Editorial: Upverter, another closed source vampire exploits open hardware for ventrue capital, PR, and profit

in Editorial, open source by Ian | 28 comments

Upverter is an online schematic layout program. Among several features, it allows a group to work on a project collaboratively. They were a big presence at the Open Hardware Summit, and have been marketing themselves as the ideal solution for open hardware collaboration. That sucks, here’s why:

This is a new class of project we’re calling closed source vampires. These outfits jump on the open source/open hardware bandwagon for venture capital, PR, and profit, but they don’t participate equally. They exploit work from the open source community for profit, but keep their special sauce to themselves.

In this case, Upverter is free to use for open hardware, but their source code is totally closed. They want to use the hard work of open source developers to build buzz for their company, but keep their own stuff closed. This is at best a side step from free-but-not-open Eagle, but with a dash of user exploitation.

If this is a platform for open source, why not make it open source too? We asked the same question of CircuitBee, a free-but-not-open online viewer for PCB and schematic files, and several similar offenders. The responses are usually similar.

“We worked hard on it and don’t want someone to seal our work”

Yeah, all the open source developers you want to use to promote your platform worked pretty hard on their stuff too. If you think reuse of your work is akin to theft, then you don’t understand open source – stop using it to market yourself.

“We need to profit to support our software as service. Servers aren’t free ya know!”

Sure, but open source in no way precludes you from making a profit. In fact, you might actually get a lot of value back from the community.

“Nobody wants to run a software as service cloud themselves anyways, so there’s no reason to open source it”

Irrelevant. Even if 99% of people never crack open the source, like most OS hardware or software projects, that’s not the point. Everyone still benefits.  The potential to make more and better open source in the future benefits the community you’re trying to exploit.

As always, we couldn’t care less if a company trades in closed source. Most do! We have no problem using free tools until an open alternative is available. But if your PR troll bangs a press release through our contact form touting your wholesome open source goodness, you had better be sure your house is in order.

This entry was posted in Editorial, open source and tagged , , , , .

Comments

  1. Denis says:

    The other big issue is that a web-based design tool doesn’t work when you’re not online and may not be available in the future if they decide their server costs aren’t worth paying. At least with Eagle, Protel, Orcad, etc, once you’ve installed the software, you have the ability to work whenever you like.

  2. logical says:

    The purpose of the web based tools is to increase collaboration and improve discussion. If I was to ask you how to build a full wave rectifier or a window comparator or an astable 555 timer circuit for example you would have to draw that up or find an image on the web because you can’t describe it really without a schematic. Open source tools like kicad are improving but they don’t provide very good collaboration ability. You can use a version control system like git but the best description of the project is a schematic. Also ,I’m biased because I’m building my own online schematic editor.

  3. Frank says:

    This is too political, I’m just going to let economics take care of this. Everything has a consequence and we’ll see how they do after a few months.

    KiCad is not streamlined enough for me to switch from Eagle, plus it keeps centering my mouse cursor when I zoom in/out using the mouse wheel, which is annoying. Open source does not guarantee perfection. Sometimes a strong economic incentive is needed to keep up quality (which is still possible with open source).

  4. Word.

    I’m looking forward to Eagle’s new XML-based file formats. Once those are in place, it should be easy enough to use them with Github. Do KiCAD & gEDA use text-based file formats?

    • Well at least kicad does.
      It looks a bit like … hmm … lets say G-code.
      Snippet from a PBCnew-file:
      $PAD
      Sh “1″ C 984 984 0 0 1800
      Dr 197 0 0
      At STD N 00E0FFFF
      Ne 7 “N-000007″
      Po -2000 0
      $EndPAD
      $PAD
      It’s as “human-readable” (note the “”) as xml to me.

      I’ve been using git and hg to keep the work on my laptop in sync with the stuff on my computer at home.
      I even produced a few merges, working on one end of the pcb at home and on an other end of the pcb with the laptop at the university.
      Some of the worked with gitmagic out of the box, some of the didn’t work I need to fix stuff with a texteditor.
      That’s not entirely kicads fault, I’ve been working on a large website project using git and our css file got messed up an most of the merges. So it’s git issue as well I suppose.
      I don’t know how git handles merges, just plain diff and patch or if there is more magic involved.

      I’d suggest to try it out yourself and tell us what you think.

      Maybe we should gather some people to hack git support into kicad.
      As far as I know there are rarely free schematic editors around that have collaboration features.

      Sebastian

  5. Zizzle says:

    Interesting.

    I’ve always found it strange to call something Open Hardware when most designs need you to install a big hunk of proprietary payware to modify them.

    I’d much prefer the online solutions to that – the barrier to entry is lower, and no more proprietary than Eagle.

    After my current project is finished I’m considering spending some time hacking KiCAD to add some more collaboration features. Specifically git backed part libraries which makes them easy to share.

    I’d love to see one of the open source CAD packages become the defacto for the Open Hardware movement.

  6. jone says:

    While I mostly agree with the rant ;) I also sort of hope it was just brought on by the OTT marketing blurb.

    IMHO there are some great closed source web services that really benefit open source projects (i.e. github). I have no problem using these, although of course there is often a choice to be completely open source if you prefer (i.e. gitorious). I’d like to think a schematics collaboration site that’s free to use for open source projects would fit in the same model (marketing nonsense aside).

    @Pierce: Yes, KiCAD and gEDA both have documented text-based file formats.

    And speaking of marketing nonsense, I did find the recent CadSoft announcement a little disingenuous with their “This is the first CAD database format to use ASCII text as a format” claim. I see their site has now clarified that statement somewhat ;)

  7. kuhltwo says:

    Open Source is Open Source. Several good points made.

    If they want to tout themselves as open source, then they need to be totally open about their products as well.

    Just my 2 cents worth.

  8. @jone: I don’t think Upverter and Github are at all the same thing. Upverter is a whole product that does not seem to be intended to exist outside of their servers. If Upverter goes away, so does all your data and your preferred method of interacting with that data.

    In contrast, Github is a site that provides a location to store Git repositories and some social sugar to make it a stickier value proposition. Git is independent of Github; if Github vanished tomorrow I’d just have to set up a Git server of my own rather than just paying them a bit of money each month to do it.

    • jone says:

      True, that’s a very good point. Must admit I was thinking more of the CircuitBee approach, where you upload existing (KiCAD) schematics to their web service. In which case it becomes purely a collaboration tool, and you still have all your original data if it ever disappears or becomes non-free…

  9. Zak Homuth says:

    Hey Guys,
    Its Zak from Upverter. Ian forwarded me the article when he published it and asked for me to comment on a couple questions. Understanding that this is intentionally controversial and sided I still thought it would be worth sharing where we stood on the conversation.

    For starters check out our promises http://upverter.com/ourpromise/
    Or our many blog posts on open hardware http://blog.upverter.com/

    My answers to Ian’s questions follow:

    >>> Why should open source use and rely on a closed source platform for development?

    I think this is a question of motivation, and I think you are very right to ask it. If I aspired to be a visionary or thought leader in open source hardware I too would want to make sure I wasn’t building on sand. But to answer your question: it doesn’t have to. But that is also not our motivation. We aren’t trying to trap anyone or disrupt the movement, we just have a product that can make the lives of the doers of open source hardware a little bit better.

    For what its worth I respectfully disagree with the requirement for massive vertical openness. I think its hampering the movement. How many doers of open source hardware do so on fully open source computers? in fully open source operating systems? Using fully open source software? And designing with fully open source components. I think at some point we should just start doing instead debating infrastructure, and I think Upverter can help with that.

    >>> Isn’t that a step backwards?

    At the very worst its a step to the side like you say in the article. But that also begs a lot of other questions. Was SourceForge bad for open source software? What about github?

    >>> Why should I help promote your closed platform with my open source work?

    I dont think anyone is asking you to do promotion. Do you like your tools? Do you collaborate or share what you do right now? Do you find that easy? You should use Upverter if it makes your workflow easier.

    >>> Why aren’t you guys willing to give anything back to the community you want to exploit for PR and profit?

    I respectfully disagree. This is a pretty loaded question, but in response I’m curious if you have looked at the site? We have open sourced every hardware design we have done since starting Upverter. We actively pay a number of hardware engineers to design building blocks of electronics – all of which we open source.

    • Zizzle says:

      >For what its worth I respectfully disagree with the requirement for massive vertical openness. I think its hampering the movement. How many doers of open source hardware do so on fully open source computers? in fully open source operating systems? Using fully open source software? And designing with fully open source components.

      The following is not so much directed at you, but to the “Open Hardware” guys in the audience.

      It seems strange that they support and encourage the use of a bunch of proprietary tools (Microchip, Windows, Eagle come to mind) but have taken exception with yours for some reason.

      If openness is not very important then why bother with open hardware at all?

      If creating/modifying a design requires a $10K per seat license how open is it? Certainly not to the casual hobbyist. But what exactly is the price cut off? And is it the same for me as an affluent westerner as a child in India? Should only rich first world types be able to participate in “Open Hardware”.

      If you care about open hardware then why not open software?

      FWIW, a complete open source stack exists and is very capable and pretty easy to use. And can be improved and built upon. A bit of effort improving KiCAD which is free to everyone benefits “Open Hardware” much more than another proprietary tool designed to lock you in to make a buck which will supposedly be reinvested into the tool. Or another blinky light “Open Hardware” design in Eagle.

      Unfortunately network effects are important. If everyone else is using proprietary payware for “Open Hardware” it makes it hard to participate or collaborate using the truely open tools.

      • Ian says:

        For what it’s worth, we don’t use any tools that anyone can’t get for free. I know Upverter’s PR mentions the 10k price point, but few in the larger oh communtiy use altium, etc. I’d ask you to name one that does.

        Beyond that, the reason that I have no problem with github/source forge/ etc is because they didn’t slam a press release through my business AND personal contact form seeking to use whatever miniscule influence this blog has to promote their exploitation of the open source community. If they ever do, you can count on a similar reply.

        I’ll have a more complete reply tomorrow during the work day.

    • EDA Ninja says:

      It’s strange I missed this one but equally timely. I see Zak’s comments and I wanted to explore just what – if anything – has massively improved about Upverter since this writing almost 2 years ago. The truth is, Upverter has failed to solve almost any of the very *hard* problems any EDA developer faces (I know having made EDA my biz for most of my professional life…and I’m old enough now to have fathered kids like Zak).

      Sure, they have built a web tool. Sure they have online web chat support. Probably a JS / Node / Mongo fueled back-end of No-SQL in the finest of modern forms. Sprinkle in some Ruby and it must be good…right? NO.

      Sure they have added online ‘collaboration and sharing’. Fancy roll-back features and persistent undo stacks. Sure you can share blobs of data. Sure you can ask others to look at your stuff. But let’s face the facts…These are NOT the hard problems of building a schematic and PCB tool. They are the problems of a couple of developers trying to outpace pissed off investors who are looking for progress…”Look boss…I managed to integrate with another well-known and easy REST API to do something that anyone with a 4 years degree in CS could bang out”.

      Schematic in upverter is not good. Profoundly bad. Painfully so. Wiring is bad, hierarchy stinks, multi-channel support doesn’t exist. Bundling of signals / busses / grouping of connections doesn’t happen. Schematic constraints management is painful, rules stink, rules-to-PCB suck. Block design and design reuse is weak, port connectivity, labeling, net classes, component classes, document classes? Where are these? How can I attribute rules to these? How can I constrain groups of DDR tracks to impedance controlled, length-tuned, differential pairs? From schematic?

      Not so common? How about every bloody computer and mobile device you have in your life and becoming more common as the days go by.

      And sure, some of this is complex, other parts of this are bloody fundamental. And without these features, you’re nothing but a pretty face.

      And what about PCB? “What about it?” you ask? How about solving important problems like polygon management, pouring around objects, clipping? connections by net, by pad, by via to pours? Split planes? Overlapping split planes? Net shorting? How about multi-track routing? Differential pairs? How about dragging corners of multiple tracks? Selecting and slicing tracks? How about multiple tracks with rounded corners? How about length constrained routing? How about impedance control?

      And libraries? Where is my library I can generate from CSV files? My link to *my own* back-end company database? Where is my integration with my supply chain management tools? My connection to the old lady in the purchasing department that manages our parts purchases?

      See, you’ve spent your energy on things like checking boxes and confirming a design is ready for production…But this is NOT the hard problem. If this is what you want to build, then stop calling this a design tool and go pretend to be Windchill or Teamcenter.

      Translators? Why do you NOT export to the same formats you import from? If you intend the system to be open? Why not try converting the really hard formats? DxDesigner? Pads? OrCAD? Engineering professors have been using OrCAD / PSPICE since you guys were a glimmer in your daddy’s eyes and yet you haven’t managed to crack it?

      See, my frustration with the guys at Upverter specifically and perhaps the web gang more broadly is that a couple of guys get some funding and think that with a good story they can market their way into relevance. But I have watched Upverter and others like it closely and it’s all smoke and no fire. There is nothing there except the easy problems. You have managed to solve what any 24 year old developer with a little bit of engineering background and access to Javascript and Mongo and some server space could solve.

      I spoke with Upverter years ago and suggested maybe I could lend a hand. Suggested maybe they should approach some of us veterans who would help out and may just do so for free. I got shot down and the attitude was “we don’t think it makes sense to hire veteran developers…”. I wasn’t looking for a job kid, I was offering to help. I *have* solved each and every one of these problems at some level and you guys want to go it alone? Fine. Just let me know what funds you’re money is coming from so I can pull mine out.

      But to Ian’s point, your closed platform and app-store mentality of “gonna-get-rich”at-seven-dollars-a-month-with-no-help-from-anyone”, means you will languish and never bloom into what a good tool might be.

      I’ve started my work on KiCAD now, just as a pet project. Why? Because it’s open. And because it’s reasonably good. Where it sucks, I know I can do better, so I’ll pitch in. Could I do the same thing for the web? Yes. Am I at all put off by JS? Hell no. Can I bang out web code with style and patterns and that make believers out of the young and old alike? Yup. And I can do it openly and honestly, without mixing my messages.

      Is it open? Is it not? You have to pick. No sitting on the fence allowed. And if it is open, you need to also be willing to cede control with the understanding that over a long enough time span, the little bumps along the way will all get smoothed out…If it’s good enough.

  10. Denis says:

    @Zak: I’m glad you’ve joined the discussion. I have a few questions:

    >>> We have open sourced every hardware design we have done

    Maybe I’m missing how this works. Is there a way I can take a design from your site and download it for use in my existing EDA tools? If not, it’s not open.

    >>> You’ve been described as “github for electronics”

    I’m not sure whether you describe yourself as this or not. However, if you truly want to match github for openness, you would either develop or adopt an open source set of design tools (which people could use without your service) and then provide the hosting/sharing/collaboration as the value-add.

    Your promises page sounds like you value openness, but your platform is proprietary and your data formats are proprietary. The big issue that many people have is that you seem to be getting a lot of value from the open hardware/software community, but not really giving much back.

    Separately, I don’t really like web-based interactive editors. They’re too dependent on reliable network connections.

    • Ben says:

      Hi Denis,

      I believe Zak and the Upverter team have said that import/export tools are on the way, so vendor lock-in probably won’t be an issue when they get to that stage. Though I can speak from experience that accurate import/export between the various EDA formats is a very hard goal to achieve, I know since this is something I am trying to do with CircuitBee. I wish them the best of luck and will lend a hand if I can.

      The approach you mention is exactly where CircuitBee is fitting in, trying to let you use whatever tool you wish to create your circuits, but then provide a service that lets you host, share, and view them online.

      The main issue raised in this article however is a tough one. The traditional models of open source software that have been profitable are going to be very hard to apply to the software as a service approach, especially for a fledgling startup. Previously successful open source projects (MySQL, Apache, Android, etc) have been engines or applications that are also used by big business, or have specific hardware requirements. This creates an ecosystem around the software that can be exploited for revenue generation, such as support, custom contracted development and hardware setup/maintenance/sale.

      This is largely not going to be the case for projects like Upverter and CircuitBee, the hobbyist community won’t pay large premiums for support, and the potential for competitors based on your own code at such an early stage in a startups lifecycle could turn out to be disastrous.

      I’d love to hear some ideas from people here about how open sourcing this type of project could be done though. If anyone has any specific ideas then please let me know.

  11. Drone says:

    “They exploit work from the open source community for profit, but keep their special sauce to themselves.”

    In my opinion Apple is the biggest offender in this respect.

  12. kuhltwo says:

    I still use the most open source ever.
    Paper & pencil. And a light table and tape for layout.
    Welcome to the “dark ages”.

    Very interesting reading. Since I have not been “enlightened” to the wonders of Eagle, KiCad, etc., this is all very interesting to read. Eventually I’ll have to learn these tools. I do agree with the OH/SW philosophy, having grown up on voiding almost every warranty I could.
    Ian is spot on from what I can tell, if you are going to talk the talk, better back it up with the walk.

    I worked for a software developer years ago, and watched Microsquirt take advantage of developer’s work an incorporate it into their own products.

  13. shuckc says:

    After about an hour of somewhat painful schematic entry, I managed to get a half-decent circuit diagram put together. It’s clear that the interface, while nicely polished, isn’t terribly reliable for some tasks – this feels more like a toy than a product. The Feedback page has feature requests for most improvements.

    Nice to have would be (1) graphic export/ hot linking, (2) component preview in chooser (3) flip as well as rotate, (4) custom parts other than rectangular ICs, (5) netlist preview/highlight to see connections actually made, (5) spice integration :-)
    Wins are: cross platform, nicely rendered, undo and rollback/version management works.

    Probably the developers too much time spent blogging about Javascript unit testing frameworks and not enough time in pre-beta usability testing!

  14. POE says:

    Ian-
    I’m glad this isn’t the case with the thingomatic’s “OPEN SOURCE HARDWARE”………………….

    Did I use enough dots to express my level of sarcasm?

    http://www.makerbot.com/docs/thingomatic/
    (2/3rds down the page)

    Has anyone been able to make a thingomatic without purchasing a single item from MakerBot? How about use their thingomatic? Nope. It’s no more opensourcehardware than these sites are opensource software.

    • Ian says:

      I did check the link, but I didn’t see what you mean. Is there a part that isn’t in the SVN repository they keep at Google Code?

  15. POE says:

    Makerbot is an online materials supplier. Among several features, it affords an individual the ability to purchase matterials for DIY 3d printers. They are a big presence at Open Source events, and have been marketing themselves as the ideal solution for open source 3d printers. That sucks, here’s why:

    This is a new class of project we’re calling closed source vampires. These outfits jump on the open source/open hardware bandwagon for venture capital, PR, and profit, but they don’t participate equally. They exploit work from the open source community for profit, but keep their special sauce to themselves.

    In this case, Makerbot links to all the resources you’ll need free of charge, but the hardware e.g. matterial sources are totally closed. They want to use the hard work of open source developers to build buzz for their company, but keep their own stuff closed. This is at best a side step from free-but-not-open (?), but with a dash of user exploitation.

    If this is a platform for open source, why not make it all open source too?
    etc. etc. etc.

    • POE:

      Huh? What are you talking about? AFAICT, Makerbot releases all the resources required for someone skilled enough to build a Makerbot from scratch to create a perfect replica. There’s full schematics, full layouts, full BoMs, and full cut files. The number of alternate sources for damn near everything in the Makerbots testifies to that. They specify what materials it takes in enough detail that neither you nor anyone else should have any particular trouble sourcing it independently. I don’t see how knowing the specific company or companies they source it from changes anything.

      • POE says:

        That’s what I thought…. Until I tried to build one without purchasing a single item from them. Figured it would be an interesting build to show how wonderful open hardware could be. Unfortunately it merely opened my eyes to the new trendy business of ‘open source’ exploitation.

        Wish I thought of it. Could have made ten million.

  16. POE: What specifically did you have trouble sourcing?

Leave a Comment

Your email address will not be published. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>

Notify me of followup comments via e-mail. You can also subscribe without commenting.