is error-prone, and this should be automated. It's clear that the property editor has automated access to all the script's variables, so why not give us a way to bundle it up and save it and load it? It wouldn't have to work for everything (I understand the problems with saving transforms and whatnot) but just letting us save all integers, floats, and strings would save a ton. 

And if not that, at least give us a way to walk through them ourselves so we could roll our own autmated saver that doesn't require me hand-adding variables to save every time I change my script. There would still be the need for special-casing things, but this would be a great head start.

Summary

I know there's no way for Untiy Technologies to make the perfect SaveMyGame() feature that would work for every game out there, but really Unity needs more than what it has at the moment. It's almost non-existant.  In other game development environments, I'd have the tools to say "these variables are going to be saved." and then that's the end of it. 

Currently, for any game this isn't dead simple, SaveGame is one of the first things I need to think about -- and think hard about -- before I do any fundimental development.  And that's typically not what Unity's about. Unity is about focusing on your game, and not your engine. 

Here it is in pastebin. 

.
 
@MenuItem ("Maquette/Batch Rename")
static function BatchRename() 
	{ // Renames all selected items in the Hierarchy to the first item selected, with numbers.
	var iname : String;
	ispacer = "0"; 
    icount = 0; 
	iname = Selection.activeGameObject.name;   // The item in the inspector
	
	istuff = Selection.gameObjects.length;  // if I wanted this to support renaming of > 99 objects correctly, I'd use this. 
	
	for (igo in Selection.gameObjects)
		{
		icount ++; 
		if (icount > 9) ispacer = ""; 
		igo.name =iname + "-" + ispacer + icount; 
		}
	}

How I imagine it is you can place objects into the crafting table, and depending on your ingenuity stat, you may see some spots around the item you place light up slightly. Rolling over these lit up spots, you may see items in your inventory highlighted.  This is giving you hints of how new crafting recipes can be discovered. This is an in-game approximation for having your character have insight into how things are built and how they can be put together.

The idea is that with low ingenuity, the highlights don't appear for some recipes, and when they highlight objects, they highlight "fuzzy" -- meaning it may highlight false positives in your inventory, or not highlight others. As your ingenuity increases, these fuzzy recipes get more defined, and new highlights start to appear for new recipes.  Each recipe has it's own skill level, so easy and basic things appear first, with more complex recipes appearing with higher ingenuity. Not all recipes would appear this way -- some could only be learned via wall carvings or parchments, etc. 

Ingeunuty, or perhaps another stat like "skill", may keep you from successfully building complex items (keeping people from looking up the reciipe online and buliding things their character couldn't have otherwise known about)

Ideally I'd like this inventory item to be a prefab that I can drop in the world. That's all well and good, but my issue comes up when the player picks it up. 

When it's in the player's inventory, there's no object in the world anymore, its just stored in an array for inventory objects.  Of course, if the player drops it, it will again be in the world. 

This *feels* like it will use two different kinds of objects, a GO with a mesh (etc) and a non-mesh object (not sure what that's called?) when it's in your inventory. BUT I DONT LIKE THIS ROUTE. It doesn't feel right to me.  Id prefer it to be one thing, although, if I can't have it that way, I understand. 

Any suggestions on how to handle this?

This is a free association sketch I did while brainstorming on Maquette. Note the simplified dome in the middle of the image.

I do a lot of free association drawing. it's just like free association writing, except for I'm drawing. But quite often I have no idea what I'm about to draw when I start. This is one of those cases. It's not about the quality of the sketch, or even the content. It's more about what my brain thinks of while I'm drawing it. 

Good news everyone! I'm currently working on Maquette full time now.   I hope to update this area periodically with something of interest regarding the game. My goal is weekly, so I imagine that will work out to once a month. 

1. As a view style for Project and Hierarchy

I'm not going to spend much time on this because it should be obvious. The Asset Picker is so much easier to see what the hell you're about to click on than the current icon/name view of Project and Hierarchy. The current icons in Unity are incredibly difficult to parse visually.  

I wouldn't want this view all the time, but  for certain things, it's a million times easier to grok this view at a glance than the current views. And it seems as though the hard part of implmenting this is already done -- just apply it as a view for Project and Hiearchy; or perhaps a 3rd panel that shows the contents of the current folder you've got chosen in Project.

2. As a Level Design tool

Again, I don't think the Asset Picker view is always the superior way to do it, but for most cases in my work flow, it would be so much easier.

Now, if the Project view supported this type of preview rendering, that would get us pretty close to what I'm looking for. 

But there's one thing that is missing from Unity, and that is more of a construction set workflow. Right now, every time you want to add a prefab to the scene, you've got to ther drag it out of the Project view and into Scene, or you've got to duplicate an already existing one in the scene itself. 

What I want is a "click to add" type of control that will allow me to place multiples in the scene without having to drag multiple times. Think about how SimCity or Rollercoaster Tycoon lets you add things to the world. It's just click, click, click.  The current method making you pogo-stick between project and scene, even when adding multiples of the same prefab.

I think it would work well as a tool, assigned to the T key by default. This would allow you to rest your hand on the qwert keys, and switch between moving (w), rotating (e), scaling (r), and adding (t). Perfect for rapid level design. 

Misc. notes

This is just my first pass at how I'd like it to work. I could just as easily see this as a level-design palette that lets you add items to Scene visually, even if Project and Hierarchy keep their text-only view. Basically implment the Asset Picker as is now, but open all the time, and showing prefabs from Project.  

But while I'm at it, here's some random notes about the current process: 

• Project view as it is feels quite unwieldy. I often just feel lost in a sea of folders and sub-folders, with these tiny icons that all look similar. 
• It would be nice if we could be shortcuts to certain folders and see JUST them using the Asset Picker view. The ability to have a palette with tabs for "power ups", "Buildings", "enemies", and to be able to ping-pong between them quickly, without having to deal with intermediary folders would be ideal. We can get close to this with a clever Project layout, but due to the tree-view nature of the project tab, it is very difficult to ping-pong between a set of folders.  A level design palette could solve this.

The lower 90%

My final thought on this is about what I call the Lower 90%. We have the upper 10% of features in an app, and these are the super-sexy marketing-ready features that wow users. 

Then we have the lower 90%. And these are not quite as sexy, but they're things we use all the time, on every project. Everything I've mentioned here are features every Unity user would use, on every project, and they'd use it all the time.  

And, as it stands, it seems like the big hurdles around a visual preview mode for level design (and for project and Hierarchy) are already done. I love Uunity and the entire UT crew. I hope this blog post inspires them to improve upon asset selection. It may not be as sexy as big shiny new features, but it seems like this is some low-hanging fruit that could really improve workflow for everyone who uses Unity.

I showed off Maquette at Game Developers Conference, and the feedback was phenominal! Thanks to everyone who took the time to check it out. There's a LOT of cool shit at GDC and I'm really honored to hear how the game resonated with people. 

Here's some of the press it's been getting: 

Kotaku: A Key Is A Bridge In This Magical Game

IndieGames.com: The Experimental Gameplay Sessions Highlights

Gamerant.com: Could Maquette be the next Portal?