Hmm, love the concept but $19/month if you want to keep up to 3 of your designs 'private'. So that's $228 a year. And no PC layout yet. If you buy something like Proteus [1] for $1000, it has a $250/yr service agreement (which is to say keeps you at the current release, gives you support, etc) and it does PC layouts and you can have unlimited private projects.
So in perspective the product is interesting, a collaborative tool. The pricing model seems a bit broken from an actual use case standpoint. Perhaps they could go to a 'price per private design' kind of thing.
We based the pricing on our pilot users, and what they were willing to spend to be able to collaborate. We would love to chat if there is a different price you are willing to pay!
And for what its worth, we are working hard on PCB layout right now, and with any luck will be able to launch it in the next few months.
Pricing in the ECAD space is weird (you're probably aware) because it goes 'free' -> $1K -> $10K. with very few intermediate points. What is more, the number of people who have been burned by having something 'in the cloud' from a smallish supplier is growing. So having private designs that would be 'gone' if your servers were offline would be an issue. And lastly the biggest thing most people pay for these days is simulation anyway, and that is something easily charged for. You might consider 'free schematics' private or public, but $20 to generate a PCB from one. You could work out a scheme where if I publish a schematic using upverter and someone comes along they can 'buy' a PCB for it and you split the revenue share with the schematic creator. You can work a partnership with one of the PCB houses so that it 'just works' or you can send them a zip file of the various bits (gerbers, drill file, tool file, etc).
Something you might offer for 'free' would be something like Google's ChartServer API but 'schematic server' API. That is something I've been building in my spare time off and on. You are a lot closer than I am and I've yet to figure out if I want to use HTTP POST or a split URL syntax.
A way to embed the shared schematic in web pages or blogs, or whatever. Feel free to contact me offline for ideas on ways to monetize this kind of stuff.
Agree with you on the pricing, its a hard one. Lets chat sometime! Were totally on the same wavelength.
We are actually working on a transcriber (https://github.com/upverter/schematic-file-converter), dropbox integration. And behind the scenes were doing "print" button style manufacturing for our more enterprise users (ie. click print and we get a one-off manufactured and mailed to you)
Every upverter schematic has an embed code and a ton of our traffic is driven from people importing their schematics and sharing them on their blogs.
What's the possibility for making the site free, and paying for it by providing an integrated manufacturing or procurement service? (I notice your library already has "buy now" buttons next to components).
Various "buy now" buttons could deliver complete BOMs, circuit boards with BOMs, or turnkey products. A future move into mechanical CAD and graphics design could see a boxed product landing on the customer's doorstep.
Here's one. Can you get the software to recognise circuit topologies as they are being drawn, and suggest better topologies or components, based on what other circuits have been entered?
It strikes me that the difference between Upverter and any other EDA software is the collaborative/community aspect. Consequently Upverter's competitive advantage will arise from features that exploit the community aspect.
Thats a really cool idea. Similarily, we have been toying with the idea of auto-complete. Ie: You add a couple parts that we've seen connected before, you start to connect them in the same way, we recommend a way to hook them all up.
Also were working on the really simple things like just suggesting parts or implementations based other users designs.
Mostly handbaked into the app. As for payments, were processing in Canada using BeanStream. We open sourced our implementation here: https://github.com/upverter/beanstream
So, is there money in this? Doesn't it make more sense to go after the HDL synthesis design flow? I can see a few hobbyists and PCB makers interested in this. But the market for those interested in digital design but not wanting to fork out a fortune to synopsys/cadence is huge, right?
I suspect you can innovate quite a bit on the HDL front and then doing everything server-side should really give you a cost advantage as well. I can also see the open source aspect of this taking off - a high quality of library of plug and play cores that can be seamlessly integrated into your designs might prove to be a significant competitive advantage.
Very neat to see this! For the past year, I've been working on a very similar collaborative-design tool in an entirely different field (an aspect of transport planning). Probably won't have anything to show until the summer, but it's reassuring to see others pursuing similar business plans, interaction models, etc. Helps to validate my feeling that the cloud and the browser are fundamentally the future of design.
My own application provides relatively high-level collaborative design functionality, and does not attempt to replicate the functionality of existing tools that are capable of very low-level simulation, analysis, and engineering. I imagine that it's a similar situation for hardware design, although probably with fewer tiers of software -- presumably Upverter users are working at a level that's much closer to the end result of the hardware-production process. Still, there must be existing (offline) design platforms with low-level functionality that you don't want to replicate in its entirety; these are presumably embedded in the workflows of existing hardware design houses in a way that will be difficult to supplant. Have you considered eventually producing plugins for these 3rd-party platforms, which would keep their designs in sync with the cloud-hosted model? This way, high-level design could be done in the browser, and low-level simulation etc. can be done in existing applications, while still taking advantage of Upverter's collaboration functionality.
This, of course, is my own long-range plan, although the web application is very much the initial focus. Just wondering if you've had similar thoughts, and if so, whether you've found such a strategy to have any particular merits or flaws that you'd care to discuss?
The schematic is a small part of a full design flow. What hooks does it have to be integrated into a full design flow? Without such hooks it will be limited to rudimentary circuits.
What support is there for hardware description languages? HDL is good for large projects and is more reusable than a schematic.
One suggestion is to have a component library that is shared among all users. Each person adds components as they need them, and the library grows, as a community resource over time. Maybe you have that already? If you do have it, what license covers the library? Is it covered by the same open source license as the schematics of the users that generated the library? In that case, is it free for any user to export and reuse in full? Such a feature would negate some of the "vampire" accusations leveled by the dangerousprototypes review.
What's the advantage over gEDA, gHDL or similar? Most people with the knowledge to design a circuit would also have the knowledge to set up a tool chain and repository. Who's the market?
OK, I'll bite. Are there tool/design flows out in the wild using HDLs in the PCB design process? I guess describing basic connectivity is relatively easy to imagine, but it's less obvious how you would do things like PCB layout etc. Would love to be pointed towards examples.
(FPGA guy here - used to using VHDL/Verilog for digital logic simulation and synthesis, never considered it for board design)
Our HDL support is non-existant right now. We do have an open file format which can be created from or parsed to HDL. Do you work in HDL right now? I'd love to chat about this with you sometime - we come from the world of very large system design and there is almost no HDL usage at the system design level.
We do have a common library, all community entered and more than half edited by someone different than the creator. The library is free and public. We have a number of users of the library who havent yet adopted the rest of the upverter tool chain.
Don't want to discourage you guys, but coming from an EDA background myself, you didn't answer the question:
> What's the advantage over gEDA, gHDL or similar?
I think your answer is, "we're focused on design & community." But I wonder if you might reconsider yourselves the "Codecademy of hardware" instead?
While I agree that companies that rely on open source software should give back to the community (new projects, bug fixes to things they're using, etc.), we have to admit maintaining new open source projects costs time and money. When you're a startup trying to get a beta product off the ground, I suspect you'll prioritize having a better product first, and open sourcing things later.
(Note: I'm not in any way, shape or form related to Upverter)
Total bullshit. Open source isn't a religion it's a way for everyone to be more efficient. Scenario: upverter goes huge, makes lots of money and ends up patching dozens of open source projects as they go. What's the problem in that? They already have plans to support every major export format so people don't even need the code in case they go under. They are tackling such a hard problem with only two full time hackers, a designer, and a CEO they don't have time for an ivory tower, clean code every where, FOSS stack.
Seems to me that Upverter are solidly aiming to be a "github" in this space, after all their business model is nearly identical.
Github generally have a pretty good standing in the open source community. The reason seems to be the early open sourcing of grit and then, as their popular blog post says, "(almost) everything".[1]
Also that github don't enforce any lockin on your data (neither does Upverter from what I can see.[2])
If Upverter manage to adopt & project this same attitude (open sourcing every useful but not-core-business piece of software), I think they could turn this piece of bad publicity around. I suspect this would be to their immense benefit if they can court the open hardware community. Having people like Dangerous Prototypes supportive or neutral, rather than openly attacking them, would definitely help.
Not to mention that software for electronics nearly universally sucks (at least at FOSS & cheap levels.) Crying out for some quality products.
one thing though - since you guys are after hobbyists and small design firms and it seems like it's a fairly small market with very little money to be made.
but it's very likely that you guys will be acquired by one of the larger CAD tool companies.
hardware design is not like software design because plenty of companies are already sell hardware building blocks that can be tweaked with software to develop an application. Very rarely you have to redesign hardware from scratch. These companies provide their customers (which include hobbyists/small design firms) with schematics. Their customers may tweak the schematics rather than designing everything from scratch, so collaboration may not be the highest pain point here.
Wow, looks promising, this would have made team/design projects much more manageable if I knew about it in college.
I wonder though, does Upverter currently support low-level logic/CMOS circuit simulation? I know that's asking a lot out of a web app, but using Cadence for simulation is often overkill for most hobby projects, and I have yet to see a modern web app/startup that serves this niche.
I met one of the founders of Upverter this week and was pretty impressed by the polish of the product during his demo. I ended up signing up myself and I've already used it for a couple of my hobbyist electronics projects. I like the product and I'm looking forward to the road ahead with options to layout PCBs and have them printed.
if it isn't there, you might think about some way for people to advertise they're looking for a design review. It would be neat to put together a circuit and set a flag for hey, everybody (or anybody), come take a critical look at it please.
That seems very interesting, and shows the "GitHub" approach to collaboration can work on other realms other than code. I wonder if it could work for product design, too?
So in perspective the product is interesting, a collaborative tool. The pricing model seems a bit broken from an actual use case standpoint. Perhaps they could go to a 'price per private design' kind of thing.
[1] http://www.labcenter.com/index.cfm