[BUG] Memory use on 2GB tablets.

[dcoredump]

Describe the bug
I use pipedal_1.4.72_arm64 on a Raspi5(8GB). The whole thing seems to work fine when I use the web interface. When using the Android app, the app (or rather the app interface) seems to hang up at some point and you can't make any changes

E.g.:

• fresh installation
• connecting via App
• creating a new preset
• adding pedals (tuner, Input, NaM, ...)
• going to "manage presets" an rename the new preset

This seems not to work. So I tried to change preset... this also seems also not to work, but after some time the change is done and the "new and renamed preset" has its new name. So it seems to work but takes much time and there is no (visual) feedback. Not sure if this is a bug...

** Environment (please complete the following information):**

• OS: Raspbian (Debian GNU/Linux 12 (bookworm))
• Linux pipedal 6.6.74+rpt-rpi-2712 Create CODE_OF_CONDUCT.md #1 SMP PREEMPT Debian 1:6.6.74-1+rpt1 (2025-01-27) aarch64 GNU/Linux
• Pi5 (8GB)
• PiPedal Version 1.4.72
  If UI-related:
• pipedal Android app (make problems), Chromium on PC works

[rerdavies]

How much RAM does your phone have? If it's < 2GB, let me know, and I'll review to see what I can do to reduce memory usage in the app.

[rerdavies]

I've tried it on a variety of phones (the smallest phone/table I have has 2gb), and I haven't been able to replicate the problem. Perhaps we should look at system logs to see if something odd is going on.

1. Reboot your raspberry pi.
2. Reproduce the issue using the steps you gave above.
3. At a system prompt on the PiPedal device, run the following command:

journalctl -b0 | grep pipedal > capture.txt

And post the output here.

[dcoredump]

Thanks for your fast response. In fact I used my old tablet (which says it has 2GB RAM), but using my much newer phone works like a charm! So no further debugging is necessary!

Thanks for this nice software. Hope I can try it soon on the PA in the rehearsal room!

[rerdavies]

Make sure dialogs are conditionally generated. Most likely a memory overcommit.