Simplify doc for first-time users

Author: VEZY

docs/src/tutorials/1.manipulate_node.md

@@ -149,28 +149,28 @@ node_5 = get_node(mtg, 5)
 get_root(node_5)
 ```
 
-### Get the attributes
+### Read and set attribute values
 
-This section has its own tutorial! Head over the next page to learn how to get the nodes attributes.
+To read one value from a node:
 
-## Columnar Attribute Backend
+```@example usepkg
+leaf = get_node(mtg, 5)
+leaf[:Width]
+```
 
-By default, `read_mtg` now loads attributes using a typed columnar backend (per symbol). The node API remains simple:
+To set one value:
 
 ```@example usepkg
-leaf = get_node(mtg, 5)
-attribute(leaf, :Width)
-attribute!(leaf, :new_var, 1.0)
-attribute_names(leaf)
-attributes(leaf, format=:namedtuple)
+leaf[:Width] = 1.0
+leaf[:Width]
 ```
 
-For schema-level operations on one symbol:
+You can also use the explicit API:
 
 ```@example usepkg
-add_column!(mtg, :Leaf, :temperature, Float64, default=20.0)
-rename_column!(mtg, :Leaf, :temperature, :temp)
-drop_column!(mtg, :Leaf, :temp)
+attribute(leaf, :Width)
+attribute!(leaf, :new_var, 1.0)
+attributes(leaf, format=:dict)
 ```
 
-`node_attributes(node)` is still available for compatibility, but new code should prefer `attribute`/`attribute!`/`attributes`.
+For more examples on querying values in a subtree (descendants, ancestors, filters), continue to [Traversal, descendants, ancestors and filters](@ref).

docs/src/tutorials/2.descendants_ancestors_filters.md

@@ -95,11 +95,13 @@ The previous notations are both equivalent to:
 node_attributes(node_5)[:Length]
 ```
 
-`node_attributes(node_5)` returns the node attribute container view for this node (columnar backend by default), so reading and writing through it does update the MTG.
+`node_attributes(node_5)` gives access to all attributes of this node.
 
-However, we still recommend the higher-level APIs (`node[:attr]`, `attribute(node, :attr)`, and `attribute!(node, :attr, value)`) because they are clearer and keep user code backend-agnostic.
+For day-to-day use, the simpler APIs are usually clearer:
 
- That is why the package implements the more generic `node_5[:Length]` notation that works with any structure used for the attributes, which helps develop more generic code.
+- `node[:attr]`
+- `attribute(node, :attr)`
+- `attribute!(node, :attr, value)`
 
 To get the names of all attributes available in the node subtree, you can use [`get_attributes`](@ref):