Micro Python development wish list

C programming, build, interpreter/VM.
Target audience: MicroPython Developers.
chrismas9
Posts: 152
Joined: Wed Jun 25, 2014 10:07 am

Re: Micro Python development wish list

Post by chrismas9 » Wed Dec 16, 2015 11:40 pm

MTP does not allow file editing, only transfer or open read only. Verified with Android 4 and Windows Phone 8 connected to WIN7 and WIN8.

Differences between MSC (memory stick) and MTP (phone) under Windows.

1. Phone does not mount with a drive letter.
2. Windows explorer does not have "New --> Text Document" in context menu, only "New Folder".
3. Phone not listed in Save As dialog box in text editor.
4. A Photo opened in Windows Photo Viewer has "Open file location", "Rotate", "Copy" and "Properties" greyed out. Only "Delete" is available.

If MTP was supported with an MSC/MTP switch in boot.py I would use MSC for development and switch to MTP for field deployment. That way users could copy log files, etc without risk of corrupting their system.

User avatar
pythoncoder
Posts: 5956
Joined: Fri Jul 18, 2014 8:01 am
Location: UK
Contact:

Re: Micro Python development wish list

Post by pythoncoder » Thu Dec 17, 2015 7:38 am

While "blue sky" thinking is great and necessary perhaps I can be permitted to offer a different perspective.
The object of the exercise of developing updated Pyboards is to make money to support the development of the language. A new board has to compete against other offerings in a crowded market, and with ports of MicroPython to other hardware. Clearly as hardware improves we need to embrace change but adding features used by a minority is liable to increase costs. The circle can be squared if users can add features as they require them.

Consider the competition. The most successful competitors in terms of numbers sold are probably the Raspberry Pi and the Arduino. The Pyboard has a number of benefits over these, some evident to the novice, others to more experienced developers. But the competition has us beat in some ways. Taking the Pyboards first, I'd identify a few of its advantages as follows, in rough order of accessibility to users of increasing experience.
  • The Python language (vs Arduino)
  • Instant boot (vs Pi)
  • Ultra low power consumption (vs almost anything). This is potentially a big deal :)
  • Onboard hardware (timers, RTC, ADC DAC Uarts I2C CAN SPI I2S)
  • Open source firmware and hardware, well documented chip (vs Pi)
  • Interrupts
  • Relatively deterministic timing (vs Pi)
  • Optimisations (code emitters, C modules, persistent byte code, inline asm)
However the competition has an overwhelming advantage in terms of support. This takes the form of readily accessible user contributed code and commercial add-on boards. A beginner can buy a Pi and a "hat", plug the two together, run the supplied code and get results. The more experienced amongst us may mock this, but it's how most of us started - building things from other peoples' recipes and gradually learning how they work and how to adapt them. In my case valve radios ;) I know I've banged on about connectors before, and I won't again, but this can't happen without them. A kid can't do much with a Pyboard that will excite them. Wouldn't it be good if it could be pugged into a simple robot which ran first time? Wouldn't the thought of hacking the code be appealing?

Readily accessible user contributed code is vital. If I want to write a device driver for the Pyboard my first port of call is the Arduino site to see if someone's already done the job in C. The point here is that code is easy to find. While we have user contributed code it's rather scattered. You can search this forum or look in the rather tatty Wiki. Would that the user wiki were as visually appealing as the dev wiki, with users being encouraged to look there for contributed code (and developers to put links there).

Then there's the vexed issue of promotion. How did you discover the Pyboard? In my case it was featured (for reasons which escape me) in a Linux magazine. Not the typical reading matter of a kid or their dad. Promotion isn't my forte but somehow the word needs to be put out if it's to sell in markets beyond the established makers, coders and geeks who trawl the web in search of new toys ;)

Or perhaps we can't bear the thought of this forum being overwhelmed with dumb questions from twelve year olds...

Just a few random thoughts...
Peter Hinch
Index to my micropython libraries.

User avatar
jgriessen
Posts: 191
Joined: Mon Sep 29, 2014 4:20 pm
Contact:

Re: Micro Python development wish list

Post by jgriessen » Sun Jan 24, 2016 10:48 pm

However the competition has an overwhelming advantage in terms of support. This takes the form of readily accessible user contributed code and commercial add-on boards. A beginner can buy a Pi and a "hat", plug the two together, run the supplied code and get results. The more experienced amongst us may mock this, but it's how most of us started - building things from other peoples' recipes and gradually learning how they work and how to adapt them. In my case valve radios ;) I know I've banged on about connectors before, and I won't again, but this can't happen without them. A kid can't do much with a Pyboard that will excite them. Wouldn't it be good if it could be pugged into a simple robot which ran first time?
I'm planning to use flat flex cables and connectors for field instrumentation. Those are not readily available to hobbyists, but as I launch into that I will sell them at easy prices so that problem goes away.
John Griessen blog.kitmatic.com

Post Reply