I trust Glowforge, but how do we avoid a Revolv scenario?


#1

Hi @dan and team. I can’t wait to receive my Glowforge and am also a big fan of cloud computing (as I work for a major provider).

Does the Revolv shutdown increase the likelihood that Glowforge will eventually produce virtual machine or container images that would enable users to run their a private “Glowforge cloud”? The topic has been raised in the Glowforge Community (Open API?) and on Reddit (https://www.reddit.com/r/gadgets/comments/4dfzwv/google_nest_is_permanently_disabling_the_revolv/)

As an example, the team over at Particle.io has published an open-source, API-compatible version of their server (https://github.com/spark/spark-server) and instructions for setting up a self-managed cloud on EC2 or GCE (https://community.particle.io/t/tutorial-local-cloud-1st-time-instructions-01-oct-15/5589).

I can’t wait to get started!

-Jamie


What if GF disappears?
#3

Yes, it could happen. Even though the cost of the cloud is built into the price, the Glowforge could become orphaned. This is a possible scenario. I recognize that I most likely will have a limited time with my Glowforge before the button turns red. I wonder which will be the juicier tidbit to buy Glowforge the company: the software or the hardware? From discussions the IP that the software represents is just going to be too valuable to open source.


#4

Dan has said they will open source the device firmware, which almost certainly won’t include any of the fancier server side software but will at least mean the device isn’t going to be totally bricked if they stop supporting it


#5

Sam,

Thanks for the link. I am super-excited to see that we will be getting intelligent laser cutters that can take advantage of the effectively limitless compute capabilities provided by the cloud.

That said, Revolv users also enjoyed the many benefits of cloud hosted services. Unfortunately, they are now running into sub-optimal business model that has caused Nest/Alphabet/Google to shut down their service, effectively bricking the device.

As more of the smarts behind the Glowforge move to the cloud, how can we ensure that we aren’t left with a large, $2,000+ paper weight. Dan and team have a plan. When they can share, I would like to see that plan. Open source firmware only goes so far when you have a device that cannot be driven by any existing software (assuming Glowforge cloud services went away for some inexplicable reason).

-Jamie


#6

If we lose access to the cloud servers (or GF decide to start charging for access or something) then we can still use the open source firmware to drive the device. It won’t be bricked.

You will have to wait for somebody to release a reliable gcode to GF-firmware tool, or write it yourself, but we shouldn’t end up fundamentally locked out of the device.

Probably more of a practical problem will be access to replacement tubes and other custom GF hardware if they decide to focus their efforts in a new direction or shut up shop. The current device will be superceded pretty quickly, and it remains to be seen how many years we can get spares for the gen 1 GF or whether it can be modded to accept off the shelf parts.


#7

This is exactly correct.


#8

My understanding is that the GPL-licensed firmware will take some amount of work by the forum community to get working. I think a firmer commitment to provide a dumbed down version of software that can be used out-of-the-box would clear up my concerns regarding a “Revolv Scenario”. @CharlesDarwin seems to have an understanding of how individuals can create a “self-managed cloud”. I second his request for Glowforge to follow Particle.io’s lead and create similar software and tutorial.


#9

I’m reasonably confident that the choice to run their app server-side was made for two reasons:

  1. it means Glowforge only has to write and maintain a single app for a single architecture (their servers) and vastly reduces their maintenance and support requirements

  2. it allows them to maintain tight control of their IP - they don’t have to publish the source for their fancier features in any form

While it would be great to get a ready-to-boot package containing their server-side app, it directly negates both benefits they were seeking when they chose the cloud in the first place.

I wouldn’t worry too much about the software - the low level firmware will let us do anything the official front end can do and more, and realistically Glowforge aren’t likely to shut down access to the gen 1 front end for a long time anyway

If they did drop support, I would be more concerned about access to spare parts


#10

I apologize if you got the wrong impression here: I’ve stated that we will not do this. I hope our commitments have been clear and firm.

Some of you may be familiar with the Electronic Freedom Foundation (EFF), a champion of freedom in hardware and software. They wrote this piece about Revolv/Nest, which is sobering and worth a read. In it, they say:

In an ideal world, Hub owners would be free to point their devices at a different central server, run by a third-party competitor or a trusted friend, or even run such a server on their own. They would likewise be free to collaborate on improved software that would unlock the potential of the Hub hardware or purchase such software from a competitor to Nest.

This is the approach we’re taking to ensure that, should something happen to us, you have options.


If GlowForge goes under
#11

The whole Revolv/Nest thing freaks me out, if google can just stop supporting a product so can you guys. Ive only read about angry customers and how google has to compensate them somehow.

Its concerning how closed the system is. I can’t write software or anything like that so if glow forge went down then I would be stuck with a brick


#12

Thanks for the link to EFF. It is hard to appreciate all the different rights and obligations of users and producers with so much happening so fast. The that Glowforge team under @dan’s leadership, and the community coalescing around what will be an amazing product are acknowledging these complicated realities and are in dialog about them has been very refreshing for me. We’ve had some enlightening threads discussing many aspects of what it takes/will take to keep the Glowforge working. I understand @dan to be committed to ensuring that options are in place to ensure the Glowforge won’t be a giant paperweight some day because it has been orphaned by the company. It will be up also up to the community to use the avenues that have been opened up. One aspect of the long wait until delivery is that we have had an amazing amount of time to get to know each other and should the day come, have a head start on keeping our forges running. That this is a new consumer electronic device that has a modest price point means that we as consumers have to do due diligence. The EFF article explains a bit more what that entails for us. The article also is very clear for the company about what is at stake in producing a project that is so intertwined with the cloud ecosystem. Perhaps forum members can share their own experiences, positive and negative about dealing with an orphaned product. That could be very helpful for us all to continue to broaden our understanding of what is involved in this issue.


#14

I have two lasers at work that perform wonderfully. Neither is connected to a server or web service to work. I fail to see any reason why it’s necessary to have GF connected to a server in order to function. I get that the processing is done on GF’s servers and then sent to our machines, but why can’t that simple processing be done on our own computers?
The whole Wi-Fi only aspect of Glowforge has been the one thing I’ve been very uncomfortable with the whole time. It’s one thing to utilize newer technology like the cloud to better your product. But when the product doesn’t function without it, it’s not necessarily a good thing. My iPhone utilizes the cloud for organization and file backup. But if that cloud server goes offline, my phone doesn’t suddenly become a useless brick…


#15

It is not necessary, but it is a lot cheaper, quicker to market and safer.

They can either write the software for a computer you connect your glowforge to, they can put a computer in the glowforge, or they can be cloud based. If they write it for a connected computer then they have to write software for Windows, at least one Apple OS, at least one flavor of Linux, and possibly a Windows Server OS, Chrome OS, and/or Android and then they have to maintain all of that software as revisions of the various OS’s come out. Also, you have to install it. Depending on the OS, writing an error free installation package can sometimes take as long as the darned software. Then people are going to call, email, text, IM, etc… and you need a tech support staff that not only needs to understand your product, but also how to play IT person to all of those OS’s.

Their other option is to put a computer in the glowforge. This is common in industrial equipment. It adds a couple hundred dollars to the BOM, which means five to eight hundred dollars (or more) to the retail price. Plus, the OS you get, and the hardware interface(s), is pretty much a direct correlation to how much you pay for your computer on a board. This matters, because the better the computer, the easier (read quicker and cheaper) it is to write the software. So there is a trade off between cost to the finished product and engineering cost. But the biggest drawback to this approach is you’re stuck with the hardware you’ve shipped. What if that computer in the glowforge has a security vulnerability? Your laser is fine, but its controller computer is a reservoir for viruses. Next thing you know your glowforge not only cuts and engraves but it also spams, infects your home network and redistributes furry porn. This is an issue in hospitals where equipment purchased ten, fifteen years ago that came with a Windows 98 (or 2000 or XP) PC to run it is now the typhoid Mary of cyber infections.

A cloud based approach leaves most of the security issues with the cloud provider (you still have to follow good coding and security practices), but always leaves you coding to the most up to date hardware and OS. You only have to write the software once, with the best tools available and deployment is a lot simpler. And tech support only has to become experts in Chrome, IE (or whatever they’re calling it these days), Firefox and Safari in addition to their product. For consumer/light business use hardware like the glowforge I think the cloud is the sanest route.


#16

Personally, ive come to terms with the server thing. I dont mind being hooked to the internet, and i look forward to having updates pushed to me. HOWEVER, i really hope there is an option to hardwire it to my PC. Just a usb port would be awesome! Wifi/routers is the weak link in my opinion.


#17

Welcome @stephendchrist to the forum. Are you a recent purchaser or ordered one a while ago? The advantages and disadvantages of cloud computer have been a huge topic from the get go; however, it seems that most folks have come to understand the practical, financial, engineering and many other advantages for having an internet slaved machine. All these advantages mean nothing if there is no internet connection or connection problems. I’ve been having issues with my home internet and can’t isolate why the connection keeps dropping. I haven’t taken the time do do all the troubleshooting necessary. I’ll tell you that if I had a Glowforge, I’d have it fixed by now. I have good friends who live at the end of a road off the beaten path in the country. Not only is their internet slow, but they have dropped connections. It definitely is going to be a problem for some people. We all need to take stock of that fundamental parameter.


#18

I had a dropping issue, and finally tracked the problem to my combination modem/wireless router. Motorola, paid around $80 for it.
Turns out the stand alone units (router) are much better.
Problem vanished when I got a TP Link wireless router for about $25.


#19

I was more concerned about the WiFi-only aspect because our home WiFi network would drop out for a couple minutes at a time, several times a day, apparently correlated to microwave use in our townhome or a next door neighbor. Switching from 2.4 GHx to 5 GHz seems to have fixed that and now my only concern about GF-WiFi is if GF goes belly up. That topic has been covered before though, and there is a fair chance at a resolution should GF close its doors so I don’t worry about it too much these days.

@caribis2 has summed up the issues nicely and what he writes makes sense from my perspective. My CNC lathe and mill use Linux-based control software that has been customised and is free from the manufacturer, but tech support from them is most complete if you buy their control computer which is around $650. That’s in addition to the main control board in each machine tool.

I’m using Onshape for 3D CAD these days, which is a completely cloud-based program. They typically update the software every 3-4 weeks and often apply simple bug fixes within a day or two of first report. It runs on most common browsers in all 3 of the major OS. All updates are pushed automatically to the users so each is always running the most recent version of the software. That has to make tech support far easier. In addition, they make it simple to share a CAD file with tech support or other users which seems to get tech support questions answered a lot more quickly than I’ve seen with traditional HD-based CAD software. I’m starting to really like cloud-based software for those reasons.