diff --git a/doc/properties.toml b/doc/properties.toml
index 95bee90b..aaaf43be 100644
--- a/doc/properties.toml
+++ b/doc/properties.toml
@@ -37,8 +37,8 @@ show_landing_text = true
 hexcasting = "https://raw.githubusercontent.com/gamma-delta/HexMod/main/Common/src/main/resources"
 
 [i18n]
-lang = "en_us"
-filename = "{lang}.json"
+default_lang = "en_us"
+filename = "{default_lang}.json"
     [i18n.extra]
     "item.minecraft.amethyst_shard" = "Amethyst Shard"
     "item.minecraft.budding_amethyst" = "Budding Amethyst"
diff --git a/doc/pyproject.toml b/doc/pyproject.toml
index 255810ee..349e0b7d 100644
--- a/doc/pyproject.toml
+++ b/doc/pyproject.toml
@@ -2,12 +2,9 @@
 requires = ["hatchling"]
 build-backend = "hatchling.build"
 
-[tool.hatch.build]
-packages = ["src/hexdoc"]
-
 
 [project]
-name = "HexDoc"
+name = "hexdoc"
 version = "0.1.0"
 authors = [
     { name="Alwinfy" },
@@ -26,13 +23,17 @@ dependencies = [
 dev = [
     "black==23.7.0",
     "isort==5.12.0",
-    "pytest==7.3.1",
-    "syrupy==4.0.2",
+    "pytest~=7.3.1",
+    "syrupy~=4.0.2",
+    "hatchling",
 ]
 
 [project.scripts]
 hexdoc = "hexdoc.hexdoc:main"
 
+[project.entry-points."hexdoc.book_data"]
+"hexcasting:thehexbook" = "hexdoc._book_data"
+
 [project.entry-points."hexdoc.Page"]
 hexdoc-patchouli = "hexdoc.patchouli.page.pages"
 hexdoc-hexcasting = "hexdoc.hexcasting.page.hex_pages"
@@ -46,6 +47,10 @@ hexdoc-minecraft = "hexdoc.minecraft.recipe.ingredients"
 hexdoc-hexcasting = "hexdoc.hexcasting.hex_recipes"
 
 
+[tool.hatch.build]
+packages = ["src/hexdoc"]
+
+
 [tool.pytest.ini_options]
 addopts = ["--import-mode=importlib"]
 markers = [
@@ -103,10 +108,6 @@ reportUntypedBaseClass = "error"
 reportUntypedClassDecorator = "error"
 reportUntypedFunctionDecorator = "error"
 reportUntypedNamedTuple = "error"
-reportUnusedClass = "error"
-reportUnusedExpression = "error"
-reportUnusedFunction = "error"
-reportUnusedVariable = "error"
 reportWildcardImportFromLibrary = "error"
 
 reportMissingTypeArgument = "warning"
@@ -116,6 +117,10 @@ reportUnknownLambdaType = "warning"
 reportUnknownMemberType = "warning"
 reportUnnecessaryComparison = "warning"
 reportUnnecessaryIsInstance =  	"warning"
+reportUnusedClass = "warning"
+reportUnusedExpression = "warning"
+reportUnusedFunction = "warning"
 reportUnusedImport = "warning"
+reportUnusedVariable = "warning"
 
 reportUnknownVariableType = "none"
diff --git a/doc/src/hexdoc/_book_data/__init__.py b/doc/src/hexdoc/_book_data/__init__.py
new file mode 100644
index 00000000..351f2846
--- /dev/null
+++ b/doc/src/hexdoc/_book_data/__init__.py
@@ -0,0 +1 @@
+BOOK_DATA_PATH = "data.json"
diff --git a/doc/src/hexdoc/_templates/__init__.py b/doc/src/hexdoc/_templates/__init__.py
new file mode 100644
index 00000000..e69de29b
diff --git a/doc/src/hexdoc/minecraft/i18n.py b/doc/src/hexdoc/minecraft/i18n.py
index e93ba824..56831844 100644
--- a/doc/src/hexdoc/minecraft/i18n.py
+++ b/doc/src/hexdoc/minecraft/i18n.py
@@ -16,7 +16,7 @@ from hexdoc.utils import (
     Properties,
     ResourceLocation,
 )
-from hexdoc.utils.deserialize import isinstance_or_raise, load_json
+from hexdoc.utils.deserialize import isinstance_or_raise, load_json_dict
 
 
 class I18nContext(TypedDict):
@@ -115,7 +115,7 @@ class I18n:
         # we could also use that to ensure all i18n files have the same set of keys
         lang_dir = props.resources_dir / "assets" / props.modid / "lang"
         path = lang_dir / props.i18n.filename
-        raw_lookup = load_json(path) | (props.i18n.extra or {})
+        raw_lookup = load_json_dict(path) | (props.i18n.extra or {})
 
         # validate and insert
         self.lookup = {}
diff --git a/doc/src/hexdoc/minecraft/recipe/abstract_recipes.py b/doc/src/hexdoc/minecraft/recipe/abstract_recipes.py
index 42f5eb83..138d22d1 100644
--- a/doc/src/hexdoc/minecraft/recipe/abstract_recipes.py
+++ b/doc/src/hexdoc/minecraft/recipe/abstract_recipes.py
@@ -3,7 +3,7 @@ from typing import Any, Self, cast
 from pydantic import ValidationInfo, model_validator
 
 from hexdoc.utils import AnyPropsContext, ResourceLocation, TypeTaggedUnion
-from hexdoc.utils.deserialize import load_json
+from hexdoc.utils.deserialize import load_json_dict
 
 
 class Recipe(TypeTaggedUnion[AnyPropsContext], group="hexdoc.Recipe", type=None):
@@ -35,7 +35,7 @@ class Recipe(TypeTaggedUnion[AnyPropsContext], group="hexdoc.Recipe", type=None)
             path = recipe_dir / f"{id.path}.json"
             if recipe_dir == context["props"].default_recipe_dir:
                 # only load from one file
-                values = load_json(path) | {"id": id}
+                values = load_json_dict(path) | {"id": id}
             elif not path.exists():
                 # this is to ensure the recipe at least exists on all platforms
                 # because we've had issues with that before (eg. Hexal's Mote Nexus)
diff --git a/doc/src/hexdoc/patchouli/book.py b/doc/src/hexdoc/patchouli/book.py
index c075191b..fdceebc1 100644
--- a/doc/src/hexdoc/patchouli/book.py
+++ b/doc/src/hexdoc/patchouli/book.py
@@ -1,3 +1,5 @@
+from importlib import resources
+from importlib.metadata import entry_points
 from typing import Any, Generic, Literal, Self, cast
 
 from pydantic import Field, ValidationInfo, model_validator
@@ -12,7 +14,7 @@ from hexdoc.utils import (
     ResLoc,
     ResourceLocation,
 )
-from hexdoc.utils.deserialize import isinstance_or_raise, load_json
+from hexdoc.utils.deserialize import isinstance_or_raise, load_json_dict
 
 from .book_models import AnyBookContext, BookContext
 from .category import Category
@@ -33,8 +35,7 @@ class Book(Generic[AnyContext, AnyBookContext], HexDocModel[AnyBookContext]):
     """
 
     # not in book.json
-    context: AnyBookContext = Field(default_factory=dict)
-    categories: dict[ResourceLocation, Category] = Field(default_factory=dict)
+    i18n_data: I18n
 
     # required
     name: LocalizedStr
@@ -65,51 +66,73 @@ class Book(Generic[AnyContext, AnyBookContext], HexDocModel[AnyBookContext]):
     custom_book_item: ItemStack | None = None
     show_toasts: bool = True
     use_blocky_font: bool = False
-    do_i18n: bool = Field(default=False, alias="i18n")
+    i18n: bool = False
     macros: dict[str, str] = Field(default_factory=dict)
     pause_game: bool = False
     text_overflow_mode: Literal["overflow", "resize", "truncate"] | None = None
-    extend: str | None = None
-    """NOTE: currently this WILL NOT load values from the target book!"""
+    extend: ResourceLocation | None = None
     allow_extensions: bool = True
 
-    @classmethod
-    def load(cls, data: dict[str, Any], context: AnyBookContext):
-        return cls.model_validate(data, context=context)
-
     @classmethod
     def prepare(cls, props: Properties) -> tuple[dict[str, Any], BookContext]:
         # read the raw dict from the json file
         path = props.book_dir / "book.json"
-        data = load_json(path)
-        assert isinstance_or_raise(data, dict[str, Any])
+        data = load_json_dict(path)
 
-        # NOW we can convert the actual book data
-        return data, {
-            "i18n": I18n(props, data["i18n"]),
+        # set up the deserialization context object
+        assert isinstance_or_raise(data["i18n"], bool)
+        assert isinstance_or_raise(data["macros"], dict)
+        context: BookContext = {
             "props": props,
+            "i18n": I18n(props, data["i18n"]),
             "macros": DEFAULT_MACROS | data["macros"],
         }
 
+        return data, context
+
+    @classmethod
+    def load(cls, data: dict[str, Any], context: AnyBookContext) -> Self:
+        return cls.model_validate(data, context=context)
+
+    @classmethod
+    def from_id(cls, book_id: ResourceLocation) -> Self:
+        # load the module for the given book id using the entry point
+        # TODO: this is untested because it needs to change for 0.11 anyway :/
+        books = entry_points(group="hexdoc.book_data")
+        book_module = books[str(book_id)].load()
+
+        # read and validate the actual data file
+        book_path = resources.files(book_module) / book_module.BOOK_DATA_PATH
+        return cls.model_validate_json(book_path.read_text("utf-8"))
+
+    @model_validator(mode="before")
+    def _pre_root(cls, data: dict[str, Any], info: ValidationInfo) -> dict[str, Any]:
+        context = cast(AnyBookContext, info.context)
+        if not context:
+            return data
+
+        return data | {
+            "i18n_data": context["i18n"],
+        }
+
     @model_validator(mode="after")
     def _post_root(self, info: ValidationInfo) -> Self:
         """Loads categories and entries."""
         context = cast(AnyBookContext, info.context)
         if not context:
             return self
-        self.context = context
 
-        # categories
-        self.categories = Category.load_all(context)
+        # load categories
+        self._categories: dict[ResourceLocation, Category] = Category.load_all(context)
 
-        # entries
+        # load entries
         for path in context["props"].entries_dir.rglob("*.json"):
-            # i used the entry to insert the entry (pretty sure thanos said that)
             entry = Entry.load(path, context)
-            self.categories[entry.category_id].entries.append(entry)
+            # i used the entry to insert the entry (pretty sure thanos said that)
+            self._categories[entry.category_id].entries.append(entry)
 
         # we inserted a bunch of entries in no particular order, so sort each category
-        for category in self.categories.values():
+        for category in self._categories.values():
             category.entries.sort()
 
         return self
@@ -120,5 +143,5 @@ class Book(Generic[AnyContext, AnyBookContext], HexDocModel[AnyBookContext]):
         return self.model if self.index_icon_ is None else self.index_icon_
 
     @property
-    def props(self) -> Properties:
-        return self.context["props"]
+    def categories(self):
+        return self._categories
diff --git a/doc/src/hexdoc/patchouli/text/formatting.py b/doc/src/hexdoc/patchouli/text/formatting.py
index 05190099..c3694dca 100644
--- a/doc/src/hexdoc/patchouli/text/formatting.py
+++ b/doc/src/hexdoc/patchouli/text/formatting.py
@@ -216,10 +216,14 @@ class ParagraphStyle(Style, frozen=True):
         return out.element("p", **self.attributes)
 
 
+def is_external_link(value: str) -> bool:
+    return value.startswith(("https:", "http:"))
+
+
 def _format_href(value: str) -> str:
-    if not value.startswith(("http:", "https:")):
-        return "#" + value.replace("#", "@")
-    return value
+    if is_external_link(value):
+        return value
+    return f"#{value.replace('#', '@')}"
 
 
 class FunctionStyle(Style, frozen=True):
diff --git a/doc/src/hexdoc/utils/deserialize.py b/doc/src/hexdoc/utils/deserialize.py
index 419fe0ff..96d25c1a 100644
--- a/doc/src/hexdoc/utils/deserialize.py
+++ b/doc/src/hexdoc/utils/deserialize.py
@@ -38,7 +38,7 @@ JSONDict = dict[str, "JSONValue"]
 JSONValue = JSONDict | list["JSONValue"] | str | int | float | bool | None
 
 
-def load_json(path: Path) -> JSONDict:
+def load_json_dict(path: Path) -> JSONDict:
     data: JSONValue = json.loads(path.read_text("utf-8"))
     assert isinstance_or_raise(data, dict)
     return data
diff --git a/doc/src/hexdoc/utils/model.py b/doc/src/hexdoc/utils/model.py
index 33553e5a..c954e786 100644
--- a/doc/src/hexdoc/utils/model.py
+++ b/doc/src/hexdoc/utils/model.py
@@ -4,7 +4,7 @@ from typing import TYPE_CHECKING, Any, Generic, Self, TypeVar, dataclass_transfo
 from pydantic import BaseModel, ConfigDict
 from typing_extensions import TypedDict
 
-from .deserialize import load_json
+from .deserialize import load_json_dict
 
 if TYPE_CHECKING:
     from pydantic.root_model import Model
@@ -56,5 +56,5 @@ class FrozenHexDocModel(Generic[AnyContext], HexDocModel[AnyContext]):
 class HexDocFileModel(HexDocModel[AnyContext]):
     @classmethod
     def load(cls, path: Path, context: AnyContext) -> Self:
-        data = load_json(path) | {"__path": path}
+        data = load_json_dict(path) | {"__path": path}
         return cls.model_validate(data, context=context)
diff --git a/doc/src/hexdoc/utils/properties.py b/doc/src/hexdoc/utils/properties.py
index aa35aeb1..ce0e3776 100644
--- a/doc/src/hexdoc/utils/properties.py
+++ b/doc/src/hexdoc/utils/properties.py
@@ -33,7 +33,7 @@ class PlatformProps(HexDocModel[Any]):
 
 
 class I18nProps(HexDocModel[Any]):
-    lang: str
+    default_lang: str
     filename: str
     extra: dict[str, str] | None = None
 
@@ -78,7 +78,7 @@ class Properties(HexDocModel[Any]):
 
     @property
     def lang(self):
-        return self.i18n.lang
+        return self.i18n.default_lang
 
     @property
     def book_dir(self) -> Path:
diff --git a/doc/src/hexdoc/utils/resource.py b/doc/src/hexdoc/utils/resource.py
index c6e34525..9ba84f30 100644
--- a/doc/src/hexdoc/utils/resource.py
+++ b/doc/src/hexdoc/utils/resource.py
@@ -35,12 +35,16 @@ class BaseResourceLocation:
         cls._from_str_regex = regex
 
     @classmethod
-    def from_str(cls, raw: str) -> Self:
+    def from_str(cls, raw: str, default_namespace: str | None = None) -> Self:
         match = cls._from_str_regex.fullmatch(raw)
         if match is None:
             raise ValueError(f"Invalid {cls.__name__} string: {raw}")
 
-        return cls(**match.groupdict())
+        groups = match.groupdict()
+        if not groups.get("namespace") and default_namespace is not None:
+            groups["namespace"] = default_namespace
+
+        return cls(**groups)
 
     @model_validator(mode="wrap")
     @classmethod