I’m a software developer interested in the programming interfaces available on Glowforge. I’m a long-time forum lurker, but first time poster, and proud soon-to-be Glowforge Basic owner. Can’t wait for day 16 to ship!
I’m excited to see how the web-based interface and server infrastructure allows GF to optimize image processing and tool path creation! The architecture is certainly unique, and is a good choice for the hardware concerned in GF.
I’m hoping to do some more… technical work with my GF. I write system code for quadcopter drones with image processing the camera feed being a particular goal, recognizing people, terrain, good and bad landing zones, etc… I’d like to use similar techniques to do material recognition and categorization in the GF cabinet.
What APIs are used internally to the GF firmware for communicating camera images to the web service? On the other side, how are the tool paths communicated to the cutter hardware? I’d love to get into the nuts and bolts of the software in GF!
@dan we met briefly at CES 2015 and talked about the topic a bit. Mind if I pick the brain of one of your software guys?
Regarding questions of what’s going on behind the scenes, Dan has replied " It took our physicists a long time to make that work, and I don’t want to deprive scientists working elsewhere the joy of discovery."
This is another reason i’m looking forward to everyone having received a forge. There are definitely other areas beyond production that need time and money dedicated to them, and once the pressure to ship is lessened I hope to see an epic butt load of updates to the software.
Beyond that I don’t recall an API ever being mentioned here, so I wouldn’t count on seeing anything for probably a year+ given all the other ideas in there that would affect more people.
There are no APIs, but it’s all written in react+redux, and fairly easy to interact with. I’ve made a couple of small adjustments to the software, so it’s definitely doable =)
You’d be better off putting together your own laser cutter and then using it as a platform to experiment on. If you get something working there will be people lining up to buy it. (protip for swindling investors: act like it’s already finished before you’ve really even started)
Well, except for the ones that have an ounce of sensibility to cut bait at the beginning. But the ones that have no will power to do it, they’ll stick around. You can keep them on the hook for months.
Not at all. The more high-quality 3rd party software that exists to work with a hardware platform, the more popular it will be. Offering an easy-to-use API enhances hardware sales and is likely in GF’s best interest. I just want to make my GF do even more amazing things!
If GF makes the API and some basic documentation available, more developers will be likely to contribute code that then ends out in other user’s hands. Otherwise DIY guys like me build cool stuff, but it doesn’t get contributed because there’s no good place to do so.
Custom software always appears for interesting hardware. It’s just that with company support, the software ends out higher quality and with greater scope. Software is hard enough without having to twiddle with GPIOs and analog interfaces with no docs!
I’d also be very interested in this, even if it was limited to things like upload, delete jobs, trace image, print image, etc. If it got more advanced into planning engraves/cuts that would be amazing for handling more repetitive jobs.
As one who drove an 18-wheeler through Autocad’s scripting language getting it to do all I wanted, it would not be difficult to hook a simple open-source scripting language like LISP to the API that the hooks can be very proprietary, but the scripting very open. That allows a way that competitors cannot just walk away with your work but many other folk can also contribute.
Not having received my GF Basic yet, I’m not sure what to build first with such an API. I’d like to see how the standard interface works before committing to anything. I hear that the unit can do relief engraves pretty well, but what about working on materials that contain variation before cutting? I mainly am interested in the possibility of using a structured light infrared dot projector and camera combo (Kinect-like) with a very tight throw for registration of varied surfaces.
In this paper first experiences with Kinect v2 sensor are related, and the ability of close range 3D modelling is investigated.
Moreover, at the nearest range of 0,8 m acquired in this test, the standard deviation is higher than for all other sensor positions. As a matter of fact, measurements realized at the minimal announced distance of 0,5 m would probably be still less reliable.
