feat: 🎸 update readme with project description

Signed-off-by: Marcin Pastecki <marcin.pastecki@inndei.com>

README.md

@@ -5,24 +5,14 @@
 
 # Polymesh Subquery
 
-This project has been initially created to replace [the harvester](https://github.com/PolymathNetwork/polkascan-pre-harvester) meaning the output of both should match as much as possible, however since the harvester serializes chain objects in slightly different ways than polkadot-js, this project contains a list of special cases handled both at the initial deserialization level (in [project.yaml](project.yaml).network.types) and at the serialization level in [serializeLikeHarvester.ts](src/mappings/serializeLikeHarvester.ts).
+Polymesh Subquery allows for querying historical events on the Polymesh chain through a GraphQL interface. This project is based on [SubQuery](https://github.com/subquery/subql), and contains the mappings specific to Polymesh.
 
 ## Running
 
 1. In the [docker-compose.yml](docker-compose.yml) file, set the appropriate [environment variables](#env-settings) for `subquery-node` container
 2. Install subql cli: `npm i -g @subql/cli`
 3. `./rerun.sh` (requires docker compose). To persist data between runs, remove the `-v` flag, which causes the docker volume to be removed
 
-## Using event_arg_x indexes.
-
-The event_arg_x columns are now text, in order for them to fit in BTree indexes they are truncated to 100 characters like in the harvester.
-This means that if you want to take advantage of the index in your query, you must use an expression like: `WHERE left(event_arg_x, 100) = 'foobar'`.
-Otherwise using `WHERE event_arg_x = 'foobar'` will result in a full table scan.
-
-## Debugging using the found_types table
-
-The found_types table generated by subquery contains all types that have been serialized like the harvester, the actual type is in column ID which is the primary key and therefore unique and the raw type is in the raw_type column, if things break this is the first place to look for potential mismatches between types.
-
 ## Version
 
 This SubQuery version works with chain versions 6.2.x
@@ -40,6 +30,20 @@ More advanced options are:
 - `NO_NATIVE_GRAPHQL_DATA` — this will set the event handler to only record what is necessary for tooling gql. This allows a stable indexer to be ran as native GraphQL handlers get developed
 - `MAX_OLD_SPACE_SIZE` — this will be passed onto the node process as `--max-old-space-size` flag. The recommendation is for this to be ~75% of available RAM. Defaults to 1536, a setting appropriate for 2GB.
 
+## Historical note
+
+This project has been initially created to replace [the harvester](https://github.com/PolymathNetwork/polkascan-pre-harvester) meaning the output of both should match as much as possible, however since the harvester serializes chain objects in slightly different ways than polkadot-js, this project contains a list of special cases handled both at the initial deserialization level (in [project.yaml](project.yaml).network.types) and at the serialization level in [serializeLikeHarvester.ts](src/mappings/serializeLikeHarvester.ts).
+
+### Using event_arg_x indexes.
+
+The event_arg_x columns are now text, in order for them to fit in BTree indexes they are truncated to 100 characters like in the harvester.
+This means that if you want to take advantage of the index in your query, you must use an expression like: `WHERE left(event_arg_x, 100) = 'foobar'`.
+Otherwise using `WHERE event_arg_x = 'foobar'` will result in a full table scan.
+
+### Debugging using the found_types table
+
+The found_types table generated by subquery contains all types that have been serialized like the harvester, the actual type is in column ID which is the primary key and therefore unique and the raw type is in the raw_type column, if things break this is the first place to look for potential mismatches between types.
+
 ## License
 
 This project uses [SubQuery](https://github.com/subquery/subql), which is [Apache 2.0 licensed](./LICENSE).