xilvar

I found another interesting bug today. For some time I've been confused as to why cura 2.3b is producing a Z20.2 unit move right after the start gcode and before the skirt starts. The 0.2 part of it was particularly suspicious since that matches various settings I have and 20.2 is a weird number to arrive at. I have retraction enabled and z-hop when retracted enabled with a 0.2mm z-hop. When that is on I end up with this gcode (among others) between start and skirt : ;LAYER:0M106 S127.500000G1 F1500.000000 E-3.00000G1 Z20.200G0 F900.000 X28.472 Y28.838 Z0.500;TYPE:SKIRT Notice the G1 Z20.200 part. When I disable z-hop that line is omitted: ;LAYER:0M106 S127.500000G1 F1500.000000 E-3.00000G0 F900.00000 X28.472 Y28.838 Z0.300;TYPE:SKIRT I'm really not sure where it's getting the 20 from though...

I've noticed a problem which I can't recall whether 15 or 2.1 had. I've been printing a lot of petg recently and adhesion has been 'difficult' to say the least. Thus I've been using brims a lot more frequently. When I print with a brim, it makes all the right lines, but it doesn't seem to be doing them in an order in which the brim would be of the most help. 1. It makes the brim lines. 2. It skips inwards to the outermost inner wall and puts that line down 3. It finishes up the other walls. Thus the first wall (which is the one I really need to adhere so it doesn't detach and destroy the rest of the piece) does not even benefit from the brim!. I sort of understand this behavior in that it helps at higher layers with overhangs but it doesn't SEEM to make that much sense to me on the initial layer. Is there some way of turning that off on the initial layer?

>> Layer view speed > Could you post the model somewhere so we can see if we can reproduce this? Sure, could I possibly send you a link or the file privately somehow? It's a model that I consider to be private since I'm considering eventually having it molded and produced. Here's a screenshot of it with main thread locked on 'processing layers'. Mouse cursor was beach ball the whole time. Progress advances in giant steps, but tends to get stuck at 100% for a long time. >> Settings saving >This was fixed last thursday. The fix will be part of the final Cura 2.3 Awesome! >> Concentric infill > Concentric infill is a feature that is just not going to work for all types of models. It is up to the user to decide on the best infill pattern for the model. Ok >> Double walls >Do you have an example/screenshot of this? Sure, here is a screenshot. In this particular case if you do up the infill to 100% it fills in the void. However there seems to be no way to make it do infill at all below 100% other than reduce wall width. The problem is that for some models with thin structures in some places and thick structures elsewhere you might not actually want to run 100% infill everywhere. Not super sure what the solution would be... option for: 'dashed line infill for infill < 2 nozzle widths' ? [media=21091] [/media]

I've been 2.3b since it was available and it generally works great compared to 2.1, however a few problems I've noticed. (some of which are in 2.1 and 15 as well). Layer view speed - Much faster most of the time, however with fairly complex models it seems to be both very slow at the 'processing layers' step AND that particular step seems to run on/block the main thread/run loop. For a model I've been working on for a few weeks which originates in openscad it can take upwards of 30s for that layer view step while nothing else I use takes that long to visualize either the stl or the finished gcode. The gcode ends up around 9MB for that and works fine. Settings saving - Seems to work far better than before but confusingly 'loses' some settings when you save to a _new_ profile. Perhaps because it doesn't save material settings with the profile settings those settings get reset? Concentric infill - It's really easy to generate a degenerate concentric infill where there's an inverted unsupported cone in the middle of the infill. I would love to be able to use concentric more frequently for transparent/translucent filaments... But I often cannot because of this problem. Double walls - There still seems to be a problem where sometimes you have to fudge your wall sizes a lot to get cura to actually fill in your walls when not on the top or bottom skins. It seems like in most cases like this, it should be either generating the outer walls and then a single inner wall between them, OR putting a little bit of (100%) infill there. Instead neither thing happens and it just ends up hollow.