Improved doc

docs/identified_problems.md

@@ -2,4 +2,11 @@
 
 - [ ] In most of the components from 'models/' we have a method to add  input and output to the model, however this linking of components is already done in oc-schedule when parsing the MxGraph. We need to determine if adding relations between components inside the objects themself is necessary.
     - When running in debug mode with a breakpoint inside the first line of computing.addLink it is only called once
-- [ ]
+- [ ]
+
+## MxGraph
+
+- [ ] The ConsumeMxGraphModel is way too long, it should refactored and broken down in different sub methods
+    - mxcell are put inside an <object> tag when the settings have been opened, wether values have been set or not. Maybe we could find a way to make mxgraph add these whenever we add a component to the graph.
+    - then identify the links only 
+- [ ] It is unclear what are the inputs and the ouputs. It seems like they were implemented to link two components, but it seems redundant with the identification of links

docs/lexicon.md

@@ -0,0 +1,6 @@
+- rType : ressource type, can only be :
+    - rtype.DATA
+    - rtype.COMPUTING
+    - rtype.STORAGE
+    - rtype.DATACENTER
+    - rtype.INVALID if it doesn't match with any of the previous type