diff --git a/UsedMemory.md b/UsedMemory.md
new file mode 100644
index 0000000..8365282
--- /dev/null
+++ b/UsedMemory.md
@@ -0,0 +1,112 @@
+--this file is still a stub--
+
+# Memory Maps
+ 
+Forth was developed as an application in old computers, where a small ROM keeps the basic input output routines (BIOS) and all memory was mapped in RAM. The applications was loaded from magnetic devices.
+
+Modern computers uses flash and sram memory. Flash memory preserves the content and could be re-writted  at least 10.000 and sram does not preserve contents and could be re-written without limits. 
+
+Flash memory write cycle needs a specific CPU/MCU routine, that erase a pages, copy contents from sram and release a page.
+
+Flash memory is better for static content, and SRAM memory for dynamic content.
+
+### Classic Map
+
+( using name and code dictionaries mixed in one linked list )
+
+In PolyForth the memory map 
+
+for a interactive task is 
+         
+_LOW RAM \[|DP0--user dictionary-->|DP----|SP<--SP0|TIB0-->IN----#TIB|----|RP<--RP0|-- user area --\] HIGH RAM_
+
+for a non-interative task
+          
+_LOW RAM \[|SP<--SP0||RP<--RP0|-- user area --\] HIGH RAM_
+
+### Review Map
+
+A review of organization, by place the user area before stacks, and considering the user area as a place holder and some rellocations
+
+_LOW RAM \[ --user area |--<--SP0|--<--RP0| \] HIGH RAM_
+
+user area as 
+
+\[ user variables |DP0 --user dicitonary-->|DP -->-- |PAD -->--|TIB0 -->IN--|TIB# \]
+
+inside user variables
+
+| name | use |
+| -- | -- |
+|    | _for each task_ |
+| UM | size of task block |
+| DP0 | start of dictionary |
+| TB0 | start of terminal input buffer |
+| TBZ | end of terminal input buffer |
+| SP0 | start of parameter stack |
+| RP0 | start of return stack |
+| | _for this task_ |
+| STATE | |
+| BASE | |
+| DP | |
+| TOIN | |
+| HLD | |
+| CSP | |
+| CRP | |
+| DSK | |
+| BLK | |
+| SCR | |
+| ERRO | |
+| WARN | |
+| CONTENT | |
+| CURRENT | |
+| LATEST | |
+| 'QUERY | |
+| 'EXPECT | |
+| 'KEY | |
+| 'EMIT | |
+| 'KEY? | |
+
+### groups
+
+For a generic Forth system,
+
+| name | type | memory | size | used |
+| --- | --- | --- | --- | --- |
+| BIOS | static | flash | vary | routines for input/output system |
+| CORE | static | flash | vary | processor specific routines for Forth, aka primitives |
+| FORTH | static | flash | vary | generic routines for Forth, aka compounds |
+| SP | dynamic | sram | 18 cells | data stack for Forth, aka the parameter data stack |
+| RP | dynamic | sram | 18 cells | return stack for Forth, aka return stack |
+| OS | dynamic | sram  | 18 cells | passing parameters between routines, aka system stack |
+| TIB | dynamic | sram | 80 bytes | terminal input/output buffer| 
+| PAD | dynamic | sram | 84 bytes | scratch area for Forth |
+| BOB | dynamic | sram | 36 bytes | scratch area for Forth format numbers | 
+| DICIONARY | static | vary | vary | generic routines for Forth, aka dictionary |
+| USER VARS | dynamic | sram | vary | keep values of Forth variables |
+| | | | |
+
+Notes:
+
+- The values for TIB, PAD, BOB are mininal size, as defined by standart Forth 1994 and Forth 2012;
+- The values for stacks are minimal enough, as said by Charles Moore
+- About TIB there is a quote, from old days of IBM paper card, " 72 is continue ";
+
+### The dictionary
+
+The dictionary of words is composed by two lists, name and code. In the beginining of Forth, these lists are separate: The name list was a linked list with references to next name and to code of word; The code list is just a sequence of pieces of codes. 
+
+The name list have a format as [reference to next word, size and flags, name of word, reference to code] and code list have no fixed format.
+
+The name list grows backwards from end of free memory and the code list grows from begin from free memory. 
+
+Nowadays those lists are merged as one with [reference to next word, size and flags, name of word, vary code].
+
+If no need of interactive interpretation or compiling...
+
+
+
+
+
+
+