personally I dont trust this consortium to actually improve anything but to introduce DRM into one time download files. that may or may not get a complete print.
Why, in that announcement, which mentions several file formats having limitations is the AMF format not mentioned? You know, the one that supports colour and probably already ticks the boxes that the 3MF format is hoping to fill.
Yeah… not real obvious how this is going to be different from AMF, which is already an ISO standard…
I’m still hoping for a file format that can contain perfect curves instead of being limited to polygons.
…and I fear a file format with DRM.
@Ryan_Carlyle problem with amf is that it is not an open standard. And that no 3D design apps are adopting it. 3mf has some important players in the mix. Other then that the base spec is not that differnt from amf.
I do not know of any DRM talk in the 3mf format.
@Erik_de_Bruijn and @Jaime_van_Kessel knows a lot more on this then i do.
@Daid_Braam Not an open standard, in the sense you have to pay ISO $90 for a copy of the spec? That’s a pretty silly reason to make a new standard. It’s frickin’ ISO. If the ISO standard isn’t good, the proper approach is to work with the committee and change it, not release your own competing standard. Maybe I’m missing something…
Autodesk was a big advocate for AMF a few years ago. SolidWorks, Blender, Cimatron, Rhino, and a bunch more all export AMF. Nobody was adopting it in 2011 when it was new, but that’s not true any more.
When Microsoft announced 3MF, we all figured it was so they could control the file ecosystem, like they have historically done with their other file format standards. The only two things people have mentioned in 3MF that AMF doesn’t already do are print settings (which are only useful for large-volume printer manufacturers with consistent hardware configuration) and digital signatures (for DRM).
It was not surprising when SSYS, 3DS, etc signed on to 3MF. It was VERY surprising when Ultimaker signed on. I can’t find any good explanation why the open 3DP community should back this. Maybe it’s legitimately better than AMF, I don’t know – nobody has publicly explained why, as far as I can find. I’d love to hear some good reasons aside from people hopping on Microsoft’s closed-source-friendly bandwagon.
@Ryan_Carlyle it’s one of the things that I found annoying of it, as an Open Source developer having a standard behind a paywall is never a good thing. My suggestions towards the AMF standard committee where simply met with a “no”.
Is 3MF better then AMF? I don’t know. Is betamax better then VHS? People say so. Still everyone bough VHS. Why? Because they could get content in VHS.
In the end, we, Ultimaker, are a content consumer, we consume 3D models. So for us, it only makes sense to read as many file types as possible. Which file type will prevail in the end? No clue. But also, no care.
Digital signatures have nothing to do with DRM. Is HTTPS about DRM? Digital signatures are to detect tampering. Not to prevent copying. The 3MF standard even specifies that if you have a file without a signature you cannot know and should never assume that it was stripped of a signature.
Why did we sign up? Well, so we can influence it. So we can make sure DRM does not end up in the base spec. So we can influence what is possible, useful and importantly, says open and free to implement for everyone.
I am strongly opposed to DRM and i’m fairly sure that Erik shares this opinion.
Also; All of the 3mf implementations are open source. Hell, they even provide them on their git hub. So how does anyone think this is a closed source thing?
The main difference between AMF and 3mf in my opinion is that there is a ‘base spec’ and a set of optional specs which can be used to expand on the base. This allows for a lot more modularity, which should make it more future proof.
@Jaime_van_Kessel That’s OK as long as any not-implemented parts of a specification don’t become a limitation to users.
I could also see an opportunity for products to include optional parts of the specification as paid upgrade components.
In that respect, fully implemented specifications are a better option for the end-user.
I can find nothing that addresses what it fixes, nothing that describes in any useful way what is broken or limiting in any of the formats. What does it embrace? What does it extend? What does it abandon?
@Nathan_Walkner The Reprap/DIY movement has been losing market share for some time now among general consumers. They want to buy, not build. It’s got nothing to do with any sort of conspiracy to to limit the DIY movement. It’s a natural progression of the tool.
The consumer 3D printer bubble has burst, there will be no mass consumer market for the machines. There is, however, considerable movement in education, small shop prototyping and small batch manufacturing.
There is still plenty of market for the DIY set but the future of the technology is in the ready to use segment. These days most people want to use them, not build them.
@Drakester It’s an engineering spec not something for direct consumer consumption. It’s not a product in and of itself. The end user is oblivious to what is happening behind the scenes. Much in the same way a desktop computer user generally doesn’t know what type of file system is used on their machines or if you are using TCP or UDP for a protocol.
If you build and/or implement machines used in commercial or production environments there is much benefit. By using a common schema there is an opportunity for portability across machine platforms. It standardizes the formatting in a similar way to how TCP standardized network communication.
As with Autodesk Spark, you need to be a developer to be able to use and implement it. It’s not a product, it’s used to make products.
EDIT: Here is a use case from trincle 3D http://3mf.io/why-we-implemented-3mf-for-the-trinckle-3d-customizer-and-you-should-too/
@dstevens_lv Thanks for finding that. Kind of reads like a marketing testimonial though. And no mention of AMF, which shares most of the same advantages.
@Ryan_Carlyle It’s a use case of an application that is in production. They’ve got a customer facing product and a software as a service product for B2B. I’d reckon Shapeways is also using it at least in a beta capacity.
The AMF standard doc license is straight ahead, old school copyright. You can’t freely redistribute the document according to the EULA on the ASTM site. Though you can freely use the specs. There is also no framework, open or otherwise. AFAIK the current AMF spec is a schema definition/standard only with the dev needing to roll their own functionality.
3MF provides an open framework in addition to the spec. The API of the lib is well documented. The license allows free distribution and modification as long as the source and the copyright are included. As a dev it’s going to be a much shorter time to market and the open source economies of scale will be more advantageous when using the same lib or versions of the same lib will help shorten time to market.
This is something that happens on the backend that users will never see. I’d venture to say, at least for the time being most regular users are going to be slicing stls and I don’t see that going away.
@dstevens_lv
It’s absolutely normal and routine for standard documents to be copyrighted and have small fees for copies. Almost all standards are like that – it helps sustain the standardization bodies (like ASTM and ISO) that generate such utterly vast economic value for society. It’s not a burdensome cost. As you note, hardly anybody actually needs a copy. If you’re seriously developing a slicer or modeling software, $90 to buy a copy of a standard is the least of your development expenses.
Now, providing a free library to use the standard is very nice. (Assuming they follow through on making it platform-independent… it currently requires more than $90 in Microsoft products to use the code…) Microsoft IS doing the right things to get 3MF adopted. BUT they could just as easily have developed equivalent open source tools for AMF. Seriously, how does Microsoft benefit here? What 3MF offers that AMF lacks seems to primarily be features designed for DRM, closed platforms, and big industrial printers.
Anybody who says “the model file will include print settings and/or digital signatures” is probably running a closed-source, chipped filament, locked-firmware, pay-to-print class of business model. There’s simply no way that stuff benefits hobbyists. At best, Microsoft included that stuff so the “bad guys” would sign on to an open standard. At worst, it’s a route to ugly pay-to-print DRM schemes baked into the operating system.
I’m not anti-3MF. I’m anti-more-standards… because they tend to ironically increase fragmentation.
In the end, it really doesn’t matter what format we standardize on, 3MF seems fine, but the motives of many of the companies backing this are questionable.
An ME prof was keen to tell us (more than once…) that the good thing about standards was there were so many of them.
It builds under both the community version of Visual Studio and gcc. No cost for the Windows tools. There are some functions that only work under Windows now but that’s more because NetFabb and MS mostly make Win software. Personally I’d like to see more OS X dev in all areas of the maker movement including this one. There is a call out to devs and users of all platforms.
DRM is not included in and of itself in the spec for 3MF. There is a signing mechanism (which is not DRM) but any DRM would need to be handled by OPC which isn’t specific to 3MF. You could just as easily implement DRM in any OPC container including one that contained AMF formatted data or any XML data for that matter.
@Nathan_Walkner It is a sad fact that much open source software is authored under Windows using Windows-specific tools.
However, as long as the resulting code can be built using open source tools without any penalty in terms of lesser performance or features then we must take that as acceptable.
@Nathan_Walkner Devs have to code for what people use, not what the devs would like them to use. You can sequester in your bunker and rage at the machine but in the end the reality is that most people still use some form of Windows.
The lib builds in a test app under gcc. If you want to make printing under Linux better put a project together and go for it and port over a full *nix version. It’s a BSD license with framework and API docs.