Not a fan having to upload projects to the cloud

I just want to voice an opinion now that I have tinkered a bit, and am nearly ready to commercialize some designs.

I am REALLY not a fan of having to upload projects to the cloud. I’m preparing to do some projects that involve topography data, and I’ve discovered that it takes forever to upload to the cloud, process the information, and then “render the design”. Every other laser cutter I’ve used in the past (Epilogue and Universal Laser Systems primarily) behaved like a printer. There are some features that I like about the GF, such as smooth ramping of power via gray scale values, but this whole cloud thing is really turning out to be a major frustration and hindrance. Furthermore, stuff seems to get mangled or lost in translation sometimes.

I’m going to make this work for now, but it might be until this thing pays for an Epilogue or ULS. I hope you’re listening GF. Please create a printer driver or something of the like.

3 Likes

Others have voiced similar thoughts, esp. when there are wifi issues, it can be frustrating not to be able to use your machine because it relies on the cloud.

BUT what makes the GF unique is the proprietary way it takes our files and prepares them for work with the laser reading with the barcode off PG material and automatically sets up the cuts/engraves, or allows us to put in our own entries for material and settings. It would likely be significantly more expensive if each unit had this computing power to on its own, as well as GF loses its proprietary edge once others hack into their unit and reverse engineer it. So don’t expect GF to change the operating model any time soon. What makes it easy to use and unique can also be frustrating to deal with.

Yes, it is a major difference from other laser systems having to rely on the cloud but this system makes the unit so easy to use in many other ways, especially for those who have not spent time already working with other laser systems. Alas it sounds like you may be better off with another system for your projects if they are are so large.

An idea may be to break up the file into multiple files, and run consecutively (without moving the material). Might make loading faster, though maybe the actual job longer…

Definitely, don’t expect them to do this. Over time they have become more hostile to the hacker/tinkerer community not more open. They have even removed the debug port from the PCB on current models, luckily mine still has one.

This is how I see the company going forward. They have after all taken $35 million in venture capital, in addition to the $27 million from the crowdfunding campaign, and are going on 5 years old. It’s time as a business to take steps to stop operating on “other peoples money” over the next 3-4 years.

  • They will never have a 3rd party development API.
  • They will never add an ethernet port.
  • There will never be a local print option.
  • They will start to monetize all the user data they are collecting.
  • They will eventually start charging for using certain cloud features.

Proofgrade is their subscription model, if enough people don’t buy this they will have to find new sources of income and that will likely be us, the users.

Things that I’m concerned about with cloud software:

  • Censoring files that we can print (DMCA, files they don’t agree with, etc.)
  • Ban users from their purchased Glowforge for violating their EULA
  • They can remotely log into your home network and do whatever they want (and based on the support tickets I’ve seen they routinely do this to collect logs)

It is a cloud offering and their EULA can change on a whim. Our choices will always be to accept those changes or to brick our device, that is the cloud.

People will probably look at this and tell me I’m wrong and being pessimistic. However, Glowforge is a company and their goal, especially as a VC funded company, is to be profitable and cash out.

@bansai8creations a couple of counterpoints to things you said there. Firstly, I doubt they would put the computing power in the laser cutter itself. It would most likely come in the form of a post-processor that runs on your laptop/desktop. Which, if it can run photoshop, it will be capable of running the processing GF is doing in their “cloud”.

Secondly, I made a chrome extension that allows users to create their own custom Proofgrade at home. It reads custom QR codes and sets the material for you. I’ve created a rubber stamp to simplify the process for me since I mask everything anyway. :slight_smile:

2 Likes

I am not sure about the processing power required to prepare the files for the Glowforge : a common laptop or desktop computer is more than capable to do that.

Lurking around makers communities online, the “mandatory cloud-based software” is the most frequent criticism. And it’s one that’s hard to argue with.
Proprietary software is bad, hidden in the cloud or not ! It is limiting our freedom as users. Look at the Snapmark feature : even if you feel adventurous, you cannot try it. Should I mention the scariest threat : What if Glowforge servers go down ?

Clearly, operating the Glowforge is easy thanks to the software. But I doubt that the company would lose much when then will release their source code (I believe they said they would :crossed_fingers: ).
What makes Glowforge special is their reputation of trust and quality (the machine itself, the customer service and the community) and the PG materials (convenient and reliable). Right now, I am more optimistic that @icirellik, and I still hope that they won’t dive into dark patterns of monetization :slight_smile:

I am using the Glowforge and loving it, but I’ll be happier once the software is open. Then we’ll be able to build a way to work computer-to-Glowforge without the Cloud middleman, and we’ll be safe from potential future corporate shenanigans. Shouldn’t that be the sales pitch ?

@bansai8creations:

I appreciate your response. I’m not sure you’re understanding my issue, however. I use high end design software such as Rhinoceros 3D and Adobe products, so it seems like silly that I should have to use some janky open source thing (Inkscape). However, if choice of software is what causes GF’s cloud to literally not process some of the stuff I’m feeding it, even though I’ve tried outputting as a vector PDF, or WMF, or AI or DXF, they’ve got a serious problem. I’ve tried various different PDF postscripting software (Bluebeam, PDF Creator, the built-in Microsoft PDF creator) as well, and they all have the same issues. I’ve also tried importing files to Inkscape and outputting to SVG files, and still have issues. The main thing is apparently solid hatches (or a “fill” if you prefer that terminology) are either totally messed up, or in the case that I encountered last night, cause it to get stuck in the “rendering your design” phase. The reason I wish they would just make the dang thing work like a printer is then there isn’t any need to output to any file format. It just acts like a printer, and to your point about computing power, the laser doesn’t require any more than your typical $80 inkjet printer. The computing is done by your computer. It would just need enough memory to store print jobs or portions of jobs. Besides nothing I’m doing at the moment should require any significant processing power on the part of the laser anyway. Most of what I’m ultimately producing is around 1 mb, but have LOTS of vector curves and a few fills. The software limitations and cloud have turned into major hurdles as noted above, though. This means nothing about their system or process is “easy”.

To give more context, I’m using Heron (GIS data processing plugin) for Grasshopper (node-based scripting plugin) within Rhinoceros 3D (3D modeling). I use the 3D topography information extracted from USGS data to produce a grayscale bitmap that can be 3D engraved, or extract topography lines to etch on a surface. The projects also include some text which would be engraved. At issue here is that, as far as I know, there is no possible way to achieve that with their preferred software, and my output, although fairly basic compared to where it comes from, is not playing nice with their cloud. I also use Grasshopper to do parametric automation so that I can rapidly customize things. It’s a major problem if their systems can’t accommodate users designing with that kind of software, and operating like that.

In summary, my beef is with the fact that I’m apparently limited to using the Pinto of software packages, the laser is bricked without a wifi connection, and I have to wait, and wait, and wait, and wait for their cloud system despite relatively small file sizes. On top of it all, I’ve been finding that their “proof grade” settings are pretty bad too, meaning I have to do test after test. And finally, after fighting the issue of solid fills causing an implosion of their cloud system for hours last night, I wound up getting stuck with the “lid open” error!!! Further research has shown that basically means this unit is a brick, and I haven’t even accomplished anything of consequence yet.

It seems as though you’re very happy with yours, which is awesome, but I can’t seem to find one good thing about mine right now.

@icirellik

You’ve added some interesting points to my technical concerns. I will almost never use their proof grade materials, as it only makes monetary sense to buy in large sheets and cut down to the bed size. The potential monetization of data would certainly be more typical, unsurprising, groan-worthy tech start up nonsense. I wonder if these guys are Facebook or Amazon alum? I also wonder if I’m going to violate my EULA for griping about their shiny garbage? Not that it matters, since some software glitch in their cloud means my unit is likely a paper weight until they send me a new one. Which brings me to…

@Julien.dubuisson

I am glad to hear that your experience is good! From the view point of my experiences, I have to take some exception to statements about quality, reliability, and service though. I’ve had the thing for just over a month and its cloud system already got it stuck in the “focusing” mode once. I got it out of that. But now that it’s stuck in the “lid open” thing as mentioned above, quality and reliability are not adjectives I can use. I’ve got a ticket in to have that issue addressed, but I hate to imagine how long it will take to get service on that based on another time I sought support and it took over a week to get a response. If I do stick it out with GF after all this, I hope you’re right about opening up the software, and removing the cloud process.

At this point I’m hoping GF can get me back up and running, so I can see if my side hustle scheme is viable. If so, I think I’ll make just enough to go buy an Epilogue Laser so that I can just hit the dang print button!

Phew… well this turned into a rant. Thanks for being a shoulder to cry on, strangers! Now if only GF could wipe away the tears.

1 Like

Sounds like your needs may exceed those the of the bulk of the market the GF was designed to support. But maybe others will chime in that have a lot of experience using complex designs and files may offer some advice on how to optimize & run such large files (it’s amazing how the community steps up if you ask for advice–I do understand you’re frustrated and in a tough spot and venting/complaining, and I hope you can a job finished soon!).

As far as PG or not, I think most of us keep a notebook or spreadsheet or post-it notes scattered about with settings that worked for our last materials, and many of us rarely use PG material (I work in leathers mostly, and very different from PG versions). But material densities vary, (e.g. grains & knots in wood); glue types vary (not only properties affecting the etching/cut, but the toxicity of glue). And acrylic properties can vary from manufacturer to manufacturer, too. So don’t be surprised if what worked great on the last job isn’t working as well as you want on the next piece. After a cutting job, I’ve learned to NOT move the material on the crumb tray until I’ve confirmed it’s loose & fully cut…

And I’m not a huge fan of relying on the cloud, but it’s more stable than relying on isolated servers located in only one building. If multiple server farms crash, then there are a lot more things I’ll be worrying about than my GF job…

1 Like

Well I do have Snap Marks, and have made use of them on some projects–and it is a shame they’ve not rolled it out to all. I think their new calibration program was to replace a lot of SM were supposed to do–but not everything (e.g. running card stock through a printer for the text, then using the SM to align for the outline cut…).

There’s a risk for any server crashing anywhere. The advantage to cloud is that the server farms can share jobs and loads, so if one has an issue, it may cause a delay while jobs shifted over, but it’s not completely down. If it’s a big enough issue that really crashes systems, I think there will be more to worry about than the GF…

And I don’t feel like my freedom as an crafter/artist is limited by GF. The GF is an amazing tool that I chose because of what it can do now to help my work and ease of using it. Now if my craft or interest was in coding and hacking tools, then I can see how having proprietary SW would be considered limiting and I’d not have chosen the GF. But most companies I worked for (not SW) only survive because much of what makes them who they are remains proprietary.

And the PG materials are great–but 90% of my jobs are not on PG. It’s a starting point to minimize testing to ensure result is what you want, and it’s great they allow us now to save our own settings for whatever material, so less a hassle wondering what happened to that post-it note (and if they were relying on the PG sales for their business model, they would not have added this feature making it easier for us to run jobs with non-PG materials…).

The good/bad of tied to the home office–the ease of sharing files/designs and nearly instant running of those jobs. Also the troubleshooting they can do…

Some comment mentioned risk of “hacking into our networks”. Suppose there’s a risk. Since the GF needs 2.4, I have it on a separate network than my other main devices, so if someone did hack into my network via the GF, they’d be pretty disappointed in what they’d find. But you could also turn off the network connection when you’re not using the GF if you’re really that worried… And I trust GF way more than things like FaceBook!

But this is only my opinion–the joy of our current market is we have choice for the tools we buy. Read the specs, choose the best tool for your needs and that you can afford (initial price, installation and operating costs).

The firmware was released early last year. (Github.)

So do I. :slightly_smiling_face:

Helps to read these tutorials to understand what your options are regarding file types, and how to save them correctly for the GFUI. Once you have them set up to save correctly, and understand how to set up your designs, life gets a lot easier.

For your specific complaint regarding topo maps being too data intense, you can make the Glowforge act just like a printer if you rasterize the data first. (Couple of clicks in Illustrator.)

1 Like

the dark side of the cloud connection only hit home yesterday when my internet service went down for a few hours. Not my fault, not GF’s fault. Net result is that my favorite tool was a brick until service was restored. I did not have the time or mental bandwidth to set the connection up with my phone (which is not a great option either given the poor cell service here) . I can touch the computer and the forge at the same time yet I was S.O.L.

Yep. Fortunately doesn’t happen very often around here any more. We had a few early issues with it when construction out on the main road resulted in the stupid cable being cut every other week, but there isn’t a hell of a lot you can do about that. (Since they also tended to take out the gas mains at the same time, the lack of internet connection was actually kind of low on the priorities scale. There was all that angst regarding the house possibly blowing up, and the shelter in place orders.) :laughing:

Seriously though, if being without the machine for a few hours occasionally is a life altering problem, you actually are better off buying an Epilog or Universal or some other laser, so save up your pennies.
I think I got one hell of a deal for what I paid for the Glowforge though, even if I can easily afford any of the other ones. This one does everything I need, and for a heck of a lot less lucre.

(And I LOVE a bargain. Makes me feel smart.) :smile:

3 Likes

Jules,

Thank you for providing those posts, but I had already read these forum posts, and others, regarding file types and output/input. Also, this very much isn’t my first rodeo with laser cutting/engraving. I ran the laser cutter shop at Texas A&M University’s School of Architecture, where we had two ULS lasers and an Epilogue. My original point to this entire thread is that the print dialogue thing is much easier to use, doesn’t rely on an internet connection, and never failed me regardless of cut, raster engrave, or score line work combined into a single project. I could simply set up layers with colors that represent different laser usage, and hit print. No cloud.

The recent issue that I raised, which was the genesis of my original posting, was not based on universal failure to get results anyway. It was based on the fact that I have really had to fight to get solid hatches produced in Rhinoceros 3D to engrave properly and consistently; not be somewhere between kinda messed up and totally messed up. Or in the recent case, cause the GF cloud system to get stuck in the “rendering your project” phase. You may have missed all of the file types that I’ve exported to as well (PDF, AI, DXF, & WMF primarily). And again, I’ve imported this stuff to Inkscape to attempt to get it into a more GF friendly file type. This is a problem I never had to suffer on other systems. Anyway, once I eliminated the hatches and simply scored the outline of the text, everything proceeded fine. But that still doesn’t change the fact that I really dislike GF’s cloud process, but according to other responses, I may as well accept it for the time being.

As for outputting the topo curves as a raster to then engrave, that is a non-starter. Some designs that I intend to produce fill the entire bed, and take 1-2 hours to simply score. Engraving that stuff would take many, many hours. Time is money… Besides, while the information import to Rhino is data heavy, the ultimate output is relatively small (few hundred kb). And again, the topo lines aren’t the issue. It’s the solid hatches that are intended to be raster engraved. Adding more steps by outputting different files for engrave and score/cut, then re-compositing them in Inkscape seems like a wholly silly affair. Time is money… My Rhino file is set up with layers, ordered in the way I would like them processed (based on a forum post for someone else with a Rhino workflow. 1) Engrave, 2)Score, 3) Cut 1 (inside), 4) Cut 2 (outside), 5)Layout (ignore). Once the file is imported to the GF cloud, I set my settings accordingly.

In summary, none of this is not based on a lack of reading, or understanding of how laser cutting works. It’s basically entirely a gripe about the GF interface, and the fact that in order to get consistent, decent raster engraves from solid hatches produced by CAD software, I would need significantly increase my cut path production time by adding a bunch of other steps. For me the GF’s lower price tag was always intended to make it stepping stone to bigger, badder lasers, but I really didn’t register how goofy their cloud system makes the whole thing.

Of course, none of this really matters right now because my unit is bricked while I have the dreaded “Lid Open” issue, and am just sitting around, waiting for GF to get around to my ticket.

Yes, it does sound like you need a different kind of laser. (AFAIK, this one is going to stay cloud based.)

1 Like

I see you have already emailed us about this and I have followed up there. So I am going to close this thread.