New app-store-like website (not for everyone!)

Hi all. I did post something similar in the object repositories forum, so forgive me for double posting, but I'm asking for help here rather than just describing the site.

First of all, let me say that this will not be for everyone. I know there are many hardcore open source members here, and I respect that. That said, here's the deal:

www.grain3d.com is a website dedicated to bringing 3D printing to the average consumer. Average consumers don't know how to design in CAD, they don't know how to manipulate .stl files or generate gcode, and they don't know about temperatures, materials, or printing speed. However, I think that they are the future with this technology. Grain 3D Ltd. Co. takes CAD files from designers like you all, and optimizes them for the most common, easiest to use printers out there. We charge a fee for our services, and the designer gets 70% commission on every sale. See [www.grain3d.com] for the idea behind this.

I know alot of you will be against this, but I think it can live in harmony with open source sites like thingiverse (as an eample, most of you will never pay for a design, since you have the skills to make your own... not everyone can do this!).

Now onto the begging for help!

At this point in the site (Launched at the end of January), I need help. In particular, I need:
1) Designers (make money and accelerate the adoption of 3D printing!)!
2) People to verify gcode on their printers (grain only supports the Makerbot Rep1 and 3DSystems Cube right now, but I want to support more common and easy to use printers... maybe this isn't the best place for "easy to use" but it's worth a shot!)!
3) Consumers (doubt I'll find anyone to pay for gcode here, haha!)!

Take a look at the site, let me know what you think, and raise any questions you might have! Forgive me if I'm ever curt in my responses (I will try not to be!)!

Please contact me if you want to participate in this project in any capacity at all! I can't wait to hear from you!

Mike Mehlman
Founder, Grain 3D Ltd. Co.

This makes no sense, so you are wanting to sell gcode? This is logistically feasible. You would never be able to remotely create correct gcode for a customer printer, not even for a branded printer. The whole idea of it is DIY.

While I know there are people that are not particularly CAD or 3D modelling oriented, I really don't think anyone is going to buy or make a 3D printer to not know how to use it.

Are you personally experienced with FDM printing?

Partial list of variables:
filament: supplier(more every day), quality, color, variable working print temps
(even from the same supplier, filament produced on a Monday can vary from filament produced on a Friday)
temps: variable thermistor sensing from even the smallest resistance difference and installation position.
printing environment: humidity, ambient temperature, Alaska .vs Hawaii.

Edited 2 time(s). Last edit at 03/11/2013 11:24PM by Dirty Steve.

Hi Steve,

Thank you for the advice! I am aware of these issues. I think that in the near future printers will be *much* better than they are now; however, by that time it will be too late to get into the "bring this to the average consumer" game. Gotta start somewhere man!

I am only supporting the most standardized printers at first, and just doing the best I can. In a year or so I expect printers to be good enough to handle this business easily. Technology evolves at an exponential rate!

All the best,
Mike

the problem with a gcode output is that it is different for each printer, then there is the other problem of how do you select things like infil rates/type? what about layer heights?

i would be considering making the stl file the thing you download ,
also the 5 meg limit probably needs to be a little higher,

---

-=( blog )=- -=( thingiverse )=- -=( 3Dindustries )=- -=( Aluhotend - mostly metal hotend)=--=( Facebook )=-

Hello!

Good points! Unfortunately, the level of consumer I'm aiming at won't know what to do with an .stl. Also, I don't think they'll know what infill or layer height even means. I know this sounds ridiculously low-level, but I'm aiming at a one button kind of demographic. I actually think there are a few printers out there that might be able to pull this off now (for the most part at least), but I'm certain there will be a few in the coming months/years. Good point about the file size though. I was trying to limit my bandwidth at first, but that might have to change!

Thank you!
Mike

I think that by the time there is a push button solution for home 3d printing, you may very likely be just loading the model file onto the printer, the printer asks the most simple of variables such as size, material, and number of copies, then the printer just does its thing.

It is VERY unlikely to be a push button solution any time soon.

Hi Dirty Steve,

Yes, there is definitely alot of work to do (as you all know!), but I'm giving it a shot! My Makerbot Rep1 was **more or less** push button (when it was working in general), and the next printer we will support is the 3DSystems Cube, which promises to be even more hands-off. I also agree that input standardization will take place, but we're in the sweet spot here when the tech is advancing but the standardization has not occurred yet. This is the perfect time, in my opinion, to get a foot hold in the market (evolution will be necessary in the future regardless!)!

I really appreciate your thought!

Best,
Mike

Standards for this is not as easy. There are different size machines, printing bed and heights, nozzle sizes, layer size limits, and quality needs that are different for each person and machine. What if someone wants a more solid object. What if someone wants a 0.4mm and someone else wants a 0.1mm print of the same thing. Thats why gcode is not used for the sharing of files stl is. Its already the standard. The software is not hard to learn.

Not to mention if you are only supporting Makerbot and 3DS at the moment, what do you expect to gain from REPRAP forums? Not many of us here work with overpriced junk.

Edited 1 time(s). Last edit at 03/12/2013 02:00AM by xclusive585.

... such 'services' could be comparable to music download sites, where the 'value' will be searching/collecting/creating 3D-objects and optimizing them for dedicated printers or printing methodes.

Thingiverse is doing something similar on 3D-format basis -- actually for free ... but could change any time ...

---

Viktor
--------
Aufruf zum Projekt "Müll-freie Meere" - [reprap.org] -- Deutsche Facebook-Gruppe - [www.facebook.com]

Call for the project "garbage-free seas" - [reprap.org]

Good morning all!

xclusive: thanks for your thoughts! This is not the only community I'm reaching out to, I just thought some people in here might have designs or thoughts to contribute, but I've been involved with several other forums and local media as well. Also, there were several posts about Makerbot equipment in this forum, so I wanted to give it a shot!

lotw: to be brief, I don't want the user to deal with **any** settings. They're going to get whatever object is most usable and most printable. I know that wouldn't fly here, but beggars can't be choosers!

Viktor: yes, I don't see thingiverse as a direct competitor right now since they appeal to a slightly more DIY (and open source leaning) crowd, but you are totally right, this could change at any moment. Gotta strike while the iron is hot, as they say!

Best,
Mike

there is no logistics solution for this. again, NO ONE is going to have a 3d printer and not know how to use it.......

Would you buy a car, hire a blind folded driver, and expect to get to where you want to go?

Hey Dirty Steve!

Thanks so much for your concern! I'm betting otherwise, but we shall see! Again, this is just my opinion!

May the force be with you,
Mike

#facepalm.....

My (inexperienced) 2p's worth :-

Maybe down the line it would have some mileage so I think you're right to get in the frame early. At some point there may be a few varieties of printers with more mass produced repetitiveness (Solidoodle jumps to mind) and branded or model-specific filament (Cube3D?). This may not be the best place to ask the questions but it would definately be silly not to ask here!

I agree that as it stands, people who can't operate a Reprap type machine would not buy one. But if there was a one click or close to that solution, then potentially they could...

I wonder how quick the software will develop generally? I assume that it will develop to the point where it knows what temperature, speed etc to run at depending on the model required. I know it's a minefield of settings at the moment but it can only improve over time?

3ddude,

Thanks for the positive attitude! I agree, it will be very interesting to see the development of the software. At one point printing on a piece of paper was a similar "chore" but today we all just hit print. Yes, 3D is a whole extra dimension, but I doubt that's going to stop the industry!

I also expect very few people to purchase products this early on. The goal for right now is to build up the designer base in anticipation of the consumers. Let me know if you have any advice or criticism!

Thanks!
Mike Mehlman

mehlmanmichael Wrote:
-------------------------------------------------------
> Good points! Unfortunately, the level of consumer
> I'm aiming at won't know what to do with an .stl.

Yes they will. They'll know that they can drop it into their printer (on a memory card or drag and drop onto the printer's wireless web interface), press "Print!" and the object pops out a while later. That is, if the 3D printers ever become common household items, and if it's any more complicated than that, they won't. And if they do, the printers will surely not accept raw G-code in, because there is no guarantee of what would happen if that were to be run. The user would most likely have to do something equivalent of iPhone jailbreaking to get raw G-code access. Voiding the warranty and taking the risk of physically breaking the printer if there's a bug or incompatibility in the G-code. Good luck convincing people to do that.

ttsalo,

I think you're right here. Soon I think 3D printers will be standardized to accept the same file types (how the industry accomplishes this will be interesting to see!). My goal is to be a product web store, not just a gcode web store. If you could just print .stl's right now, Thingiverse would be a much more significant competitor, but, since you can't at this time, this gives grain3d a better chance of gaining market traction with the non-existing "average consumer". I'm not married to selling gcode, but have chosen this path as the most direct way to get involved in the market.

Best,
Mike

the reality of a 3d printer in every home will happen when you can say 'Computer, Earl Grey Tea, hot.'

It doesn't seem that you are very experienced in 3D printing.

If the customer doesn't know ANYTHING, how would you ever be able to remotely troubleshoot a bad print when the customer can't tell you any thing more than it didn't print right? How may times are you going to rip g-code for one person? How many times do you think a customer is going to try to print an object?

Hi Dirty Steve!

I caught your reference there!

I think the customer will try to print something once, or *maybe* twice. That's why we have to get it right the first time (bring on the comments about how 3D printers are environment specific, filament specific, etc.!). Right now, getting it right the first time entails using the most standardized printers and only printing simple enough objects. In a year or two, I think all modern printers will be able to print right the first time. Again, only my opinion, let the record state it!

Back to my order of business: does anyone know a better forum to look for testers? There aren't many Cube forums, and the Makerbot ones are slightly hostile too.

Anyone else up for trying out a design for the site?!

Finally, does anyone else know of other forums where I might have success enlisting help or getting advice?

Thank you Dirty Steve et al.!

What I have been wondering about is how you would address that your customers might re-upload the GCode files elsewhere, in effect enabling others to completely circumvent your shop? The printers won't be able to read custom file types with DRM (though I can imagine Makerbot or 3D Systems being willing to adopt just that).

Other than that, distributing printable files via GCode is a really bad idea, that we pretty much agree with here. Even if you had customers who own printers of the same make, calibration might look completely different. As Dirty Steve already mentioned, filament supply variations play a big role in printability and print quality. I have heard people reporting that PLA from one supplier extrudes at way lower temperatures than the one from another supplier for example. And that isn't even covering that you can print with multiple materials. You have to consider all that while generating GCode.
I would suggest that customers get a minimal interface from which they can choose their filament and other basic parameters (with the option to save that profile for future use), so that they can get tailor-made GCode files on demand, delivered freshly sliced from your servers. After the initial parameters are all set up, CTRL+P simplicity might be possible, but you will also have to calibrate your 2D printer after switching spent cartridges out and occasionally fiddle with some settings for a few print jobs.

Hey uGen,

I like the idea of a few savable online settings. I don't think I can pull that one off right away, but maybe down the line (though by that point maybe printers are more standardized).

Regarding the DRM issue: I'm not worried about people redistributing the gcode. It's just like MP3s. People can if they want to, but if you make it cheap enough and if they have any desire to support designers, at least some fraction won't "pirate".

Version 2 uGen, version 2! Those are some good ideas.

Mike

... I'll say: "try this!" -- with G-code you'll only address customers with exact/specific sets, that meets your basic settings ... this is common for CNC-data, when programming NC-jobs for specific machines.

If you want a broader acceptance, then you'll need a more exchangeable data format - take a look in the gaming- and CNC-"worlds", what's common there for spread/interchanging 3D-content.

The STL-format was created to fit the smallest possible set of defines, needed for the first generation of 3D-printeres ("STL" means "STereo-Lithography") and is only a subset of the original 3D-object, representing a coarse assumption of the surface ... not referencing on material, colour (but this is possible with modified subsets), subsets/assembly or parametrics ...

---

Viktor
--------
Aufruf zum Projekt "Müll-freie Meere" - [reprap.org] -- Deutsche Facebook-Gruppe - [www.facebook.com]

Call for the project "garbage-free seas" - [reprap.org]

I see; this DRM-free approach is better than artificially limiting the usefulness of a file. If you offer enough incentive for customers to come (back), I can see this working on a trust and convenience basis.

Actually, it shouldn't be too difficult to save customer specific data. Popular slicers like Skeinforge and Slic3r also run in text console mode where you can call the program with specific parameters.
All you have to do is having a slicer installed on the server and saving some basic settings that apply to all printers in a text file (maybe add a second line of files for different printer models). The user settings are being saved into another file. When the customer orders a 3D file, the text files get read out and the output piped to the call parameters for the slicer.

Some people I know are already working on a remote slicing service similar to this because if we are to use Raspberry Pi's as host computers for printers, slicing would take ages to finish, unless we outsource the processor intensive process to a more powerful server.

Yes, I think it's a race between standardization and adoption by "average consumers". If standardization occurs (maybe built in slicing in printer hardware or something) then there's no need for remote slicing, but if consumers start to adopt the tech before standardization is achieved, remote slicing could be a huge deal.

uGen, I would be very interested to speak with the people you know who are working on remote slicing, if only to get their opinion on the subject.

Thanks!
Mike

I think you got this a little bit backwards. Because there is absolutely no standardization, built-in slicing makes more sense since your printer knows / can store its configuration and calibration. The remote slicing idea was more or less a compromise between distributing the final, printable GCode, thus abstracting the process from the user's point of view, and being able to cater to the biggest set of configurations possible.
However, slicing takes a lot of processing power / time, so it wouldn't be too economical to equip every printer with the necessary hardware in order to make slicing swift and comfortable.

The one responsible for the bulk of the code is actually member of this forum: Just search for dronus; you will probably find his latest contribution in the developers forum in the Vis Printer thread.

What came to my mind just now would be to copy Cura's approach and write a dumbed-down interface that sits atop a slicer preset with basic parameters. The user would be able to download standard 3D files like STL's or OBJ's and load them into the program, just like it is now, but with less options and more explanations.
Already, you can call the slicer of your choice with the settings you like from Pronterface, so practically, you just have to modify this piece of software a little bit, bundle it with a slicer (also already done for the precompiled Windows/Mac versions of Pronterface) and offer it as a free download.
The disadvantage would be that people with weaker computers will have to wait longer for their files to be sliced, though.

... or offer a cloud-service, where the customer have to set all the properties of his printer, material and/or support options ... and last the server calculates all slices, paths and presets and returns a ready-to-print file

---

Viktor
--------
Aufruf zum Projekt "Müll-freie Meere" - [reprap.org] -- Deutsche Facebook-Gruppe - [www.facebook.com]

Call for the project "garbage-free seas" - [reprap.org]

mehlmanmichael Wrote:
-------------------------------------------------------

>
> Finally, does anyone else know of other forums
> where I might have success enlisting help or
> getting advice?
>
> Thank you Dirty Steve et al.!


Maybe Solidoodle forum? They're going for the cheaper market with as much hype as Makerbot...