For a guy with such a subdued demeanor you have a great sense of humor.
You are always standing by to tether us to reality. 
3 Likes
The lack of any official notice or response here from support during an hours-long situation when multiple customers are reporting being unable to use their machines is a bit disheartening. I am deliberately not looking at any internal dashboards so I can honestly say I donāt know if thereās a GCP issue that I wouldnāt be allowed to mention anyway. Publicly, https://status.cloud.google.com hasnāt listed any problems so this would seem to be application-level issues under Glowforgeās control.
At the same time, weāve all accepted that the system is still in beta, so I say ādishearteningā rather than āunacceptableā. Iād like to think that having a good incident management plan including customer communication is in the part of the hopper that will get done soon. Spes consilium non est!
6 Likes
Original post had little to do with our discussion. I find that Support often does not even look at additional posts when trying to diagnose the original issue. One problem per topic.
2 Likes
Yes. By design.
I agree that it is way better to flip the whole sheet. Iāve done that on some cardboard tests, especially if I cut a square around the whole set of tokens and then flip. I make a large jig of cardboard that lies on the crumb tray and cut out a square from the usable area, then cut a rectangle of the material to fit it. Works best with full sheets, but smaller ones that are symetric work. Very important to get the layout of the tokens symmetric. Thatās where I go astray most.
I canāt say that I have clear evidence that there has been a change in the software. I was able to do a full sheet of tokens that I had done before. But I hadnāt run into issues before with multiple objects that are all vectors. Curious as to what the limitations are.
3 Likes
Machine is back on line. Both sides of the current run are good to go.
6 Likes
We know that one limitations is the job has to complete in three hours. With 77 tokens that would be only two and a half minutes per token.
Another token run successful. All good now, support feel free to close the ticketā¦
It would be nice to have some understanding of what the issue was.
2 Likes
I did a print of tokens that was 3 hours and ten minutes and that worked. Iāve had longer. My basement is cool and I have never had a heat pause or warning.
Here is a test upload:
It makes a difference if you combine the objects of the same color or not for engraving so itās not simply a per object equation. I usually set the tokens up to engrave the full width of the bed.
4 Likes
OK. Dan said it was 3 hours, it must be a bit more, but there will definitely be a time limit.
Three worth for on board memory. I wonder though if that is the same thing as capping a print at three hours.
2 Likes
Every time this thread pops back up I think itās talking about something else.
9 Likes
I think currently it has to fit in one buffer but the plan is to be able to have multi-buffer jobs.
2 Likes
I havenāt seen a bottle of that stuff since I was a kid in N.C.
1 Like