3d laser printer
September 24, 2015 2:39 PM   Subscribe

The Glowforge 3d laser printer "The Glowforge simplifies laser cutting by moving software to the cloud and making use of smartphone sensors. " [via]
posted by dhruva (43 comments total) 7 users marked this as a favorite
 
was excited about the small form factor and price, but I don't see why they had to go pour a bucket of stanky cloud all over what could have been a nice home laser.
posted by potch at 2:55 PM on September 24, 2015 [10 favorites]


I really want this. In my opinion a laser cutter is way more useful than a 3D printer -- they're reliable, fast, and can make objects out of a huge variety of materials. Sadly I'll probably never spend $2000 on a toy.
posted by miyabo at 2:55 PM on September 24, 2015 [1 favorite]


I don't see why they had to go pour a bucket of stanky cloud all over what could have been a nice home laser.

Easier centralized updates, less computer power required on the device, the usual cloud reasons.
posted by GuyZero at 2:57 PM on September 24, 2015 [2 favorites]



For normal CNC type laser cutting, this doesn't seem like that big a step up from what's already out there. I think the coolest feature is the one where you can draw on stuff and then it will cut/etch where you drew, which seems cool because you don't need a computer, but it's disappointing if while operating that way it still needs a network and won't just do its thing on its own the way all the other non CNC tools do.

I fairly recently got into laser cutting, and most of the features of this seem like they would make the process a little less fussy but not in a way that would really transform it. The database of settings for different materials is a little better than having the same thing on paper. Automatically making the test cuts wouldn't save that much time over just doing it myself.

The one that seems potentially game changery is the one where you can draw on your material and it'll use the camera to etch the stuff you drew, but I'm great with a computer and terrible with a pen, so that doesn't personally excite me. Being able to use a laser cutter without a computer seems like maybe a big deal, but my excitement for that is dampened by the requirement for a net connection. If you're actually seriously using it to make things, it's the kind of tool where you want it to be 100% reliable, and you don't want anything to change without you changing it. Neither of which is compatible with cloud computing.

I'm of the opinion that they'd be better off putting an extra few hundred bucks worth of computing power in there, increasing the price accordingly, and then make it easy to update the firmware and materials DB and whatever.
posted by aubilenon at 3:16 PM on September 24, 2015 [1 favorite]


Easier centralized updates, less computer power required on the device, the usual cloud reasons.

i.e., a pretty successful long con.
posted by brennen at 3:18 PM on September 24, 2015 [12 favorites]


For normal CNC type laser cutting, nobody would use this. It's not remotely an industrial tool.

This is for a maker/crafter who sells things at conventions/shows/expos/fairs/whatever.

And that guy who has a basement full of tools that he never uses, but they're all new and shiny and there's book full of project ideas that's going to be used *any minute* now.

The price point is just right for that, particularly since there's that whole bit about "buying materials from us". It's a razor blade...
posted by jefflowrey at 3:22 PM on September 24, 2015 [2 favorites]


It will be super useful for these guys.
posted by Chuckles at 3:24 PM on September 24, 2015


The cutting isn't nearly as interesting as the engraving possibilities. That's the real money application, imho.
posted by Thorzdad at 3:24 PM on September 24, 2015


That steady flame coming off the cut? Yeah that is bad.

You don't want that, it screws up your lens and is just this side of needing a fire extinguisher.

Otherwise I want one.
posted by poe at 3:40 PM on September 24, 2015


I've recently organised the purchase of and extensively refurbished a lasercutter for my local hackspace and, while this is a nice looking machine, it's really not as revolutionary as they're making out.

Using a camera pointed at the bed to superimpose your vector files on a live image of the materials is a nice trick, and you can do it with any USB webcam and Visicut, a great piece of Free software that works on linux, windows and (I'm told) osx. You can also give it materials presets, so the user just says "I'm cutting 3mm acrylic" and gets a sensible default. This software --like others-- will happily accept standard vector or raster files from the list of packages he read out (inkscape is free and easy, of course) to be placed or dragged onto your image of the materials, so it's just as easy to get your templates into.

The software plots the paths that the laser will take, which is trivial enough that quite intricate parametric designs take no appreciable time on a laptop that was midrange when I bought it in 2005. I'm sure his iPad could do it faster, certainly faster than by going via his company's servers. It then spits out a few kB file of machine code over the network, which our cheap, open source (LAOS) control board picks up and uses to drive the stepper motors and trigger a couple of pins to fire the laser.

I know I sound like I'm shitting all over his product, and I don't mean to. Good user interfaces and aesthetic design are hard, and often underrated by techy types. It does look like a nice machine. (Although, I agree, that steady flame is far from ideal.) But a lot of the stuff he's implying is new and exciting about this machine really isn't, and he must know that. I know marketers gonna market, but the dishonesty is irritating.

In fairness, having the software infer cuts and engraving from what it sees on the webcam is a nice trick that I haven't seen before. I'd be interested to see how it fares with balancing smoothing against faithfully capturing intricate details.
posted by metaBugs at 4:31 PM on September 24, 2015 [8 favorites]


Sadly I'll probably never spend $2000 on a toy.

I, on the other hand, would do terrible things for one of these.
posted by louche mustachio at 4:45 PM on September 24, 2015 [1 favorite]


I know I sound like I'm shitting all over his product, and I don't mean to.

No, this is MetaFilter, where people make their toast for breakfast with a blowtorch or ideally an open pit fire burning wood they cut down the previous season.
posted by GuyZero at 4:46 PM on September 24, 2015 [4 favorites]


Easier centralized updates, less computer power required on the device, the usual cloud reasons.

i.e., a pretty successful long con.


C'mon these guys have $9 million in funding, they would never go bankrupt and turn your cloud-dependent printer into a brick.
posted by benzenedream at 4:47 PM on September 24, 2015 [9 favorites]


They might go broke and leave your device a brick, but the point is that it's a plug and play appliance. It's pretty clear that it's aimed at people who want a low barrier to entry and consumers, maybe not even "prosumers."

Yes, the cloud model has drawbacks. The advantage is that you're supposed to be able to unbox this thing, plug it in and have it start laser engraving your Macbook cover in a few minutes. Plus it's relatively cheap.

Sheesh, it's not like you-better-wear-eye-protection Shenzhen laser cutters are going away with their crazy proprietary Windows-based device drivers that may or may not be backdoored. And you can always buy an Epilog Zing for $8,000.

The world has not somehow gotten smaller or worse with the release of this product.
posted by GuyZero at 4:55 PM on September 24, 2015 [3 favorites]


they would never go bankrupt and turn your cloud-dependent printer into a brick.

Backup plan: Run it with a Chumby.
posted by Bringer Tom at 4:56 PM on September 24, 2015 [7 favorites]


This is a 2D laser cutter, not a 3D laser printer.
posted by Ansible at 5:01 PM on September 24, 2015 [7 favorites]


Is there a lot of competition in the 2-3k USD range for laser cutters with similar working size and capabilities, though?

I realize that's preorder price but I wasn't aware the entry cost was less than about 5k for a 12x20" machine that didn't require a computer to program and operate.
posted by a halcyon day at 5:18 PM on September 24, 2015


I know I sound like I'm shitting all over his product

I similarly feel that while it does make some important and innovative steps forward, it's not quite far enough forward for me to buy. I think it's great though - these features will be improved and become more widespread. At these price points the software simplicity will help the market expand, funding more products and competition, and there will be more iterations offering a range of less-underpowered wattages, and one day... ONE DAY... all those stars shall align and the perfect laser cutter for me will exist, and I shall be parted from my money.
posted by anonymisc at 5:23 PM on September 24, 2015


Is there a lot of competition in the 2-3k USD range for laser cutters with similar working size and capabilities, though?

Yes and no. There are machines of this work-size and wattage and price range, but they are trickier to set up and use and trickier to maintain and have fewer Nice Things. They're not consumer-ready the way this device sounds like it tries to be.
posted by anonymisc at 5:30 PM on September 24, 2015


This discussion is oddly reminiscent of the iOS vs android debates.
posted by dhruva at 5:31 PM on September 24, 2015 [1 favorite]


Wait. There is absolutely no technical reason for this to require an internet connection.

Offloading processing to a computer is totally a reasonable thing, but there's no valid reason for it to be somebody else's computer.

tl;dr; Watch for subscription fees to suddenly appear in a year. Can't make printers without $50 ink cartridges.
posted by schmod at 6:08 PM on September 24, 2015 [8 favorites]


Nthing others here who are seeing through the alleged benefits of a cloud requirement.

For those not into the fab-hobbies, would you accept a regular old printer that would not work without an always-present internet connection to its manufacturer? That doesn't twig just a few concerns in your mind?

Because the 'benefits' would be the same.
posted by rokusan at 6:33 PM on September 24, 2015 [1 favorite]


less computer power required on the device,
AAHAHHHHAAAAHHA
My fucking amiga from 20 years ago has enough computing power to run this well.
posted by lalochezia at 6:58 PM on September 24, 2015 [8 favorites]


Yeah, I am kinda baffled by this. Aspects of it are neat, but I cannot for the life of me come up with any answer for what computations this could possibly be pushing off to "the cloud" that couldn't be accomplished by about $50 of commodity parts (e.g. something similar to a Raspberry Pi).
posted by tocts at 7:06 PM on September 24, 2015 [6 favorites]


After being on the receiving end of a fair amount of grumbling from Hacker News today over the mandatory internet connection, Glowforge has promised to open-source its firmware.
posted by teraflop at 8:34 PM on September 24, 2015 [6 favorites]


It was suggested to me yesterday that I may be in charge of my library's upcoming makerspace. Despite the legitimate complaints from folks who work with these kinds of tools all the time, this lowers the bar for the folks I expect to see coming into our space without any experience, to jump in and use. If, as the Glowforge folk suggest, that they'll release the firmware open source, full price, this would still be a decent piece of kit for our uses.
posted by evilDoug at 9:29 PM on September 24, 2015


If all the useful computing happens in the cloud, open-source firmware isn't necessarily all that useful.
posted by GuyZero at 9:54 PM on September 24, 2015 [1 favorite]


I guess it would make it easier to deduce the command set and build your own software.
posted by benzenedream at 10:25 PM on September 24, 2015 [1 favorite]


If all the useful computing happens in the cloud, open-source firmware isn't necessarily all that useful.


Exactly! However, releasing the firmware source code for a printer with an embedded controller that "doesn't have the computing power" to run it's own cutting operations is a flat-out irresistible challenge for a certain sort of old-school hacker. Somewhere, although they might not know it yet, some crusty old assembly programmer has just had their day made.
posted by Jon Mitchell at 10:32 PM on September 24, 2015


I suspect the important thing about the 'cloud' bit of this (aside from the cynical thought that some marketing guru thought that this word would make it more rather than less attractive) is the database of materials. Presumably there are important features of these that need to be known in order to configure cutting parameters in some way and having an online database makes it possible for them to maintain this information for you. OTOH the danger of having this change without warning seems like a bad thing, as pointed out upthread.

I can't really imagine why this device couldn't be controlled by a local (even onboard) computer that could use, but wasn't dependent on, a connection to this cloud resource though.
posted by merlynkline at 12:12 AM on September 25, 2015


Yes, the cloud model has drawbacks. The advantage is that you're supposed to be able to unbox this thing, plug it in and have it start laser engraving your Macbook cover in a few minutes.

Other cloud enthusiasts have already covered most of my points, let's just note for completeness I can't see the reason a device needs to connect to some remote server to be user-friendly and easy to use: if anything, one more step in the process is one more thing that can break, making it actually less user-friendly. Skipping the cloud though means the manufacturer would lose out on the ongoing relation with the customer, not to mention all that sweet, sweet metadata.
posted by Dr Dracator at 12:45 AM on September 25, 2015


My fucking amiga from 20 years ago has enough computing power to run this well.

Yea, aren't the systems driving most cutters and CNC machines often wimpier than even the cheapest smartphone, and often also old as fuck unless it's a brand new installation, or not cutting edge anyways but just what the manufacturer/installer was supporting at the time?

I've always got the feeling that this was something that not only could a potato run, but often did. It's basically a fancy plotter, and those worked fine in the 80s or even before.
posted by emptythought at 3:54 AM on September 25, 2015


In their defense, their "special sauce" appears to be an algorithm that translates a scanned image into CNC instructions (possibly with some feedback from the camera in cases where the design is traced directly onto the target).

Admittedly, this is *not* a trivial problem, and admittedly, I'd rather have a lot of processing power at my disposal for the initial implementation.

Still: No excuse for not providing binaries that their customers can run locally.

GuyZero: "If all the useful computing happens in the cloud, open-source firmware isn't necessarily all that useful."

Oh, good. So inexperienced hobbyists will be able to control the 50W laser, but not the useful parts of the machine.
posted by schmod at 6:41 AM on September 25, 2015


"Old man shouts at cloud computing"
posted by Ambient Echo at 7:57 AM on September 25, 2015 [3 favorites]


Still: No excuse for not providing binaries that their customers can run locally.

There are plenty of excuses for not shipping customers binaries, mostly that it's a goddamned nightmare. A company I once worked for that sold VM migration tools once took a database offline because of a bug in an IBM ethernet driver.

For a consumer product, end-user binaries mean windows software. Sure, it's not impossible, but it's also a huge pain and having the cutter talk directly to a server that the manufacturer owns and operates is much, much easier.

I'm all for owning what I buy as much as the next person, but this company is selling "cut things", not lasers. You don't buy this device to own a laser cutter. You buy this device to cut things.
posted by GuyZero at 8:09 AM on September 25, 2015


First, this isn't exactly something that a random person is just going to buy. Slick and packaged as it may be, it's still aimed at an audience (the "maker" crowd) that is going to have a lot more expertise in technology than the average person. We are, let's not forget, talking about a $2k (retail $4k) tablet-controlled networked laser cutter.

Second, the thing is already networked, and could talk to a tablet directly if they wanted it to. There's absolutely no consumer benefit to having the control go from the cutter, through the internet, to the manufacturer's data center, back through the internet, and to their tablet, versus going from the cutter to the tablet over the local network they're both on. Nothing about what they've described is so computationally expensive that it couldn't be handled by relatively cheap local hardware built right into the device.

This isn't "old man yells at cloud computing" -- I work in software, and there's definitely applications where it is appropriate. In this case, though, it seems like the only people benefiting from the always-connected nature is the manufacturer, as it functions as an internet-based dongle and recurring revenue stream.
posted by tocts at 8:27 AM on September 25, 2015 [1 favorite]


There's absolutely no consumer benefit to having the control go from the cutter, through the internet, to the manufacturer's data center, back through the internet, and to their tablet, versus going from the cutter to the tablet over the local network they're both on.

There are plenty of consumer benefits. When the manufacturers find a bug or inefficiency in their algorithms, everyone's cutter just gets better. There is no firmware upgrading, no tablets prompting users to install an update (and people like me generally won't let apps auto-update without oversight unless they're security related).

When someone uses a new material that isn't in the materials library, everyone's library gets bigger and more detailed (but the footprint doesn't)

And obviously there are lots of benefits for the manufacturer, which will likely translate into a better and cheaper experience for the consumer. Additionally, the added ease of offering after-market services often results in big-spenders in the optional after-market enabling thinner margins on the device as the company tries to grow the market, thus lowering the cost for everyone who has budget as a higher priority.
posted by anonymisc at 10:51 AM on September 25, 2015 [1 favorite]


There is no firmware upgrading, no tablets prompting users to install an update (and people like me generally won't let apps auto-update unless they're security related or the update addresses an issue I'm having).

So ... tying the device to a cloud service is good, because it lets the manufacturer force updates on you that you otherwise would have said 'no' to? I'm failing to see how the updates are somehow better or more acceptable when they're taken out of your hands, and doing network updates of an already networked device is trivial.

When someone uses a new material that isn't in the materials library, everyone's library gets bigger and more detailed (but the footprint doesn't)

Having a network-accessible library of added materials and patterns is not in any way related to tying the operation of the device to a cloud service. You can do the former without doing the latter, with zero negative impact on the consumer.

And obviously there are lots of benefits for the manufacturer, which will likely translate into a better and cheaper entry experience for the consumer...

Again, we're talking about a device that retails for $4,000 that could be made to operate independently (and not become a brick if the company goes out of business) for maybe $50 worth of parts -- and that's being generous.

I'll restate that I do honestly think this is neat, and there's cool stuff that can be done with it. However, nothing about what they've demo'd justifies a cloud requirement, and having that aspect diminishes the long-term value of it.
posted by tocts at 11:03 AM on September 25, 2015


So ... tying the device to a cloud service is good, because it lets the manufacturer force updates on you that you otherwise would have said 'no' to? I'm failing to see how the updates are somehow better or more acceptable when they're taken out of your hands, and doing network updates of an already networked device is trivial.

There's a big difference. Public devices are fragmented into millions of hardware and software configurations. I have a lot of experience with apps downloading an update to fix a bug on someone else's configuration that accidentally introduces a bug on my (previously working) configuration. This doesn't happen on the cloud - if it works on their hardware, then it works on their hardware, and that's the only configuration it needs to work on, and it works, and I don't need to do shit to get it to work, and I don't need to pay the cost of them having to spend hundreds of thousands to test everything on a fragmented environment every time they want to improve or fix anything.

not become a brick if the company goes out of business

That isn't a concern any more. The code will be open source. If they go out of business, you end up in the same boat as if you bought a regular laser cutter of that price bracket - having to run a patchwork of software and drivers.

You seem determined to handwave away any advantages of the cloud. I agree with you that it would be preferable to have the ability to have everything on your own hardware, and that there are advantages to having it on your own hardware which make that very desirable, but it's going WAY further than that to claim that a cloud approach offers literally nothing. It offers plenty, and recognizing that does not mean you're conceded that the cloud is better.
posted by anonymisc at 1:34 PM on September 25, 2015


How is this different from a programmable paper-cutter, like a Cricut?
posted by gregor-e at 1:36 PM on September 25, 2015


How is this different from a programmable paper-cutter, like a Cricut

Do you mean how different/similar is the business model? (Cricut is more a platform for selling fonts and clipart patterns in a proprietary format rather than letting people cut what they want), or how is the output different? (Laser cutters can cut wood, plastic, leather, etc, not just paper or the surface of leather, and it can engrave with heat, and it can cut some fine pieces because it doesn't exert force on the material, but it does char the edges of some materials, and smoke residue can build up inside the machine. Speaking of which, the Cricut is less likely to accidentally start a fire and burn down your house :) )

(Higher end laser cutters cut in a nitrogen atmosphere to enable aggressive cutting without charring or fire)
posted by anonymisc at 1:45 PM on September 25, 2015


The code will be open source.

Unless I've missed something, they've said the firmware for the device will be open source, but have said nothing about the cloud component. If they really are pushing important aspects of the functionality to the cloud, having the open source firmware with no implementation of the cloud is pretty much worthless if they go out of business.

You seem determined to handwave away any advantages of the cloud.

I'm really not handwaving here. I work in software, and in fact, on cloud-based applications in settings where it makes sense. This does not seem to be one of those settings. Making updates more mandatory is not even close to worth the trade-off being proposed here.
posted by tocts at 2:07 PM on September 25, 2015


Oh this is still open - I heard through the grapevine today that Chris DiBona, who heads open-source programs at Google and well-known FLOSS person, is an investor here so if it makes any difference to your opinion, the GlowForge people definitely have advisors on-hand to set them straight on open sourcing things.
posted by GuyZero at 2:57 PM on October 2, 2015


« Older Teacher required ... may need their own boat   |   A is for Autumn. Newer »


This thread has been archived and is closed to new comments