Thanks for that! That goes a long way, I think.
Yes, and they made it very clear that the code that does the fancy stuff in cloud is very very unlikely to be released to the public. It is they key IPā¦those algorithm will be worth a good chunk. And seeing that currently there is no simple way to just home the head, there will be a lot of work involved to make the firmware with user/non cloud control usable.
Hopefully they have some ideas up their sleeve on at least a basic non cloud conversion, but until production units are out, focus is to get the current service working solidly. Considering the significant investor commitment, I would think theyāll be around for a while, but I like to keep my tools running for a long time, not toss and replace after a year or two.
Judging from the success GF has had, I have no concern they will dry up and blow away. Much more likely the company would be aquired.
My concern there would be the new iteration may start charging for a subscription. A captive audience presents an opportunity for a revenue stream.
A small fee, like a magazine subscription likely wouldnāt cause many customers to abandon their investment. Multiplied out across a growing customer base of tens of thousands - it would be attractive for itās potential.
I donāt think Glowforge will go away, but the business model might evolve, and with it - likely my motivation to hack the machine.
It will never be a brick, but it might become a shadow of itās former self.
yeah this is sort of how i see the worst-case scenario playing out
If they werenāt so broad with their marketing I agree. But since itās not solely geared towards businesses I donāt see that model happening. SaaS models like that arenāt common to the general public. Certainly possible though so I hope not! I think the recurring revenue theyāre looking at is Proofgrade⢠materials, Catalog purchases (Iām guessing they get a cut of those), and future accessories.
Dunno, I see subscription models replacing stand alone software sales.
I donāt think Glowforge under itās current management would do that, but any entity that aquired the company might feel differently.
I have no idea what the expense is to maintain the servers, in fact I have no idea regarding any of it - just speculation.
Frankly, I expected to see more software models being cloud-based and subscription by now. So Iām surprised. But I suspect itās the reason I said⦠folks just arenāt used to that and will rebel against it. Guess itāll be a slow-moving process, if anything.
My understanding is that there is no g-code-like interpreter hardware on the glowforge. Most machines take input from a computer and have hardware that converts it into code for the stepper motors. I donāt think the glowforge has this hardware. I thought @dan said something like the Glowforge gets the stepper motor control signals from the cloud service. This is one of the main reasons why you always need wifi, and reverse engineering this sounds to me like a huge undertaking.
I would feel much more comfortable and confident that I could use this machine years into the future, if the Glowforge team would commit to publishing a watered down version of their cloud software. Similarly to how the people at Blynk (IoT software company) publish code to create your own Blynk server.
Thatā's right⦠I recall that being said some time ago.
I would expect that the solution to GF going away would be a swap on the controller to something like the Smoothie board. If there are steppers in the GF, a different controller could do that so I wouldnāt worry about that issue. The unknown things would be the variable power supply and the cameras. I think thereās some work in the Smoothie camp to control power supplies so Iād expect that would be able to be handled.
It would require someone to either replace a board or add one for the GF. Folks have created drop-in replacements for the chineses lasers (like the guys at Cohesion) so itās not as simple as pointing to a new cloud address but it would allow GF owners to have something other than a paperweight and not require a lot of technical expertise.
I agree that producing code that for standard laser cutting would be faster to generate than the extra cutting and engraving features promised for our GFs, especially the Pro models. Hopefully it never comes to that and in a few months time this be a forgotten concern.
Most users donāt want to be forced into monthly fees just to access data files that they already have produced. In GFs case, at some point down the road you may see a fee imposed for your continued use of the GF cloud services. GF has committed to making that a free service for current users, but that commitment can change for a variety of reasons, say a change of ownership.
Itās also not clear to me if all of the GF cloud features, currently promised and those developed new in future, are included with the purchase of a GF Basic or Pro. It would seem fair to have to pay for new cloud features in some manner and it would be interesting to see how GF plans to deal with that eventuality.
Possibly. Thing is, they canāt charge to use the device you spent thousands on. Imagine if you bought a car, you own the thing, and it wouldnāt start unless you paid your bill. (Iām not talking about supplying it fuel.)
Thatās not to say anything about new features. But, still⦠Thatād be one lame company that did that to its customers.
Agreed and itās happened to me before. Iād just as soon not make the same mistake twice. As I may have suggested elsewhere, maybe @dan could comment on cloud pricing policy, now and in future, in the next update.
We donāt charge for any features right now, and we donāt have any future pricing announcements to share beyond what Iāve said before about existing features not being charged for.
ādan
and participles not being dangling from
???
(Yes, Discourse, that is a comment.)
Absolute worst case scenario:
Glowforge goes under, cloud services lost, no firmware of any sort exists anywhere for us to use. Heck, all of the programming team even disappears from the face of the world so we cannot even hope one happens to have a random old backup to sneakily distribute.
You still have motors, belts, mirrors, a power supply and a tube. All of these are very well known entities, and they have inputs and outputs that behave in known manners.
Laserweb or any other programming inclined group would easily slip in a smoothieboard or a beaglebone and have the whole system up and running (sans camera) within a month of dedicated work.
Camera is down, so homing is lost. That would mean grabbing a drill and buying a few limit switches.
So, a grand total of $40 gets you back up to a functioning laser cutter. You can now ditch the case and buy new rails/belts to increase your bed size as well (ideally design and produce a new case before you ditch the old one. Acrylic is a nasty stink)
Should have just included the link to the original topic: