Hopefully, I only have to endure 6 more weeks with the K40.
(Checks Email) Sigh…
Untrue. The only thing at issue is the timing. Some folks seem to be upset that the Glowforge interface does not include all of the functionality now that took the developers of the Silhouette Studio software many years to develop. (Just about eight of them, IIRC.)
I worked with WinPCsign and the Robomaster software, the precursors to Silhouette Studio, and they were the kludgiest things I’d ever seen, with absolutely zero functionality. Manual tracing was literally point-click-point-click-point-click-drag-click. There was no auto-trace at all. Text had to be individually traced. Each letter. By hand. Painstakingly joined by editing. The Silhouette Studio software took from six to eight years to develop from those humble beginnings. A lot longer than this did.
Don’t discount what this software is currently capable of doing, they have written something almost as good as the Silhouette Studio software in about a year and a half. It’s already a powerhouse. The core functionality for the rendering is rock solid. (And better than the Studio software, IMO.) They did an absolutely astounding job with it, that very few people will be able to appreciate. (That part really does make me very sad.)
I fully expect to see a few bells and whistles pop up over time. They’ve already added quite a few of the goodies that we’ve requested - they’ll add more as they get the time to deal with them.
But that’s just my take on it. (Having to drag a shape over from the desktop and drop it onto an open project instead of clicking on a button in the interface isn’t really slowing me down much, so I’m totally cool with it. Obviously it’s a big deal to other people and GF knows that as well. It’s one of the items that has been discussed as an addition and has been added to the hopper before this, so they do know about it.)
The improvements will come. Just takes some time. 
5 Likes
I am happy with the software as-is. Sure, I’d like the auto-alignment with passthrough, and more precise visual positioning, and the ability to save projects as new “documents”, but none of those prevent me from using the GF just fine now. So while there are lots of things I’d love to see added, they’re not blockers, they’re 'nice to have’s".
My basic workflow is - design something, one color each for etch, score and cut. Use a 12x20 page size if I care about precise positioning (e.g. cutting a sheet of keychains) In the GFUI, create a new project from upload. On the bar on the left. make sure that the colors are assigned to cut/score/engrave as I want. If it’s not ProofGrade, select the material or manually set the cut/score/engrave settings and material thickness. Click print. Press the glowing button.
Compared the Epilog, it’s a dream. Yes, I can make the Epilog do what I want, but it’s “work” to do so. The GF is easy and fun. Plus, my kids are using it more than I am.
The ‘scan an image and cut/engrave it’ works surprisingly well, and is very easy. That being said, a clean EPS/SVG file is a lot sharper than a camera scan. So it’s great for getting kids’ drawings onto a physical object, and in making the GF accessible to people who are more comfortable drawing on paper than a computer or tablet.
The only problems I’ve had are when I’m on a saturated or crashing WiFi network. The GF hated that. My home network is fine, so the GF is happy.
11 Likes
This is a very complex issue.
Let’s begin by defining software. You know what my favorite print experience is? Ctrl + P from within InDesign. I have done a cookbook. It is print ready. It’s a combination of document parts that just comes out of my 2D laser printer perfect, ready for my plastic comb binder. I can do document presets that make the software needed for the printer hardware invisible to me.
The most complicated procedure for me in turning ideas into real objects is the workflow for the 3D printer. Onshape > Export as STL > Open in Slic3r [haven’t integrated Cura or Slic3r into Octoprint yet, but whether or not integrated into the print software, slicing is still a discrete step] > Prep settings for slicing > Slice > Export gcode > upload into Octoprint > Print. [There may be more efficient work flows, automation scripts and presets to assist, but basically this is what works.]
Between these two workflows lies an extremely complicated context of hardware. From design to print lies an extremely complicated constellation of values to select, prioritize. and build out.
One experience I have had recently with working with a some new hardware is the Silhouette. I had to pay to be able to upload SVGs. Before I paid for the upgrade, I laughed out loud on so many occasions (@jules’s enthusiasm notwithstanding) because even though the software had some decent design functionality, it wasn’t quite as responsive as I was used to in Inkscape. I pretty well dumped designing in Silhouette as soon as I had the SVG capability and used it primarily to adapt the design elements to different operations for controlling the cutter.
So Glowforge software:
-
Design storage: Let’s call it pre-basic. You can upload designs. You can retrieve them for later. You can delete them. You can’t sort, tag, or put into folders.
-
Trace: You can do some basic tracing to cutting and engraving a design in the bed. You can change contrast a bit to get an image that works better for your needs. It’s useful and is a fun trick. The results can not be compared to high resolution laser scanners or custom lensed large format photo scanners. Hobbyists and the average person with few expectations will enjoy this function. Pofessionals and artists will find ways to play to its strengths or will discard it as not relevant to their workflows.
-
Hardware control: This is where we should be focusing our criticisms and expectations. We’d like as much of it automated or controlled by presets we can save and recall. It’s a complex mixture of many things, each of which deserves its own analysis and punch list. For example: the hardware has the ability to focus on different planes in the Z axis. Standard stuff for any laser cutter/engraver. Basic functionality. Proofgrade QR codes is an interesting solution that most of the time for Proofgrade allows you to ignore any issues with material thickness and concentrate on other aspects. We can however enter in material thickness manually to control lid camera image processing. And then we can do granular control of the focus for individual operations. There are many other aspects of the hardware that deserve their own analysis and opinions about how well or poorly the command and control software works with the hardware. Head camera control would be a very robust discussion at that moment.
-
Design Control: This is the software that takes individual design elements and runs it through the laser beam itself. Boy, is this a challenge to evaluate. It’s so complex and the software is at many different levels of maturity. Take the very simple choice to map color to operations. It worked pretty well from the start. The funny thing was that the whole ordering of operations by RGB color values was undocumented, as were/are many other features. A major change to operations software was defaulting engraving to first order. Another feature that is software controlled is engraving settings. They have evolved quite a bit from 11 months ago. Recent changes have been a welcome improvement. There is currently no way to save how you set a design up. That is a pre-basic functionality at the moment. There is pretty good granularity you can get with the designs and operations settings. It’s a big issue. No indications from how things have progressed that this feature is close, much less on the horizon at all. Another design control issue is text processing and clipping paths. This is a feature that could be available but it would invisible in general. Yes a software feature but a good example of the hidden software that people who have little experience with graphic design or laser cutting and engraving might never miss or might even expect 'it off the bat. “Hey, I put some text in my design and Glowforge says it can’t process text yet. What gives? It’s a printer. Printers process text. I don’t understand?” The 20x12 artboard feature is the unsung hero of the Glowforge. Couple that with some jig work, and you can do some precision work. That it bypasses the lid camera placement feature, ironic, isn’t it.
-
Design capabilities: From my needs and expectations, this is low priority. Let’s say that manipulation of uploaded designs is part of this feature set as a basic functionality. We can rotate (vectors), we can resize, we can reposition. We are able to use some constraints that are keyboard activated. There is room for growth there, especially as many have pointed out in using numeric entry for object manipulation. At the moment, native design elements are limited. You might want to include the trace feature as part of the design capability. That’s a pretty big rabbit hole of a discussion. To be able to have access to primitive geometric shapes, that would be handy. But from my experience with Silhouette software, I would rather not be inflicted with a half-baked array of tools that over promises and under delivers in comparison to how I use design software. Once the feature went live to be able to upload SVGs in the operations space of the GFUI, I pretty well let it go. Biggest to do in this category is once again, saving your work for later.
-
User Interface: I defer to others on this. I haven’t decided yet if Glowforge is going for minimalistic or it’s been a challenge to devote needed resources to improving the user experience. When I first opened up the app and started processing a design, I was puzzled. What is a feature and what is just a placeholder for future development. It’s changing. Ditching the image thumbnails for choosing engraving settings is a big change. I welcome it. It’s a text based menu. The icon deal just never sat right with me. I could use it, but I could never quite puzzle out how the intended results were mapped onto the icons. I have suggested what I think is a basic interface functionality that really cuts into my normal productive workflow. Tab through the settings input boxes and that works but using a keystroke to go to the next operation, it isn’t a feature. Please click somewhere off the popup, best place the next operation, to close this popup. UX evaluation is very complex. It is a matter of taste, priorities, and I would imagine the state of the underlying control software. I can do what I need efficiently enough, but look for improvements.
-
Lid camera image processing and fidelity to design placement. Nope, not going to touch this issue other than it fits with what I use the Glowforge for. Would I use the Glowforge differently and do different designs if this worked differently? Maybe. I’ve seen improvements. I’ve seen some true defects. I’ve seen remediation of problems. Not sure about this, but the FUD generated by the differing opinions, evaluations and experiences with this feature must be a major marketing headache. Ironic since this is often the defining super power of the Glowforge. For me and my use cases, it is pretty handy. The folks who come by and use my Glowforge are super impressed. They see it and start going on and on about how they could use a Glowforge. It does nothing more, and in some cases less, than what any laser has accomplished before with hard stop margins, zero positioning and and numeric entry of coordinates. But the camera positioning has an attraction to folks that makes a pretty persuasive dog and pony show. Perhaps the cake is a lie in the end, but it’s pretty impressive as is. Flip and process accuracy? It was part of the wow factor that rationalized the purchase, that I could do 1/2" hardwood. I would welcome this, but have no idea how it would even work or be implemented. One huge plus and is very functional is that you can use the lid camera to easily place small designs and odds and ends on bits of material.
-
Hardware development status: This would include the case, the lid the button, network connectivity choices, the crumb tray, the gantry, the cooling, the exhaust, the dimensions of the machine and the dimensions of accessible areas. Add in the pass through slot. Laser power profiles. Holy cow. This is a big subject. I can’t muster the energy to tackle this one. Gut reaction? I am impressed by the build and design of the machine. I have no other experience with lasers. I have experience with a wide range of tools and consider this on the upper end of things. It looks deceptively simple and in the end is remarkable uncomplicated as to hardware. But then laser cutters and engravers are pretty simple machines from a parts viewpoint. Power supply is pretty important. Looks like that was quite the challenge. Tube quality, reliability and life span are crucial points to consider. The cooling system was/is ambitious. The summer really challenged the system. That the hardware seems to be fairly capable is evident, but operating temperature was a huge issue that got little discussion before this past summer. I certainly didn’t consider fully enough all the ramifications. I just heard “built in chiller” and said sign me up. That said, I have had zero cooling issues with my pre-release or production unit. So glad I have air conditioning in my house.
-
Cloud-based system control. Deserves a discussion. Can find some good evaluations in the forums of the concept in and of itself. But once you go beyond questioning this development choice and ask how it works, wow, that’s complicated. Yes it works. Yes it becomes transparent and I am able to ignore this aspect, It just works, that’s all I care for. But sometimes it becomes obvious that this is a cloud based computing model. Hard to tell if some things are a bug or a feature. In any case, wifi connectivity issues, timeouts in processing or glitches in the matrix have a lot of potential for undermining this design choice. Still trying to discern the nature of the printing errors that pop up with large engraves. Absolutely is an essential issue for the evolution of the Glowforge. Probably could stand some more documentation of the conditions that create errors and also offer a more robust error reporting/explanation routine. Not just documentation but above all information rich error messages. Aside from problems that are inherent in cloud based computing, I am so thankful that I can control my Glowforge from any of my three major desktop platforms from anywhere in the house without having to run a server like Octoprint or do remote desktops. I would hate to be tied to one OS and bespoke software that you have to install as an individual app. Device agnosticism is pretty cool, but has some more to go when getting to the mobile OSs.
-
Shipping status: Deserves its own three hour ramble. Special shout out to International. One thing I would like to say: given what we know now if anyone starts up a new hardware enterprise that has significant regulatory and shipping encumbrances to it and they don’t caution those who would be potential international customers thinking about a pre-order in their initial marketing prior to full product development, they would be pretty much jerks. The narrative of Glowforge international marketing and shipping should be required reading for startups with this potential for screw ups. Ignorance can now be no excuse. You have the power to market internationally, use that power responsibly. And on the customer side, international or not, do not let the marketing video alone sell you. Probably I will avoid the drama of a pre-order from now on. It just takes too much work to have due diligence.
Ok. I’m stopping now. I’ve pretty well convinced myself that evaluating any aspect of the Glowforge in detail and addressing the wide variety of use cases takes a lot of time, energy and discretion and an almost impossible level of objectivity to write a compelling narrative that gets all the details and the big picture right.
Lucky we have a forum and all this stuff is there, just waiting to be collated, cross-referenced, indexed and summarized. Luckily we can do that with a simple search engine.
17 Likes
I don’t know how you can say that when it falls over with no sensible error messages if your file is too big, too complex, has an image transform, etc. And it gives incorrect results if it uses fill-rule:evenodd.
I think I have a different definition of rock sold software. It should accept all standard conformant SVG files and render them correctly when they are within the hardware capabilities of the machine or give a descriptive error message when they are not.
6 Likes
I appreciate the time and energy you have taken here. This is a useful sketch of what it’s like to work with the software. Objectivity is largely an illusion in any case; I wouldn’t worry much about that.
I am a little puzzled, though: several other users’ responses were to the effect of, “What software? There’s not much to it; you press a button and it lases.” That is decidedly not the impression I walk away from your post with.
1 Like
There is a lot of software, but you only directly interact with the tip of the iceberg. Before I received my unit I didn’t understand how truly little there is to do. Now that I have my unit I agree releasing this to people doesn’t make any real sense.
You indirectly interact with a little more of the software when you upload a design and the GFUI either refuses to process it or doesn’t do what you expected. As this appears to be all cloud based they could allow anyone to try processing designs to understand what does and doesn’t work. Without the ability to actually process material would this be of benefit? I’m not sure. There are some things it’s hard to understand until you have seen how it cuts and engraves and other things where that knowledge is not relevant. I believe there is a way to make this beneficial to some people, but will they spend the coding resources to do this? I doubt it.
6 Likes
The best way to think about it, is this would be like releasing the print dialog from your mac. That’s about the level of functionality. You can set a few options (placement, power settings, what does what (engrave/cut) but that’s about it. Not that there aren’t a few subtle things, but in general it’s about as complex as printer settings in MS word…
6 Likes
I was trying to get at this point in making distinctions between the command and control software that is invisible and should be invisible to the user and the user interface that allows some input, but really can appear simplistic. In reality the heavy lifting is going to be done in your design software where I’d have more robust tools than Glowforge can provide at this time baked into the user interface.
Ctrl + P is a very simple software command. Deceptively simple because one combined keyboard operation takes control of vast lines of code.
It is helpful to define the many facets of software at work in the Glowforge project. It has to be divided up into categories or areas of control and interface. I’m trying to examine my own understanding of what constitutes software, especially since my primary model or understanding of software is a stand alone application that one must download and interact with to make function. Then I am trying to understand the hardware functionality. Then my use cases.
I resonate with your desire to test drive the interface. From early discussions on the forum, folks wanted some type of sandbox to play with so they could get a sense for what to expect. Just to explore the user interface and see what features are present on screen that require or allow input. Interface evaluation when you don’t have a way to test it out is a hard problem. It can be solved with some effort without having to do a physical meetup: find someone who will allow you to being a subuser for their Glowforge or find someone who will allow you to remote into their computer. (The third option of just giving credentials out is a possibility, but as that in normal understanding would be contrary the user agreement, I am fine with pointing out the possibility but wouldn’t suggest that folks do it.) I did the remote access thing using Google/Chrome remote desktop with a few folks from the forum in the early months of my pre-release use to assist with this very thing. Their experience of the app and interface did raise questions in their mind about the software development process, but did not lead them to cancel. In fact, both ended up getting pre-release units, one having some burdensome issues with the unit they got, but they both stuck with it and are both looking forward to their pros but continue to make some amazing stuff in the meanwhile.
The other facet of software is how robust is the command and control feature set. While there are some functionalities that everyone can agree on which must be present and which are now operating in a high capacity, e.g., turning a vector in a drawing into a cut, turning a bitmap into an engrave, there are things that were promised and expected by many people but aren’t present, e.g. automatic pass through slot registration. Getting some data points about all the features, assigning values to these features, making a punch list of their present capabilties and then displaying that data in some type of Edward Tufte-endorsed visualization would be pretty helpful. Until that takes place, many folks are stuck with a huge information deficit.
Could you provide a link to the posts about this that you are referring to?. I’d like to review their perspective. I can recall someone selling for whatever reasons but didn’t register it well enough to factor into my own narrative.
Would love to hear more about your own plans to use the Glowforge, I note you have a Carvey and thus should have a pretty good context to evaluate the Glowforge’s capabilities. Have you made a punch list of go/no go features that can be added up to see if the Glowforge is right for you? Are there software horror stories you might like to share that might contribute to your uneasiness about what’s in store for using the Glowforge? In my case, I am hyper-sensitive to software that becomes orphaned, only works on a certain platforms, and becomes a money sink. I gravitate to open source projects because there is a control or access to information about the software that makes me feel a bit more secure about investing my time and mental space learning the software. For example, I used Finale music scoring software for many years. I loved it and it was the best wysiwyg music processing software available to me, and since I taught, the school bought the software. Once I left teaching, I could no longer take advantage of the upgrade cycle, or chose not to because I could not justify the price increases for upgrades. Concurrently I started getting into open source software and found lilypond, a totally text based scoring software that I could use on multiple platforms. It came with a huge library of music already scored. It makes beautify scores, I guess one could think of it as the LaTex for music. Fast forward a few years. MuseScore comes out. Boom. I was able to switch easily, have access to a free scoring program, and get productivity as I never had before. So when I evaluate the Glowforge cloud approach to software, that is the personal context I evaluate it. Functionality is crucial, but in one sense just a part that has to play well with other values at work. My experience of using the Glowforge and getting the job done is not frustrating. I don’t have negative emotions or have to force myself grumbling to engage in an interface that fights me at every turn. I figured it out. The software then becomes invisible to me in the workflow. Not that there aren’t error messages or compromises in productivity. I am almost at the point to say that if I first heard of Glowforge today and had to pay $3,000 plus shipping, and got to see first hand of its capabilities, I would imagine myself ordering it. I say almost because the whole shipping/delivery time-table would have quite an impact on my decision. The rest of the decision would be on my evaluation of Glowforge sticking around as a company and keeping the servers running.
Hope this helps you clarifying your decision making strategy.
3 Likes
Thank you again – and to everyone else on the thread – for the time you are taking to address this question. I appreciate it.
I am quite certain there was a fairly recent post to this effect on Reddit, but I cannot now find it. I am not sure if it has been deleted or if I am searching wrong. That post did not give much detail, but as I recall it described the software as “clunky” and that it yielded inappropriate toolpaths. I did not give much credit to the latter point as CAM often comes up with toolpaths that are unintuitive for human users.
I recently had the realization that by not making a decision to cancel, I am effectively making the decision to purchase. I now have more information than I did when I made the decision to hand over non-trivial amounts of money 2+ years ago, and so it seems like I should be revisiting that decision accordingly. One new piece of information is that Glowforge does not reliably deliver on what it says it will do, and so it seemed to me like I should take a close look at what the state of the software is right now rather than bank on the notion that “it’s getting better every day!”
I run a small hardware lab for a company that does international development work. I sometimes need to cut plastic (acrylic and polypropylene) for internal and external mounts for devices, and this means a trip to a local makerspace, which means cutting parts takes me several hours. Having a laser in the lab would save a lot of time. I also prototype the shapes of PCB designs and other components to make sure they will fit well in an enclosure with all the wires attached, etc.; right now I do that kind of thing with the Carvey or a Silhouette Cameo, but lasing them would be faster.
So my needs are actually pretty simple, and while the impression I am getting from this thread is that the software is kind of half-baked, it sounds like it is mostly adequate to what I need. It does not sound much worse than, say, ULS drivers.
Yes, this bothers me, too, and I am getting increasingly uncomfortable with the cloud-based model as my confidence in the company declines. I am starting to wonder what will happen if the company gets bought and the buyer has no incentive to keep the servers running.
3 Likes
For your use case, assuming you can find materials of the appropriate thickness for stackups, I can’t see the software not working very well for you. What CAD programs are being used there you would want to export shapes from? Posting a question asking if anyone has had a conversion issue from the program(s) used where you work to svg files the glowforge likes would net you good feedback. Or, if you can do it at your local makerspace you can probably do it easier with a 
. Without knowing anymore than you’ve written I’d think the delay between laser and filter would be a big issue for you.
You never know with new companies, but using it I find it hard to believe somebody won’t keep the lights on trying to make it work for at least 5-7 years. Or as I think of it, enough time for viable alternatives to appear. The only way I believe they can fail is if the market for a prosumer laser cutter just doesn’t develop.
3 Likes
Agree with you on the cloud based software making me nervous as I am no longer very trusting of glowforge as a company. So much so that im debating spending 3x the money on a domestic laser cutter. But I also think that if I can get my 2k ise out of it before glowforge screws it up for me then I might not mind tearing it apart and re-wiring the motion system to be controlled by mach 3 or something.
1 Like
The only problems I have with dxf export from onshape is when I forget to set the workspace units to inches, then nothing works right…
1 Like
And from Fusion 360 I had to do mm until this latest version of Inkscape, now it does not matter.
If a company buys Glowforge, it will have every incentive to keep the cloud software running. That is an integral part of the value of the company. The reason to acquire such a company is to make more machines and sell them, growing the user base so the consumables (Proofgrade) business has more of a market.
5 Likes
More often it’s to target some intellectual property or absorb a team with a set of expertise and experience you’re interested in.
3 Likes
I don’t know about more often but definitely often, I don’t think I personally could work with investors money for this reason. Every dollar you take you lose say in your own business.
1 Like
I bought a Seagate NAS for backup a few years. That had a cloud service that allows you to access your files anywhere. I never activated it because it seems a a very bad idea to put all your files in the cloud. A few months ago the cloud service got shut down with no replacement. So anybody wanting to access their files remotely is now screwed.
2 Likes
There’s alot of well written info here and I won’t argue with anyone.
But from my take the Glowforge software is quite adequate.
I cannot store my designs on my HP printer, my Cricut, my 3D printer, my CNC router. I store them on my hard drive and access them as needed. Why should we expect the Glowforge to do more than that? I wish I could save them with prior cut settings though.
Same goes with file manipulation. This is better suited in design software. Our $1,000,000 lasers at work don’t do this. We design in CAD/CAM and sent the cut files to the machine. I don’t expect anymore from the Glowforge.
I feel Glowforge is doing a wonderful job trying to make the ability to laser cut user friendly and easy. It’s always a compromise between features and intuitive seamless interfaces.
6 Likes
But in that case, printing your files is basically, just hitting print from the application, at least for the HP Cricuit. For the 3D printer and CNC, you can probably save the print job to a file, so you can just hit print later.
Here we have the web browser for the Glowforge, an intermediary step between application and actual printing. On top of that, there are lots of options for settings and materials etc.
People who do production work won’t want to have to re-enter settings each time they need to print for a client.
3 Likes