mindforger icon indicating copy to clipboard operation
mindforger copied to clipboard
verified publisher icon dvorka

[Question] ARMv7 Debian package?

Open lordofscripts opened this issue 1 year ago • 3 comments

I am using a Raspberry π with Debian Bullseye. It's an armv7l platform. Any plans to include a release for this architecture?I was using Kanri Kanban but that Tauri contraption doesn't seem to be stable.

lordofscripts avatar Dec 27 '24 17:12 lordofscripts

So I'm curious and have a few questions if I may:

  1. ARMv7 is a 32-bit architecture so which edition of the 32-bit Raspberry Pi OS are you running then?
  2. What model of Raspberry Pi do you want to run MindForger on and
  3. How much RAM is it equipped with?

johnblommers avatar Dec 27 '24 22:12 johnblommers

So I'm curious and have a few questions if I may:

Sure!

  1. ARMv7 is a 32-bit architecture so which edition of the 32-bit Raspberry Pi OS are you running then?

Indeed it is a 32-bit architecture. I have been using it as a graphical desktop for over a year.

It is a Raspberry π 2B which has been running Raspbian Debian Bullseye since I reinstalled from scratch a year ago.

  1. What model of Raspberry Pi do you want to run MindForger on and

It is a Raspberry π 2B.

  1. How much RAM is it equipped with?

That I know the π 2 has 1 GB.Kanri ran well for some time until suddenly it stopped working. I think it is the Tauri framework that the app uses (he is still porting to Tauri).

lordofscripts avatar Dec 31 '24 16:12 lordofscripts

MindForger is too useful to not build on a Raspberry Pi. So I gave it a shot on a Raspberry Pi 4B with 8 GB RAM running the latest Raspberry Pi OS. So yes this is not a 32-bit system running 1 GB of RAM.

But I followed the build instructions for Debian systems. I had to install libcurl4-gnutls-dev to satisfy a dependency for curl.h, but after a few long minutes I had a working build of MindForger. And that was pretty darn satisfying!

So let me suggest that you follow the MindForger Debian installation instructions on your 32-bit system and cross the fingers that all the 32-bit dependencies are present in the Raspberry Pi operating system repository.

Given the 1 GB RAM limitation of your Pi, it may be necessary to limit the make step to a single thread, so the build process does not run out of RAM+swap and crash. Worse case, you'll have to increase the amount of swap and possess the patience of Job while the build crawls along.

I hope that you have the bandwidth to attempt this. It will be very interesting to hear about your progress.

johnblommers avatar Jan 03 '25 02:01 johnblommers