Yes, I left the masking on one kitty as an accent. I thought it made the design pop a bit.
I’m pretty happy with it, but made a few boo-boos. Luckily, I was able to use superpowers to fix one of them.
Boo-boo #1: Forgot to take into account material thickness when I resized the SVG in the GFUI. When the print was complete, tab A did not fit into slot B. This is where the superpower came in handy. I drew a simple rectangle in Affinity Design and exported it as a SVG. I placed the faceplate on the GF bed and dropped the box over each slot and sized them to slightly enlarge the slot so the tabs would fit. A quick 12 seconds later, I had wider slots.
Boo-boo #2: Forgot that optical placement has precision limits. Keen-eyed observers will note that the base doesn’t align exactly with the face. When fixing the slot sizes, I was shaving a couple of hundredths of an inch off of each side. That worked fine, but I was working at the extreme left edge of the placement area and ended up with about 1/8" displacement side-to-side.
Boo-boo #3: Under-estimated weeding time. The print itself took 25 minutes. Weeding took 30. 'Nuff said.
Thanks for sharing the process, I especially found your “fix” to miss sized tabs encouraging. hopefully the alignment problems on the edge of the bed will get tuned in soon.
Because some of us use OctoPrint that runs on an RPI locally and gives the same benefits with no disadvantages. We upload our files to 3D printers from a web browser anywhere while watching them on a live stream. We don’t need an internet connection if we are on the same network as the machine. Software updates announce themselves in the GUI and we chose when to install them. It is then just one click and a service restart. We can roll back to an older version if we want to as well.
An RPI isn’t as powerful as the cloud but it can do machine vision with OpenCV and it can slice 3D files so it seems powerful enough to do what GF promises to do.
I imagine splitting the work between realtime stuff running on the machine and the rest of the processing in the cloud is quite difficult, especially things like machine vision and real time focus. Much easier when you have a high bandwidth low latency connection to a local host.
I think the main reason for using the cloud is to keep the secret sauce secret.
superpowers to fix one of them.
my 
I knew that I needed to pick some up (and masking and magnets and X, Y, & Z), but this all got amazingly real this week.
