An enhanced description of the problem that we're trying to solve

cloud.ganttproject.colloboque/DesignDoc.md

@@ -4,12 +4,38 @@ The goal of Colloboque project is to create a protocol and software for real-tim
 relational structured data with the option for creating offline data branches with subsequent automated or 
 assisted merging.
 
+Suppose that our project data is stored on GanttProject Cloud in a task table:
+
+| id | start      | duration | color |
+|----|------------|----------|-------|
+| 1  | 2023-09-01 | 1d       | red   |
+
+and there are two GanttProject client apps working with that project. One of the client apps is online and 
+the user changes the task color to yellow, while another is offline, and the user changes the task start date to "2023-10-01", adds a new task with id=2, and goes online in a few hours after that.
+
+We want our table to look like this when both changes are merged:
+
+| id | start      | duration | color  |
+|----|------------|----------|--------|
+| 1  | 2023-10-01 | 1d       | yellow |
+| 2  | 2023-09-01 | 1d       | red    |
+
+It doesn't seem to be a trivial problem, because concurrent update of the same row would've been an issue even if 
+we had a centralized database with both users being online at the moment of change. In our case any user may go offline
+for quite long time, and we want a reliable protocol that lets them receive the updates or merge the changes that they created when they were offline.
+
 ## Alternatives and Related Technologies
 
 There is a number of related technologies and possible alternatives.
 
 ### Logical Replication
 ### RAFT 
 ### PostgreSQL Multimaster
+https://pgconf.ru/en/2021/290216
+
 ### PostgreSQL branching
-### CRDT
+https://neon.tech/docs/introduction/branching
+
+### CRDT
+
+### Text data merging