MTF Files

ReadMe for version 2.2 Release of MTF. 26-03-2025

Here you should find:
mtf.py          : microPython module
mtf-doc.pdf     : documentation for the release
test-suite.py   : set of tasks which test and demonstrate MTF features

To Install:
Download mtf.py and install it to your Pico using Thonny

If you want to try out the test suite, download it and open it using Thonny. 
Then you'll be able to run it.  The test suite consists of 7 tests which 
exercise most of the service calls. It ends by displaying metrics and 
other status information.

For your own project make sure you include an "import mtf as MTF" at the 
start of your file.  See Section 4 of the PDF documentaion for more help.

Changelog

2.2
Changes:
- Task registration now only needs the python function name. When referring to tasks
  in service calls you may use either the python function name or a string equivalent
- MTF now supports use of the watchdog timer
- Pipes have been enhanced to include read locking
- the tick-rate must be between 1ms and 100ms (previously 1000ms was permited)

2.1
Major new version
Changes:
- attempts to mask the effect of the garbage collector by creating
  "fat-ticks" when gc runs. See documentation. Improves short timimgs
  by an order of magnitude.
- metrics now permanently enabled

1.6
Major update
Additions:
- task alarms
- simple logging system
- traps run-time exceptions and logs to the filesystem, with stack-trace
- support for setting the time from wlan on PicoW
Changes:
- Metrics functions renamed in-line with Python style-guide
- Metrics FIFO now holds 7 values and the order has changed
- Automatic run loop timing adjustment is now enabled by default
Removed:
- support for regular Raspberry Pis

1.5
Minor revision which now detects if it is running on a Pico W and 
MTF.status( "SYSTEM" ) reports platform as "PicoW"

1.4
MTF.status() now returns a list of the current pipes when the system 
status is requested

Some minor changes to the documentation.

1.3
Bug fixed in status() service call which caused MTF to crash when a signal was
detected waiting for delivery.

Pipes have been added to MTF. New service call MTF.mkpipe() creates a new pipe

Status() now shows waiting signals which are blocked due the signal mask

All service calls now return 0 on success or -1 on error and set the 
global MTF.errno with an error code.  
The new service call MTF.error() can be used to print error messages.

Added service call MTF.pause() which can be used to pause a running 
(scheduled) task

1.2
The run loop timing error can be automatically adjusted by setting 
RLTE_auto True.  This is done through the rlt() service call.

1.1
First public release

AFL, 10/02/2024