Just had a thought…
If the cloud-processed job was sent back via your device, instead of directly to the machine, that would make storing for future re-running a breeze.
Of course, that would mean completely changing how things work today, both for the machine and the app. Which means, it’s never going to happen…
I saw that message two/three days ago, I guess it got worse the next day. An F5 fixed it for me, although it threw an error on the third of four files I uploaded that day. It was a standard toast error message and I just reloaded the file.
Obviously running into this on any sort of consistent basis is not acceptable. Whether it is a couple day thing or an ongoing thing is what is important and tells you about the current state of glowforge the company.
I want to be clear as to my suggestion here. While I think it would be cool to locally cache jobs on a personal computer for offline use, that is not in the scope of my current recommendation.
What I am suggesting is that they introduce a caching system on their side to store commonly used jobs on a per customer basis. This would be one completely rendered motion control file that could immediately be sent back down to the glowforge, instead of sitting in the job queue and waiting for rendering.
90% of my glowforge jobs are the same stuff I do constantly, and require incredibly high rendering time. I always use the same settings, the same positioning, and the same material. Now the important question is how many people do this like I do. If its not a lot, it wont help them much, but if it is something where someone normally just throws new material in, pushes go, and repeats, then it will reduce their compute time drastically.
If I were to run jobs back to back 24 hrs a day, they would save 4 HOURS of compute time, just from my jobs. Thats 4 hours less people would be waiting as a whole to get their work done, and mine would be done instantly.
Take that amount and multiply it by the amount of people reusing the same exact job, and it could potentially lead to large amounts of compute savings.
Now IF they decide to instantiate this process, it would definitely be possible to offload the ability to local machine, but that would be a much larger amount of work. It would be pretty cool though.
Sure makes sense to me as a worthwhile feature. And I’m ok having it all server side. It is how computing is moving and I don’t see it stopping.
JUST got it here… Friday 4:50pm EDT
We’re sorry, too many people are accessing this website at the same time. We’re working on this problem. Please try again later.
It’s that Christmas crunch. Just gonna get worse IMO
This topic was automatically closed 32 days after the last reply. New replies are no longer allowed.
