metaseq/qseq: ms set setting not included in save

[nofoto]

A metaseq set assigned to a sequence in qseq is lost when the netpd set is saved and reopened.

[reduzent]

I can't reprodue. The metaseq set selection in qseq3 set is restored properly. I wonder if this might be some edge case or if order plays a role. However, I tried loading qseq3 after metaseq and vice versa and it still produced a preset file that correctly loaded all settings.

Maybe you can provide a detailed step-by-step guide, starting from an empty unpatch.

[reduzent]

This is how it looks after loading the preset:

[reduzent]

This is the preset file I used for testing: test-meta-qseq-not-saving-bug.zip

[nofoto]

well yeah it's strange - if you go back to the file i put on discord with the bug it just refuses to save the setting in qseq; you can change it, save the netpd set and when you open it, it's gone back to master. so the problem is in that setup somehow - are you able to read the way it saves the settings to see what happens? i can't reproduce it starting from scratch either.

[reduzent]

Thanks for the hint and the example preset. I attached it here as reference. There is definitely something fishy going on. I used chat's debug feature (/debug to on) to print every message when loading the preset. And it contains this:

NETPD_SND: [..b] /3/qseq3/2/metaseq/length 128
NETPD_SND: [..b] /3/qseq3/2/metaseq/offset 0
NETPD_SND: [..b] /3/qseq3/2/metaseq/set -1
NETPD_SND: [..b] /3/qseq3/1/metaseq/length 16
NETPD_SND: [..b] /3/qseq3/1/metaseq/offset 0
NETPD_SND: [..b] /3/qseq3/1/metaseq/set 0

So, set 1 should be set to metaseq set 0 (-1 means master).

b2grdz261023b.zip

[reduzent]

@nofoto Thanks for reporting this. This was a hard one to find.