Release UI on G26 priming timeout

[tpruvot]

+ remove a duplicated ifdef

Unsure if the default extrusion requiring user click (G26 P) is the right thing to do.
Here, that needed 20cm of filament extrusion (make a huge blob in the middle of the bed)

and nothing was explained on the LCD to interrupt it... edit: nvm, MSG_G26_MANUAL_PRIME is shown after a closer look.

[tpruvot]

hold on before the merge, i detected a printer freeze after this prime... unsure yet what is causing that

wait_for_release maybe, unsure

[tpruvot]

yep, the workflow is kinda problematic in the UBL assistant.. step 3 (to check PLA mesh) requires the heat of the bed+nozzle, then manual prime...
That take a while to heat and stabilize heat, and if the user doesnt click, the whole process is aborted... after 20cm (or more) of wasted filament

[tpruvot]

well lol, far from usable test print... but now works...

2 big issues there : no retraction and lines are shocking/moving a part of the circles. And the prime extrusion made just over the first point (bottom middle)... instead of a corner/border of the bed

[ManuelMcLure]

Do you have ARC_SUPPORT enabled? I made a change to G26 about a year ago that would use the ARC_SUPPORT feature to make the circles instead of drawing them as a series of short lines - it might be interesting to toggle ARC_SUPPORT and see if you get a better result from G26 with the opposite setting.

[tpruvot]

yes, ARC_SUPPORT is enabled in my config. Its not what i meant... There is a grid line passing in the circles.... after they are printed

[ManuelMcLure]

Yeah - my thought was that perhaps the non-ARC_SUPPORT code would end up with a circle that had more extrusion so it sticks better to the bed and thus the line doesn't pull it. My circles don't get pulled up by the lines (even with ARC_SUPPORT enabled) - maybe I'm running my nozzle closer to the bed or my bed has better adhesion?

[tpruvot]

its not "pulled", the lines should stop at the circle, i mean... to passing through them on the same layer
they are pushed :p right well, some were pulled too, now i see

[ManuelMcLure]

I believe the lines are supposed to do that so that it's easier to determine the center of the circle. But @Roxy-3D would be the person to ask about the rationale for the pattern.

[tpruvot]

anyway, was not the purpose of the PR... now it should be a bit less disappointing :p

[tpruvot]

maybe a G26 P10 (instead of G26 P) in this ubl assistant could help a lot of users... step 1 is "cold bed probe"

[ManuelMcLure]

At one point I was thinking of adding an option to G26 that would only print the four points closest to each corner (i.e. 16 points) to help in manual bed levelling before the complete probing occured. I haven't gotten around to it, though.

[thinkyhead]

if the user doesnt click, the whole process is aborted... after 20cm (or more) of wasted filament

Is this addressed now? If not, should we fix that here too?

[Roxy-3D]

yes, ARC_SUPPORT is enabled in my config. Its not what i meant... There is a grid line passing in the circles.... after they are printed

Yes... Originally that wasn't the case. But the grid lines intentionally go a little bit inside the circle just for aesthetics. It kind of looks like cross hairs on a rifle's scope.

[Roxy-3D]

I believe the lines are supposed to do that so that it's easier to determine the center of the circle. But @Roxy-3D would be the person to ask about the rationale for the pattern.

Yes... But since that time, another reason has evolved. In the case of IDEX... It probably would be good to print the G26 pattern partially with one nozzle, and the rest with the other nozzle.

So, maybe it would make sense to print the circle with one nozzle, and all the straight lines with another nozzle. You would see very quickly where you had bad offset numbers. (And if doing this way... It probably makes sense to have the lines go a little past the circle and continue inward some.
That would make it easier to see how much the two nozzles are out of square.)

[thinkyhead]

Multi-nozzle alignment could be helped by G26, but I bet we could develop a better procedure to get X, Y, and Z dialed in for multi-nozzles. The E3D Toolchanger group developed a bunch of good procedures during the beta testing period, so maybe we can borrow some of those concepts.