Maintainers? Calibration.py cleanup?

[folfy]

I see @KartoffelToby being overloaded a bit with the project and number of open items - I'd love to help ya more dealing issues and providing some more PRs, but would any of you @RubenKelevra, @wtom be priviledged to also discuss, review and pull those changes then? Would need at least some guideline / feedback and support from an active maintainer 🙃

For example, I see an attempt to improve Calibration.py with 1.7.0b1 a bit, but honestly I'd love to properly clean up some things there and discuss a few major improvements, like deduplicating the core calibration routine between local and target temp calibration, and making it independent of the mode (like 90% of it should be/is doing the same). But for any such "more critical" or major changes, I'd need to talk about some things (like what the idea/intention behind some of the modes is), and some kind of OK, before I'm wasting my time on something that you'd maybe not wanna merge anyway.

[RubenKelevra]

@folfy I've worked with @KartoffelToby in the past on this project, just was missing time to do that for a while, but got now more time and I'll be happy to review code. Just ping me.

[folfy]

Actually relevant:
Can you clean up issues? For example #1440 is a duplicate, or #1398 is lacking any information, even though requested, so should be at least marked as stale or soemthing - Would you want me to just make a list of all the issues to update, or how can I help best? I guess only @KartoffelToby can give any kind of permissions?

[KartoffelToby]

@RubenKelevra was disappeared for a long time, and i removed him. Because I take security risks and integrity very seriously

But i havent any Problem to readd him to the
contributors again.

I'll add you (@folfy) too, if i checked your PRs and when you plan to do more PRs in the future.

Anyway thank you both for your work and time!

I will have more time in the next few days, I have been suffering from a long period of illness but I am on the road to recovery

I try to check and merge all PRs tomorrow or this Weekend

[folfy]

Hey @KartoffelToby, rly glad to hear from ya!

Yeah, it would be great if there's at least a second person around that can review and approve PRs again 👍

Regarding permissions for me, I e.g. don't mind having to poke someone for merging stuff, I'd anyway ask someone to at least briefly review the changes. Mainly there are a lot of issue, and I'd love to at least throw some of the labels on for stuff needing more info/logs, closing duplicates, and assigning/taking over some of them. So would be great to get some permission for that maybe😉

I know life always can get in the way, no worries, that's why I tried to poke @RubenKelevra for whatever he could do as well. Hope you're rly doing better now, don't wanna burden you too much.

What are ya thinking about some changes to calibration.py and utils like:

• fixing rounding from rounding-down to nearest step - Does not round correctly #1475
• Reducing rounding functions to simple math.floor, math.ceil or round() calls (where possible)
• Smaller improvements, like replacing "(x < y + a) or (x > y - a)" with "abs(x - y) < a"
• Most important: The previously mentioned merge of local and target temp calibration into one common subroutine (the previous changes would also make life easier for doing this step afterwards)

Just short question on that still - Any objections regarding such improvements? Idk how attached you're to the current state of the algorithms and e.g. calibration.py, but I'd love to do some proper maintainance 😅

Thanks!

[folfy]

One more thing:
I've been working on main, but there are quite a lot of changes in v1.7.0b1 -> Before I continue further, should I right away try to rebase on/merge v1.7.0b1? It is quite a substential change already, so merging is already bit of a pain as of now :/

[folfy]

FYI, there's rly a lot to fix in the calibration, for example what I just fixed in fe622d8 now:
We add the current (i.e. previously set) calibration to the newly determined temperature difference on every iteration of the calibration loop, meaning this will just quickly shoot towards min or max local calibration (depending on if delta_t is greater or smaller 0). So instead of applying just a calibration offset, we're integrating it on every update, which I highly doubt making sense. This applies at least to local calibration mode with aggressive or default calibration, because in that case the result of this calculation is not completly overwritten right after.

I threw that commit just on top of my rounding branch now, which I know is not nice, but if I keep making a new branch from master for all the fixes and improvements within calibration.py, things will just turn into an even worse merging madness. Unfortunately I don't have any TRV with local calibration for testing that stuff in the real world, to proof how bad it was and how this fix drastically should improve the control loop.

I'll soon be busy with christmas stuff, and since I don't wanna risk wasting more time with working from the wrong version potentially, or not getting stuff merged because it's too many changes or something.... I'll just have to let it be for now, till I got some more feedback on the open questions from you @KartoffelToby or @RubenKelevra 😓