Differences

This shows you the differences between two versions of the page.

--- nodemcu-unofficial-faq [2015/06/03 23:22]
terrye Add Q&A on implementing complex actions as a function
+++ nodemcu-unofficial-faq [2016/05/14 00:59]
gandibowo [How does DEVKIT use DTR and RTS to enter download mode?]
@@ Line 1: / Line 1: @@
-====== nodeMCU Unofficial FAQ ======
+-====== nodeMCU Unofficial FAQ ======
 ** * * * Work in Progress * * * **
@@ Line 6: / Line 6: @@
 ===== What is this FAQ for? =====
-This FAQ does not aim to help you to learn to program or even how to program in Lua.  There are plenty of resources on the Internet for this, some of which are listed in [[#Where to start|Where to start]] .  What this FAQ does is to answer some of the common questions that a competent Lua developer would ask in learning how to develop Lua applications for the ESP8266 based boards running the [[http://nodemcu.com/index_en.html|nodeMCU]] firmware.
+This FAQ does not aim to help you to learn to program or even how to program in Lua.  There are plenty of resources on the Internet for this, some of which are listed in [[#Where to start|Where to start]].  What this FAQ does is to answer some of the common questions that a competent Lua developer would ask in learning how to develop Lua applications for the ESP8266 based boards running the [[http://nodemcu.com/index_en.html|nodeMCU]] firmware.
 ===== Lua Language =====
@@ Line 14: / Line 14: @@
 The nodeMCU firmware implements Lua 5.1 over the Espressif SDK for its ESP8266 SoC and the IoT modules based on this.
-  * The official lua.org [[http://www.lua.org/manual/5.1/manual.html|Lua Language specification]]  gives a terse but complete language specification.
+  * The official lua.org **[[http://www.lua.org/manual/5.1/manual.html|Lua Language specification]]**  gives a terse but complete language specification.
   * Its [[http://www.lua.org/faq.html|FAQ]]  provides information on Lua availability and licensing issues.
-  * The [[http://www.luafaq.org/|unofficial Lua FAQ]]  provides a lot of useful Q and A content.
+  * The **[[http://www.luafaq.org/|unofficial Lua FAQ]]** provides a lot of useful Q and A content, and is extremely useful for those learning Lua as a second language.
   * The [[http://lua-users.org/wiki/|Lua User's Wiki]]  gives useful example source and relevant discussion. In particular, its [[http://lua-users.org/wiki/Learning|Lua Learning Lua]]  section is a good place to start learning Lua.
-  * The best book to learn Lua is Programming in Lua by Roberto Ierusalimschy, who is one of the creators of Lua. It's first edition is available free [[http://www.lua.org/pil/contents.html|online]] . The second edition was aimed at Lua 5.1, but is out of print. The third edition is still in print and available in paperback.  It contains a lot more material and clearly identifies Lua 5.1 vs Lua 5.2 differences. This third is widely available for purchase and probably the best value for money.
+  * The best book to learn Lua is //Programming in Lua// by Roberto Ierusalimschy, one of the creators of Lua. It's first edition is available free [[http://www.lua.org/pil/contents.html|online]] . The second edition was aimed at Lua 5.1, but is out of print. The third edition is still in print and available in paperback.  It contains a lot more material and clearly identifies Lua 5.1 vs Lua 5.2 differences. **This third edition is widely available for purchase and probably the best value for money**. References of the format [PiL **n.m**] refer to section **n.m** in this edition.
   * The Espressif ESP8266 architecture is closed source, but the Espressif SDK itself is continually being updated so the best way to get the documentation for this is to [[https://www.google.co.uk/search?q=Espressif+IoT+SDK+Programming+Guide|google Espressif IoT SDK Programming Guide]] or to look at the Espressif [[http://bbs.espressif.com/viewforum.php?f=5|downloads forum]] .
-  * The [[http://www.nodemcu.com/docs/|nodeMCU documentation]]  is available online. However, please remember that the development team are based in China, and English is a second language, so the documentation needs expanding and be could improved with technical proofing.
+  * The **[[http://www.nodemcu.com/docs/|nodeMCU documentation]]** is available online. However, please remember that the development team are based in China, and English is a second language, so the documentation needs expanding and be could improved with technical proofing.
   * As with all Open Source projects the source for the nodeMCU firmware is openly available on the [[https://github.com/nodemcu/nodemcu-firmware|GitHub nodemcu-firmware]] repository.
 ==== How is nodeMCU Lua different to standard Lua? ====
-nodeMCU Lua in an implementation of [[http://www.eluaproject.net/overview|eLua]]  over the ESP8266 SDK. eLua is a full-featured implementation of Lua 5.1 that is optimized for embedded system development and execution to provide a scripting framework that can be used to deliver useful applications within the limited RAM and Flash memory resources of embedded processors such as the ESP8266.
+Whilst the Lua standard distribution includes a host stand-alone Lua interpreter, Lua itself is primarily an //extension language// that makes no assumptions about a "main" program: Lua works embedded in a host application to provide a powerful, light-weight scripting language for use within the application. This host application can then invoke functions to execute a piece of Lua code, can write and read Lua variables, and can register C functions to be called by Lua code. Through the use of C functions, Lua can be augmented to cope with a wide range of different domains, thus creating customized programming languages sharing a syntactical framework.
-A key goal of eLua is to reduce the RAM requirements for the Lua runtime system. One of the key techniques used in this implementation is to use read-only tables and constants wherever practical for library modules.  On a typical build this approach reduces the RAM footprint by some 20-25KB and this makes a Lua implementation for the ESP8266 feasible. This technique is called LTR and this is documented in detail in an eLua technical paper: [[http://www.eluaproject.net/doc/master/en_arch_ltr.html|Lua Tiny RAM]] .
+The ESP8266 was designed and is fabricated in China by [[http://espressif.com/new-sdk-release/|Espressif Systems]].  Espressif have also developed and released a companion software development kit (SDK) to enable developers to build practical [[wp>Internet of Things|IoT]] applications for the ESP8266.  The SDK is made freely available to developers in the form of binary libraries and SDK documentation.  However this is in a //closed format//, with no developer access to the source files, so ESP8266 applications //must// rely solely on the SDK API (and the somewhat Spartan SDK API documentation).
-The Espressif SDK is the interface that is freely available //albeit in closed format// to developers building applications for the ESP8266.  The nodeMCU eLua implementation must therefore use this SDK as its kernel layer and work within any design constraints that the SDK API imposes.  In particular, the SDK employs an event and task-oriented structure, where individual events can trigger an associated task; this task then runs to completion uninterrupted, at which point the next event queued can be initiated.  (Note that the SDK contains device drivers which are interrupt driven. However, these are internal to the SDK, so it treats all event triggered //application// tasks as atomic.)
+The nodeMCU Lua firmware is an ESP8266 application and must therefore be layered over the ESP8266 SDK. However, the hooks and features of Lua enable it to be seamlessly integrated without loosing any of the standard Lua language features.  The firmware has replaced some standard Lua modules that don't align well with the SDK structure with ESP8266-specific versions.  For example, the standard  ''io'' and ''os'' libraries don't work, but have been largely replaced by the nodeMCU ''node'' and ''file'' libraries.  The ''debug'' and ''math'' libraries have also been omitted to reduce the runtime footprint.
-The API calls for each type of event typically use a callback parameter to bind a C function implementing a given task to a given event.  In the case of the nodeMCU Lua implementation, this task is a wrapper around a developer-provided Lua function. This event-driven model imposed by the SDK is very different to a conventional procedural implementation of Lua.  Some standard Lua modules and eLua platform modules don't fit well within this structure, and so the nodeMCU implementation replaces these by ESP8266-specific versions.  For example, the standard  ''io'' and ''os'' libraries don't work, but have been largely replaced by the nodeMCU ''node'' and ''file'' libraries.
+NodeMCU Lua is based on [[http://www.eluaproject.net/overview|eLua]], a fully featured implementation of Lua 5.1 that has been optimized for embedded system development and execution to provide a scripting framework that can be used to deliver useful applications within the limited RAM and Flash memory resources of embedded processors such as the ESP8266. One of the main changes introduced in the eLua fork is to use read-only tables and constants wherever practical for library modules.  On a typical build this approach reduces the RAM footprint by some 20-25KB and this makes a Lua implementation for the ESP8266 feasible. This technique is called LTR and this is documented in detail in an eLua technical paper: [[http://www.eluaproject.net/doc/master/en_arch_ltr.html|Lua Tiny RAM]].
-The ''debug'' and ''math'' libraries have also been omitted to reduce the runtime footprint.
+The mains impacts of the ESP8266 SDK and together with its hardware resource limitations are not in the Lua language implementation itself, but in how //application programmers must approach developing and structuring their applications//. As discussed in detail below, the SDK is non-preemptive and event driven.  Tasks can be associated with given events by using the SDK API to registering callback functions to the corresponding events.  Events are queued internally within the SDK, and it then calls the associated tasks one at a time, with each task returning control to the SDK on completion. //The SDK states that if any tasks run for more than 10 mSec, then services such as Wifi can fail.//
+The nodeMCU libraries act as C wrappers around registered Lua callback functions to enable these to be used as SDK tasks. **//You must therefore use an [[wp>Event-driven programming|Event-driven programming]] style in writing your ESP8266 Lua programs//**.  Most programmers are used to writing in a procedural style where there is a clear single flow of execution, and the program interfaces to operating system services by a set of synchronous API calls to do network I/O, etc.  Whilst the logic of each individual task is procedural, this is not how you code up ESP8266 applications.
 ===== ESP8266 Specifics =====
@@ Line 44: / Line 46: @@
 ==== How is coding for the ESP8266 different to standard Lua? ====
-  *  The ESP8266 use onchip RAM and offchip Flash memory connected using a dedicated SPI interface.  Both of these are //very// limited (when compared to systems than most application programmer use).  The SDK and the Lua firmware already use the majority of this resource: the later build versions keep adding useful functionality, and unfortunately at an increased RAM and Flash cost, so depending on the build version and the number of modules installed the runtime can have as little as 17KB RAM and 40KB Flash available at an application level.  This Flash memory is formatted an made available as a **SPI Flash File System (SPIFFS)** through the ''file'' library.
+  *  The ESP8266 use onchip RAM and offchip Flash memory connected using a dedicated SPI interface.  Both of these are //very// limited (when compared to systems that most application programmer use).  The SDK and the Lua firmware already use the majority of this resource: the later build versions keep adding useful functionality, and unfortunately at an increased RAM and Flash cost, so depending on the build version and the number of modules installed the runtime can have as little as 17KB RAM and 40KB Flash available at an application level.  This Flash memory is formatted an made available as a **SPI Flash File System (SPIFFS)** through the ''file'' library.
   *  However, if you choose to use a custom build, for example one which uses integer arithmetic instead of floating point, and which omits libraries that aren't needed for your application, then this can help a lot doubling these available resources.  (See Marcel Stör's excellent [[http://frightanic.com/nodemcu-custom-build/|custom build tool]] that he discusses in [[http://www.esp8266.com/viewtopic.php?f=23&t=3001|this forum topic]]).  Even so, those developers who are used to dealing in MB or GB of RAM and file systems can easily run out of these resources.  Some of the techniques discussed below can go a long way to mitigate this issue.
   *  Current versions of the ESP8266 run the SDK over the native hardware so there is no underlying operating system to capture errors and to provide graceful failure modes, so system or application errors can easily "PANIC" the system causing it to reboot. Error handling has been kept simple to save on the limited code space, and this exacerbates this tendency. Running out of a system resource such as RAM will invariably cause a messy failure and system reboot.
@@ Line 56: / Line 58: @@
  node.restart(); for i = 1, 20 do print("not quite yet -- ",i); end
 </code>
-  * You therefore //have// to implement ESP8266 Lua applications using an [[wp>Event-driven programming|event driven]] structure.  You have to understand which SDK API requests schedule asynchronous processing, and which define event actions through Lua callbacks.  Yes, such an event-driven approach makes it difficult to develop procedurally structured applications, but it is well suited to developing the sorts of application that you will typically want to implement on an [[wp>Internet of Things|IoT]] device.
+  * You therefore //have// to implement ESP8266 Lua applications using an event driven approach.  You have to understand which SDK API requests schedule asynchronous processing, and which define event actions through Lua callbacks.  Yes, such an event-driven approach makes it difficult to develop procedurally structured applications, but it is well suited to developing the sorts of application that you will typically want to implement on an [[wp>Internet of Things|IoT]] device.
 ==== So how does the SDK event / tasking system work in Lua? ====
-Any SDK-based application for the ESP8266 uses a startup hook ''void user_init(void)'' defined by convention in the C module ''user_main.c''.  The system invokes this hook on boot. The ''user_init()'' function can by used to do any initialisation required and to call the necessary timer alarms or system functions to bind and callback routines to implement the tasks needed in response to any system events. Individual task callbacks need to implement their actions and return control to the SDK as soon as practical, as the SDK framework is not pre-emptive so any further event tasks are queued on a pending list within the SDK kernel.
+  * The SDK employs an event-driven and task-oriented architecture for programming at an applications level.
+  * The SDK uses a startup hook ''void user_init(void)'', defined by convention in the C module ''user_main.c'', which it invokes on boot. The ''user_init()'' function can be used to do any initialisation required and to call the necessary timer alarms or other SDK API calls to bind and callback routines to implement the tasks needed in response to any system events.
+  * The API provides a set of functions for declaring application functions (written in C) as callbacks to associate application tasks with specific hardware and timer events.  These are non-preemptive at an applications level.
+  * Whilst the SDK provides a number of interrupt driven device drivers, the hardware architecture severely limits the memory available for these drivers, so writing new device drivers is not a viable options for most developers
+  * The SDK interfaces internally with hardware and device drivers to queue pending events.
+  * The registered callback routines are invoked sequentially with the associated C task running to completion uninterrupted.
+  * In the case of Lua, these C tasks are typically functions within the Lua runtime library code and these typically act as C wrappers around the corresponding developer-provided Lua callback functions.  An example here is the Lua ''tmr.alarm(id, interval, repeat, callback)'' function.  The calls a function in the ''tmr'' library which registers a C function for this alarm using the SDK, and when this C function is called it then invokes the Lua callback.
-Excessively long-running tasks can therefore cause other system functions and services to timeout, or allocate memory to buffer queued data, which can then trigger either the watchdog timer or memory exhaustion, both of which will ultimately cause the system to reboot.
+The nodeMCU firmware simply mirrors this structure at a Lua scripting level:
+  * A startup module ''init.lua'' is invoked on boot. This function module can be used to do any initialisation required and to call the necessary timer alarms or libary calls to bind and callback routines to implement the tasks needed in response to any system events.
+  * The Lua libraries provide a set of functions for declaring application functions (written in Lua) as callbacks (which are stored in the [[#So how is the Lua Registry used and why is this important?|Lua registry]]) to associate application tasks with specific hardware and timer events.  These are non-preemptive at an applications level.
+  * The Lua libraries work in concert with the SDK to queue pending events and invoke any registered Lua callback routines, which then run to completion uninterrupted.
+  * Excessively long-running Lua functions can therefore cause other system functions and services to timeout, or allocate memory to buffer queued data, which can then trigger either the watchdog timer or memory exhaustion, both of which will ultimately cause the system to reboot.
+  * By default, the Lua runtime also 'listens' to UART 0, the serial port, in interactive mode and will execute any Lua commands input through this serial port.
-SDK Callbacks include:
+This event-driven approach is very different to a conventional procedural implementation of Lua.
-  *  Timer alarm callbacks
-  *  Wifi scan callbacks
-  *  Network (ESPCONN) callbacks for connection, disconnect, send, receive, etc. (roughly equivalent the ''socket:on()'' callbacks in Lua)
-  *  GPIO and other hardware related interrupts.
-The eLua implementation sits within this framework:
-  *  ''app/user/user_main.c'' contains the ''user_init()'' entry point.  This reinitialises the UART, the volatile sections of flash memory (if necessary), the RomFS and SPIFFS before calling ''lua_main()'' with the command-line ''lua -i''.
-  *  The Lua RTS (see ''app/lua/lua.c'') then sets up a timer to poll the input UART every 80 mSec to assemble a complete execution chunk which it then executes with a ''lua_pcall()''.
-  *  The running Lua script can initialise one or more callbacks associated with events such as a timer.  The library code will typically store the link to this Lua callback function in the [[#So how is the Lua Registry used and why is this important?|Lua registry]] .  When the callback hook is subsequently invoked, this hook code then retrieves this function reference from the registry and executes it with a ''lua_call()''.
-  *  There are no concurrency or interlock issues with this approach as the SDK will only initiate a callback after the previously running task has completed, and in the case of Lua when the previous Lua chunk has completed -- Lua chunks are executed one-at-a-time.
-Consider an simple telnet example given in ''examples/fragment.lua'':
+Consider a simple telnet example given in ''examples/fragment.lua'':
 <code Lua>
  s=net.createServer(net.TCP)
@@ Line 94: / Line 97: @@
  end)
 </code>
-This example doesn't use upvalues and all declarations are global, so we can reorder this code for clarity (though doing this adds a few extra globals):
-<code Lua>
+This example defines five Lua functions:
- function c_receive(c,l)
-   node.input(l)
+^      Function        ^      Defined in      ^     Parameters        ^ Callback? ^
- end
+| Main                 | Outer module         | ... (Not used)        |           |
- function c_disconnection(c)
+| Connection listener  | Main                 | c (connection socket) |           |
-   con_std = nil
+| s_output             | Connection listener  | str                   |    Yes    |
-   node.output(nil)
+| On Receive           | Connection listener  | c, l (socket, input)  |    Yes    |
- end
+| On Disconnect        | Connection listener  | c (socket)            |    Yes    |
- function s_output(str)
-   if(con_std~=nil) then
+''s'', ''con_std'' and ''s_output'' are global, and no [[#Why is it importance to understand how upvalues are implemented when programming for the ESP8266?|upvalues]] are used.  There is no "correct" order to define these in, but we could reorder this code for clarity (though doing this adds a few extra globals) and define these functions separately one another.   However, let us consider how this is executed:
-     con_std:send(str)
-   end
+  * The outer module is compiled including the four internal functions.
- end
+  * ''Main'' is then assigning the created ''net.createServer()'' to the global ''s''.  The ''connection listener'' closure is created and bound to a temporary variable which is then passed to the ''socket.listen()'' as an argument.  The routine then exits returning control to the firmware.
- function s_listen(c)
+  *  When another computer connects to port 23, the listener handler retrieves the reference to then connection listener and calls it with the socket parameter.  This function then binds the s_output closure to the global ''s_output'', and registers this function with the ''node.output'' hook. Likewise the ''on receive'' and ''on disconnection'' are bound to temporary variables which are passed to the respective on handlers.  We now have four Lua functions registered in the Lua runtime libraries associated with four events.  This routine then exits returning control to the firmware.
-   con_std = c
+  *  When a record is received, the on receive handler within the net library retrieves the reference to the ''on receive'' Lua function and calls it passing it the record. This routine then passes this to the ''node.input()'' and exits returning control to the firmware.
-   node.output(s_output, 0)
+  *  The ''node.input'' handler polls on an 80 mSec timer alarm. If a compete Lua chunk is available (either via the serial port or node input function), then it executes it and any output is then passed to the ''note.output'' handler. which calls ''s_output'' function.  Any pending sends are then processed.
-   c:on("receive",c_receive)
+  *  This cycle repeats until the other computer disconnects, and ''net'' library  disconnection handler then calls the Lua  ''on disconnect'' handler. This Lua routine dereferences the connected socket and closes the ''node.output'' hook and exits returning control to the disconnect handler which garbage collects any associated sockets and registered on handlers.
-   c:on("disconnection",c_disconnection)
- end
+Whilst this is all going on, The SDK can (and often will) schedule other event tasks in between these Lua executions (e.g. to do the actual TCP stack processing).  The longest individual Lua execution in this example is only 18 bytecode instructions (in the main routine).
- s=net.createServer(net.TCP)
- s:listen(23,s_listen)
-</code>
-So let us consider how this is executed:
-  *  The main routine executes defining 4 functions in the global variables: ''c_receive'', ''c_disconnection'', ''s_output'', ''s_listen''; the server ''s'' is bound to port 23 registering ''s_listen'' as the initialisation callback. The main routine then exits, with the global variables retained and the main routine code garbage collected.
-  *  When another computer connects to port 23, the listener handler retrieves the reference to ''s_listen'' from the registry and calls it with the socket parameter.  This function then binds ''s_output'' to the ''node.output'' hook registering it in the registry, and likewise the ''c_receive'' and ''c_disconnection'' are bound and registered to the respective on handlers.  We now have four routines registered in the registry associated with four events, and this routine then exits with only the routines execution frame garbage collected.
-  *  When a record is received, the ''on_receive'' handler retrieves the reference to ''c_receive'' from the registry and calls it passing it the record. This routine then passes this to the ''node.input()'' and exits. (The node input handler marshals these records into a complete Lua chunk).
-  *  The node input handler is polling on an 80 mSec alarm and if a compete Lua chunk is available, it executes it.  Any output is then passed to the note.output handler which retrieves and calls ''s_output'' which exits on completion.  Any pending sends are then processed.
-  *  This cycle repeats until the other computer disconnect which triggers the on_disconnect handler.  This retrieves the ''c_disconnection'' reference from the registry and calls it. This routine dereferences the connected socket and closes the ''node.output'' hook and exits returning control to the disconnect handler which garbage collects any associated sockets and registered on handlers.
-The SDK can and will often schedule other event tasks in between these Lua executions (e.g. to do the actual TCP stack processing).  The longest individual Lua execution in this example is only 20 bytecode instructions (in the main routine). The original version was a few instructions shorter in that temporary locals were used to hold the closure references instead of globals, but the runtime and memory footprint aren't materially different.
 Understanding how the system executes your code can help you structure it better and improve memory usage. Each event task is established by a callback in an API call in an earlier task.
+==== So what Lua library functions enable the registration of Lua callbacks? ====
+SDK Callbacks include:
+^ Lua Module ^ Functions which define or remove callbacks                                   ^
+| tmr        | ''alarm(id, interval, repeat, function())''                                  |
+| node       | ''key(type, function())'', ''output(function(str), serial_debug)''           |
+| wifi       | ''startsmart(chan, function())'', ''sta.getap(function(table))''             |
+| net.server | ''sk:listen(port,[ip],function(socket))''                                    |
+| net        | ''sk:on(event, function(socket, [, data]))'', ''sk:send(string, function(sent))'', ''sk:dns(domain, function(socket,ip))'' |
+| gpio       | ''trig(pin, type, function(level))''                                         |
+| mqqt       | ''client:m:on(event, function(conn[, topic, data])''                         |
+| uart       | ''uart.on(event, cnt, [function(data)], [run_input])''                       |
 ==== So how is context passed between Lua event tasks? ====
   *  It is important to understand that any event callback task is associated with a single Lua function. This function is executed from the relevant nodeMCU library C code using a ''lua_call()''. Even system initialisation which executes the ''dofile("init.lua")'' can be treated as a special case of this.  Each function can invoke other functions and so on, but it must ultimate return control to the C library code.
-  *  By their very nature Lua ''local'' variables only exit within the context of an executing Lua function, and so all locals are destroyed between these ''lua_call()'' actions. //No locals are retained across events//.
+  *  By their very nature Lua ''local'' variables only exist within the context of an executing Lua function, and so all locals are destroyed between these ''lua_call()'' actions. //No locals are retained across events//.
   *  So context can only be passed between event routines by one of three mechanisms:
     *  **Globals** are by nature globally accessible.  Any global will persist until explicitly dereference by reassigning ''nil'' to it.  Globals can be readily enumerated by a ''for k,v in pairs(_G) do'' so their use is transparent.
-    *  **Upvalues** are the Lua mechanism for implementing inheritance during a function closure.  The Lua runtime system does  behind-the-scenes magic to remap any upvalues into a hidden linked list when the calling stack frame is exited, so they just work as you would expect them to.  The hidden values will be garbage collected correctly when fully dereferenced.  However, some API calls don't correctly dereference expired callback references and as a result upvalues may not be correctly garbage collected and manifest themselves as memory leaks. So using them can cause more frequent and difficult to diagnose PANICs during testing. So my general recommendation is to stick to globals for this specific usecase of passing context between event callbacks, and nil them when done.
+    * The **File system** is a special case of persistent global, so there is no reason in principle why it can't be used to pass context.  However the ESP8266 file system uses flash memory and this has a limited write cycle lifetime, so it is best to avoid using the file system to store frequently changing content except as a mechanism of last resort.
-    * The **File system** is a special case of persistent global, so there is no reason why it can't be used to pass context, in principle.  However flash memory has a limited write cycle lifetime, so it is best to limit using file content that frequently changes to a mechanism of last resort.
+    *  **Upvalues**.  When a function is declared within an outer function, all of the local variables in the outer scope are available to the inner function.  Since all functions are stored by reference the scope of the inner function might outlast the scope of the outer function, and the Lua runtime system ensures that any such references persist for the life of any functions that reference it.  This standard feature of Lua is known as //closure// and is  described in [Pil 6].  Such values are often called //upvalues//.  Functions which are global or [[#So how is the Lua Registry used and why is this important?|registered]] callbacks will persist between event routines, and hence any upvalues referenced by them can be used for passing context.
+==== So how is the Lua Registry used and why is this important? ====
+So all Lua callbacks are called by C wrapper functions that are themselves callback activated by the SDK as a result of a given event.  Such C wrapper functions themselves frequently need to store state for passing between calls or to other wrapper C functions.  The Lua registry is simply another Lua table which is used for this purpose, except that it is hidden from direct Lua access.  Any content that needs to be saved is created with a unique key. Using a standard Lua table enables standard garbage collection algorithms to operate on its content.
+Note that we have identified a number of cases where library code does not correctly clean up Registry content when closing out an action, leading to memory leaks.
+==== Why is it importance to understand how upvalues are implemented when programming for the ESP8266? ====
+Routines directly or indirectly referenced in the globals table, **_G**, or in the Lua Registry may use upvalues. The number of upvalues associated with a given routine is determined by the compiler and a vector is allocated when the closure is bound to hold these references.  Each upvalues is classed as open or closed. All upvalues are initially open which means that the upvalue references back to the outer functions's register set.  However, upvalues must be able to outlive the scope of the outer routine where they are declared as a local variable.  The runtime VM does this by adding extra checks when executing a function return to scan any defined closures within its scope for back references and allocate memory to hold the upvalue and points the upvalue's reference to this.  This is known as a closed upvalue.
+This processing is a mature part of the Lua 5.x runtime system, and for normal Lua applications development this "behind-the-scenes" magic ensures that upvalues just work as any programmer might expect.  Sufficient garbage collector metadata is also stored so that these hidden values will be garbage collected correctly //when properly dereferenced//. However allocating these internal structures is quite expensive in terms of memory, and this hidden overhead is hard to track or to understand.  If you are developing a Lua application for a PC where the working RAM for an application is measured in MB, then this isn't really an issue.  However, if you are developing an application for the ESP8266 where you might have 20 KB for your program and data, this could prove a killer.
+One further complication is that some library functions don't correctly dereference expired callback references and as a result their upvalues may not be correctly garbage collected (though we are tracking this down and hopefully removing this issue).  This will all be manifested as a memory leak. So using upvalues can cause more frequent and difficult to diagnose PANICs during testing. So my general recommendation is still to stick to globals for this specific usecase of passing context between event callbacks, and ''nil'' them when you have done with them.
 ==== Can I encapsulate actions such as sending an email in a Lua function? ====
-Think about the implications of the last Q&A.
+Think about the implications of these last few answers.
   * An action such as composing and sending an email involves a message dialogue with a mail server over TCP.  This in turn requires calling multiple API calls to the SDK and your Lua code must return control to the C calling library for this to be scheduled, otherwise these requests will just queue up, you'll run out of RAM and your application will PANIC.
   * Hence it is simply **//impossible//** to write a Lua module so that you can do something like:
@@ Line 146: / Line 168: @@
   -- move on to next phase of processing.
 </code>
-  * But you could code up a event-driven task to do this and pass it a callback to be executed on completion of the mail send, something along the lines of:
+  * But you could code up a event-driven task to do this and pass it a callback to be executed on completion of the mail send, something along the lines of the following.  Note that since this involves a lot of asynchronous processing and which therefore won't take place until you've returned control to the calling library C code, you will typically execute this as the last step in a function and therefore this is best done as a tailcall [PiL 6.3].
 <code Lua>
   -- prepare message
   local ms = require("mail_sender")
-  ms.send(to, subject, body, function(status) (require("process_next")).init(status) end)
+  return ms.send(to, subject, body, function(status) loadfile("process_next.lua")(status) end)
-return
 </code>
   * Building an application on the ESP8266 is a bit like threading pearls onto a necklace.  Each pearl is an event task which must be small enough to run within its RAM resources and the string is the variable context that links the pearls together.
-==== So how is the Lua Registry used and why is this important? ====
+==== When and why should I avoid using tmr.delay()? ====
-So all Lua callbacks are called by C wrapper functions that are themselves callback activated by the SDK as a result of a given event.  Such C wrapper functions themselves frequently need to store state for passing between calls or to other wrapper C functions.  The Lua registry is simply another Lua table which is used for this purpose, except that it is hidden from direct Lua access.  Any content that needs to be saved is created with a unique key. Using a standard Lua table enables standard garbage collection algorithms to operate on its content.
-Note that we have identified a number of cases where library code does not correctly clean up Registry content when closing out an action, leading to memory leaks.
-==== When and why should I avoid using tmr.delay()? ====
 If you are used coding in a procedural paradigm then it is understandable that you consider using ''tmr.delay()'' to time sequence your application.  However as discussed in the previous section, with nodeMCU Lua you are coding in an event-driven paradigm.
 If you look at the ''app/modules/tmr.c'' code for this function, then you will see that it executes a low level  ''ets_delay_us(delay)''.  This function isn't part of the nodeMCU code or the SDK; it's actually part of the xtensa-lx106 boot ROM, and is a simple timing loop which polls against the internal CPU clock.  It does this with interrupts disabled, because if they are enabled then there is no guarantee that the delay will be as requested.
-''tmr.delay()'' should be correctly used if you want to have exact timing control on an external hardware I/O (e.g. lifting a GPIO pin high for 20  μSec).  It will achieve no functional purpose in pretty much every other usecase, as any other system code-based activity will be blocked from execution; at worst it will break your the code and create hard-to-diagnose timeout errors. A good indication here is if you want a delay of more than 10 mSec or so, then //using ''tmr.delay()'' is the wrong approach//.  You should be using a timer alarm or other library callback, to allow the other processing to take place.  As the nodeMCU documentation correctly advises (translating Chinese English in to English): //''tmr.delay()'' will make the CPU work in non-interrupt mode, so other instructions and interrupts will be blocked. Take care in using this function.//
+''tmr.delay()'' is really intended to be used where you need to have more precise timing control on an external hardware I/O (e.g. lifting a GPIO pin high for 20  μSec).  It will achieve no functional purpose in pretty much every other usecase, as any other system code-based activity will be blocked from execution; at worst it will break your application and create hard-to-diagnose timeout errors.
+The latest SDK includes a caution that if any (callback) task runs for more than 10 mSec, then the Wifi and TCP stacks might fail, so if you want a delay of more than 8 mSec or so, then //using ''tmr.delay()'' is the wrong approach//.  You should be using a timer alarm or another library callback, to allow the other processing to take place.  As the nodeMCU documentation correctly advises (translating Chinese English into English): //''tmr.delay()'' will make the CPU work in non-interrupt mode, so other instructions and interrupts will be blocked. Take care in using this function.//
 ==== How do I avoid a PANIC loop in init.lua? ====
@@ Line 171: / Line 191: @@
 Most of us have fallen into the trap of creating an ''init.lua'' that has a bug in it, which then causes the system to reboot and hence gets stuck in a reboot loop.  If you haven't then you probably will do so at least once.
   * When this happens, the only robust solution is to reflash the firmware.
-  * The simplest way to avoid having to do this is to keep the ''init.lua'' as simple as possible -- say configure the wifi and then start your app on a one-shot ''tmr.alarm()'' after a 2-3 sec delay.  This delay is long enough to issue a ''file.remove("init.lua")'' through the serial port and recover control that way.
+  * The simplest way to avoid having to do this is to keep the ''init.lua'' as simple as possible -- say configure the wifi and then start your app using a one-time ''tmr.alarm()'' after a 2-3 sec delay.  This delay is long enough to issue a ''file.remove("init.lua")'' through the serial port and recover control that way.
-  * Also always test any new ''init.lua'' by creating it as ''init_test.lua'', say, and manually issuing a ''dofile("init_test.lua")'' through the serial port, and then only rename it when you are certain it is working as you require.
+  * Also it is always best to test any new ''init.lua'' by creating it as ''init_test.lua'', say, and manually issuing a ''dofile("init_test.lua")'' through the serial port, and then only rename it when you are certain it is working as you require.
 ===== Techniques for Reducing RAM and SPIFFS footprint =====
-==== How do I minimise the footprint of an application ====
+==== How do I minimise the footprint of an application? ====
-  * Perhaps the simplest aspect of reducing the footprint of an application is to get its scope correct.  The ESP8266 is an IoT device and not a general purpose system.  It is typically used to attach to real-world monitors, controls, etc. to an intranet and is therefore design to implement functions that have limited scope.  We commonly come across developers who are trying to treat the ESP8266 as a general purpose device and can't understand why their application can't run.
+  * Perhaps the simplest aspect of reducing the footprint of an application is to get its scope correct.  The ESP8266 is an IoT device and not a general purpose system.  It is typically used to attach real-world monitors, controls, etc. to an intranet and is therefore designed to implement functions that have limited scope.  We commonly come across developers who are trying to treat the ESP8266 as a general purpose device and can't understand why their application can't run.
-  * The simplest and safest way to use IoT devices is to control them through a dedicated general purpose system on the same network.  This could be a low cost system such as a [[https://www.raspberrypi.org/|RasberryPi (RPi)]] server, running your custom code or an open source [[wp>home automation|home automation (HA)]] application. Such systems have orders of magnitude more capacity than the ESP8266, for example the RPi has 2GB RAM and its SD card can be up to 32GB in capacity, and it can support the full range of USB-attached disk drives and other devices.  It also runs a fully featured Linux OS, and has a rich selection of applications pre configured for it. There are plenty of alternatives system available in this under $50 price range, as well as proprietary HA systems which can cost 10-50 times more.
+  * The simplest and safest way to use IoT devices is to control them through a dedicated general purpose system on the same network.  This could be a low cost system such as a [[https://www.raspberrypi.org/|RaspberryPi (RPi)]] server, running your custom code or an open source [[wp>home automation|home automation (HA)]] application. Such systems have orders of magnitude more capacity than the ESP8266, for example the RPi has 2GB RAM and its SD card can be up to 32GB in capacity, and it can support the full range of USB-attached disk drives and other devices.  It also runs a fully featured Linux OS, and has a rich selection of applications pre configured for it. There are plenty of alternative systems available in this under $50 price range, as well as proprietary HA systems which can cost 10-50 times more.
   * Using a tiered approach where all user access to the ESP8266 is passed through a controlling server means that the end-user interface (or smartphone connector), together with all of the associated validation and security can be implemented on a system designed to have the capacity to do this.  This means that you can limit the scope of your ESP8266 application to a limited set of functions being sent to or responding to requests from this system.
-  * //If you are trying to implement a user-interface or HTTP webserver in your ESP8266 then you are really abusing its intended purpose.  When it come to scoping your ESP8266 applications, the adage **K**eep **I**t **S**imple, **S**tupid truly applies.//
+  * //If you are trying to implement a user-interface or HTTP webserver in your ESP8266 then you are really abusing its intended purpose.  When it comes to scoping your ESP8266 applications, the adage **K**eep **I**t **S**imple **S**tupid truly applies.//
 ==== How do I minimise the footprint of an application on the file system ====
@@ Line 190: / Line 210: @@
     * Keep a master repository of your code on your PC or a cloud-based versioning repository such as [[https://github.com/|GitHub]]
     * Lay it out and comment it for ease of maintenance and debugging
-    * Use a package use as [[https://github.com/4refr0nt/ESPlorer|Esplorer]] to download modules that you are debugging and to test them.
+    * Use a package such as [[https://github.com/4refr0nt/ESPlorer|Esplorer]] to download modules that you are debugging and to test them.
     * Once the code is tested and stable, then compress it using LuaSrcDiet before downloading to the ESP8266.  Doing this will reduce the code footprint on the SPIFFS by 2-3x.
   * Consider using ''node.compile()'' to pre-compile any production code.  This removes the debug information from the compiled code reducing its size by roughly 40%.  (However this is still perhaps 1.5-2x larger than a LuaSrcDiet-compressed source format, so if SPIFFS is tight then you might consider leaving less frequently run modules in Lua format. If you do a compilation, then you should consider removing the Lua source copy from file system as there's little point in keeping both on the ESP8266.
-==== How do I minimise the footprint of running application ====
+==== How do I minimise the footprint of running application? ====
-  * The Lua Garbage collector is very aggressive at scanning and recovering dead resources.  It use an incremental mark-and-sweep strategy which means that any data which is not ultimately referenced back to the Globals table, the Lua registry or in-scope local variables in the current Lua code will be collected.
+  * The Lua Garbage collector is very aggressive at scanning and recovering dead resources.  It uses an incremental mark-and-sweep strategy which means that any data which is not ultimately referenced back to the Globals table, the Lua registry or in-scope local variables in the current Lua code will be collected.
-  * Setting any variable to ''nil'' dereferences the previous context of that variable.  (Note that reference-based variables such as tables, strings and functions can have multiple variables referencing the same object, but once the last reference has set to ''nil'', the collector will recover the storage.
+  * Setting any variable to ''nil'' dereferences the previous context of that variable.  (Note that reference-based variables such as tables, strings and functions can have multiple variables referencing the same object, but once the last reference has been set to ''nil'', the collector will recover the storage.
-  * Unlike other compile-on-load languages such as PHP, Lua compiled code is treaded the same way as any other variable type when it comes to garbage collection and can be collected when fully dereferenced, so that the code-space can be reused.
+  * Unlike other compile-on-load languages such as PHP, Lua compiled code is treated the same way as any other variable type when it comes to garbage collection and can be collected when fully dereferenced, so that the code-space can be reused.
-  * Lua execution is intrinsically divided into separate event tasks with each associated with a Lua callback.  This, when coupled with the strong dispose on dereference feature, means that it is very easy to structure your application using an classic technique which dates back to the 1950s know as [[wp>Overlay (programming)|Overlays]].
+  * Lua execution is intrinsically divided into separate event tasks with each bound to a Lua callback.  This, when coupled with the strong dispose on dereference feature, means that it is very easy to structure your application using an classic technique which dates back to the 1950s known as [[wp>Overlay (programming)|Overlays]].
-  * There are various approaches to implementing this.  One is described by DP Whittaker in his [[http://www.esp8266.com/viewtopic.php?f=19&t=1940|Massive memory optimization: flash functions]] topic.  Another is to use //volatile modules//.  There are standard Lua templates for creating modules, but the ''require()'' library function creates a reference for the loaded module in the ''package.loaded'' table, and this reference prevents the module being garbage collected.  To make a module volatile, you should remove this reference by setting it to ''nil''.  You can't do this in the outermost level of the module (since the reference is only created once execution has returned from the module code), but you can do it in any module function, and typically an initialisation function for the module, as in the following example:
+  * Various approaches can be use to implement this.  One is described by DP Whittaker in his [[http://www.esp8266.com/viewtopic.php?f=19&t=1940|Massive memory optimization: flash functions]] topic.  Another is to use //volatile modules//.  There are standard Lua templates for creating modules, but the ''require()'' library function creates a reference for the loaded module in the ''package.loaded'' table, and this reference prevents the module from being garbage collected.  To make a module volatile, you should remove this reference to the loaded module by setting its corresponding entry in ''package.loaded'' to ''nil''.  You can't do this in the outermost level of the module (since the reference is only created once execution has returned from the module code), but you can do it in any module function, and typically an initialisation function for the module, as in the following example:
 <code Lua>
   -- . . .
   local s=net.createServer(net.TCP)
-  s:listen(80,function(c) (require("connector")).init(c) end)
+  s:listen(80,function(c) require("connector").init(c) end)
 </code>
   * **''connector.lua''** would be a standard module pattern except that the ''M.init()'' routine must include the lines
@@ Line 218: / Line 238: @@
 </code>
   * This approach ensures that the module can be fully dereferenced on completion.  OK, in this case, this also means that the module has to be reloaded on each TCP connection to port 80; however, loading a compiled module from SPIFFS only takes a few mSec, so surely this is an acceptable overhead if it enables you to break down your application into RAM-sized chunks.  Note that ''require()'' will automatically search for ''connector.lc'' followed by ''connector.lua'', so the code will work for both source and compiled variants.
+  * Whilst the general practice is for a module to return a table, [PiL 15.1] suggests that it is sometimes appropriate to return a single function instead as this avoids the memory overhead of an additional table. This pattern would look as follows:
+<code Lua>
+  -- . . .
+  local s=net.createServer(net.TCP)
+  s:listen(80,function(c) require("connector")(c) end)
+</code>
+<code Lua>
+  local module = ...  -- this is a situation where using an upvalue is essential!
+  return function (csocket)
+    package.loaded[module]=nil
+    module = nil
+    -- . . .
+  end
+</code>
+  * Also note that you should //**not**// normally code this up listener call as the following because the RAM now has to accommodate both the module which creates the server //and// the connector logic.
+<code Lua>
+  -- . . .
+  local s=net.createServer(net.TCP)
+  local connector = require("connector") -- don't do this unless you've got the RAM available!
+  s:listen(80,connector)
+</code>
 ==== How do I reduce the size of my compiled code? ====
@@ Line 225: / Line 266: @@
   - The second is to use ''loadfile()'' to load the source file into memory, followed by ''string.dump()'' to convert it in-memory to a serialised load format which can then be written back to a ''.lc'' file.  This approach creates a bytecode file which retains the debug information.
-The memory footprint the bytecode created by method (2) is the same as when executing source files directly, but the footprint of bytecode created by method (1) is typically **60% of this size**, because the debug information is almost as large as the code itself.  So using ''.lc'' files generated by ''node.compile()'' considerably reduces code size in memory -- albeit with the downside that any runtime errors are extremely limited.
+The memory footprint of the bytecode created by method (2) is the same as when executing source files directly, but the footprint of bytecode created by method (1) is typically **60% of this size**, because the debug information is almost as large as the code itself.  So using ''.lc'' files generated by ''node.compile()'' considerably reduces code size in memory -- albeit with the downside that any runtime errors are extremely limited.
 In general consider method (1) if you have stable production code that you want to run in as low a RAM footprint as possible.  Yes, method (2) can be used if you are still debugging, but you will probably be changing this code quite frequently, so it is easier to stick with ''.lua'' files for code that you are still developing.
@@ Line 233: / Line 274: @@
 ==== How do I get a feel for how much memory my functions use? ====
-  *  Given the limited resources available to applications it is highly desirable that by you understand the VM model.  The essential reference here is [[http://luaforge.net/docman/83/98/ANoFrillsIntroToLua51VMInstructions.pdf|A No Frills Introduction to Lua 5.1 VM Instructions]] .  This explain how the code generator works, how much memory overhead is involved with each table, function, string etc..
+  *  You should get an overall understanding of the VM model if you want to make good use of the limited resources available to Lua applications.  An essential reference here is [[http://luaforge.net/docman/83/98/ANoFrillsIntroToLua51VMInstructions.pdf|A No Frills Introduction to Lua 5.1 VM Instructions]] .  This explain how the code generator works, how much memory overhead is involved with each table, function, string etc..
   *  You can't easily get a bytecode listing of your ESP8266 code; however there are two broad options for doing this:
-    * **Generate a bytecode listing on your development PC**. The Lua 5.1 code generator is basically the same on the PC and on the ESP8266, so whilst it isn't identical, using the standard Lua batch compiler ''luac'' against your source on your PC with the ''-l -s'' option will give you a good idea of what your code will generate.  The main difference between these two variants is the size_t for ESP8266 is 4 bytes rather than 8bytes found on modern 64bit development PCs; and the eLua variants generate different access references for ROM data types.  If you want to see what the ''string.dump()'' version generates then drop the ''-s'' option to retain the debug information.
+    * **Generate a bytecode listing on your development PC**. The Lua 5.1 code generator is basically the same on the PC and on the ESP8266, so whilst it isn't identical, using the standard Lua batch compiler ''luac'' against your source on your PC with the ''-l -s'' option will give you a good idea of what your code will generate.  The main difference between these two variants is the size_t for ESP8266 is 4 bytes rather than the 8 bytes size_t found on modern 64bit development PCs; and the eLua variants generate different access references for ROM data types.  If you want to see what the ''string.dump()'' version generates then drop the ''-s'' option to retain the debug information.
-    * **Upload your ''.lc'' files to the PC and disassemble then there**. There are a number of Lua code disassemblers which can list off the compiled code that you application modules will generate, ''if'' you have a script to upload files from your ESP8266 to your development PC.  I use [[http://luaforge.net/projects/chunkspy/|ChunkySpy]]  which can be downloaded [[http://files.luaforge.net/releases/chunkspy/chunkspy/ChunkSpy-0.9.8/ChunkSpy-0.9.8.zip|here]] , but you will need to apply the following patch so that ChunkySpy understands eLua data types:
+    * **Upload your ''.lc'' files to the PC and disassemble then there**. There are a number of Lua code disassemblers which can list off the compiled code that you application modules will generate, ''if'' you have a script to upload files from your ESP8266 to your development PC.  I use [[http://luaforge.net/projects/chunkspy/|ChunkSpy]]  which can be downloaded [[http://files.luaforge.net/releases/chunkspy/chunkspy/ChunkSpy-0.9.8/ChunkSpy-0.9.8.zip|here]] , but you will need to apply the following patch so that ChunkSpy understands eLua data types:
 <code diff>
  --- a/ChunkSpy-0.9.8/5.1/ChunkSpy.lua   2015-05-04 12:39:01.267975498 +0100
@@ Line 275: / Line 316: @@
 Of course you should still use functions to structure your code and encapsulate common repeated processing, but just bear in mind that each function definition has a relatively high overhead for its header record and stack frame (compared to the 20 odd KB RAM available).  //So try to avoid overusing functions. If there are less than a dozen or so lines in the function then you should consider putting this code inline if it makes sense to do so.//
-==== What other resources are available ====
+==== What other resources are available? ====
+  *  Install lua and luac on your development PC.  This is freely available for Windows, Mac and Linux distributions, but we strongly suggest that you use Lua 5.1 to maintain source compatibility with ESP8266 code.  This will allow you not only to unit test some modules on your PC in a rich development environment, but you can also use ''luac'' to generate a bytecode listing of your code and to validate new code syntactically before downloading to the ESP8266.  This will also allow you to develop server-side applications and embedded applications in a common language.
+===== Firmware and Lua app development =====
+==== How to save memory? ====
+  * The NodeMCU development team recommends that you consider using a tailored firmware build, which only includes the modules that you plan to use before developing any Lua application. Once you have the ability to make and flash custom builds, the you also have the option of moving time sensitive or logic intensive code into your own custom module. Doing this can save a large amount of RAM as C code can be run directly from Flash memory. If you want an easy-to-use intermediate option then why note try the [[http://frightanic.com/nodemcu-custom-build|Cloud based nodeMCU custom build service]]?.
+===== Hardware Specifics =====
+==== Why file writes fail all the time on DEVKIT V1.0? ====
+  * NodeMCU DEVKIT V1.0 uses ESP12-E-DIO(ESP-12-D) module.  This module runs the Flash memory in [[#What's the different between DIO and QIO mode?|Dual IO SPI]] (DIO) mode. This firmware will not be correctly loaded if you uses old flashtool version, and the filesystem will not work if you used a pre 0.9.6 firmware version (<0.9.5) or old. The easiest way to resolve this problem s update all the firmware and flash tool to  current version.
+    - Use the latest [[https://github.com/themadinventor/esptool|esptool.py]] with DIO support and command option to flash firmware, or
+    - Use the latest [[https://github.com/nodemcu/nodemcu-flasher|NodeMCU flasher]] with default option. (You must select the ''restore to default'' option in advanced menu tab), or
+    - Use the latest Espressif's flash tool -- see [[http://bbs.espressif.com/viewtopic.php?f=5&t=433|this Espressif forum topic]] (without auto download support). Use DIO mode and 32M flash size option, and flash latest firmware to 0x00000. Before flashing firmware, remember to hold FLASH button, and press RST button once. Note that the new nodeMCU our firmware download tool, when released, will be capable of flashing firmware automatically without any button presses.
+==== What's the different between DIO and QIO mode? ====
+Whether DIO or QIO modes are available depends on the physical connection between the ESP8266 CPU and its onboard flash chip. QIO connects to the flash using 6 pins as compared to DIO's 4, and QIO is twice as fast to read/write as DIO.
+==== How to use DEVKIT V0.9 on Mac OS X? ====
+<TODO>
-  *  Install lua and luac on your development PC.  This is freely available in Windows, Mac and Linux distributions, but we strongly suggest that you use Lua 5.1 to maintain source compatibility with ESP8266 code.  This will allow you not only to unit test some modules on your PC in a rich development environment, but you can also use ''luac'' to list off bytecode listing of your code and syntactically validate new code before downloading to the ESP8266.  This will also allow you to develop server-side applications and embedded applications in a common language.
+==== How does DEVKIT use DTR and RTS to enter download mode? ====
+<TODO>
+http://sakacamprung.blogspot.co.id/2016/05/jelly-gamat-qnc.html

User Tools

Site Tools

Differences

Page Tools