Where's the next version?

classic Classic list List threaded Threaded
1 message Options
Dan AE9K Dan AE9K
Reply | Threaded
Open this post in threaded view
|

Where's the next version?

The release cycle for several months leading up to July was pretty fast.  This was partly due to a quick evolution of features in firmware by Gordon at RF Concepts - some of which were intended to support planned features in the monitor.  The goal was to get stable monitor and matching firmware out in July.

I was on track to do just that until the last two features which were contemplated back in April seemed too good to push off to a future release.  The result was typical scope creep.  Adding to that my day job got a lot more demanding at the beginning of July and mid-July another analyst left leaving us less than half staffed.  Instead of coming home and developing at night, I was either still working or not able to stay awake.

The two features Gordon easily added to firmware yet I'm slow to accommodate in the monitor are an EEPROM configuration dump and streaming telemetry.  

The EEPROM dump capability is the start of a set of related features that will ultimately allow backup and restore of amp configuration as well as the potential for factory-defined or factory-modified configurations that can be sent to the user to upload into the amp.  In the first iteration, there will be a dump (backup) capability, then a parser to display what the dumped data block means.  Sometime after that will be a means to upload.  This last feature needs lots of safeguards so it will take a while to develop.

The streaming telemetry feature is meant to lighten the telemetry processing overhead on both the amp and the monitor.  Currently all telemetry is polled (the monitor asks for data, the amp sends it).  Streaming telemetry allows the monitor to ask the amp just once to keep sending specific telemetry ten times per second.  When doing so the amp no longer needs to receive and parse the monitor's requests and the monitor no longer needs to spend time building and sending the requests.

These two features aren't easily added to the monitor without breaking existing functionality.  Some of my design decisions are invalidated by the way these features need to be implemented.  Thus it's somewhat of a re-work rather than an addition.

Hopefully in August I'll be able to devote more time to these features as well as a more robust threading model to overcome the elusive hang issue with serial port closure.

Thank you for your patience.

Dan
AE9K