Sparkfun Splits with Adafruit

[Kean]

No, the drama which led to SparkFun refusing to engage in future business with Adafruit wasn't about open source.  Or at least what I know of it doesn't seem to have been about open source, and certainly not Teensy.

Talk of open source is coming up now only because the one thing Adafruit was buying from SparkFun which they can't make themselves was Teensy.

I'm not going to comment on the drama, partly because it's not my fight, but mostly because I didn't directly observe any of it.  I only just learned of it a few weeks ago.  Like you, I've only been able to view the aftermath of this conflict and try to piece together some understanding of what actually happened.

But I will point you to the best info I've seen so far, a "Hacker News" headline on January 14 which received over 500 comments.  A Google search for "hacker news adafruit sparkfun" should bring it right up.  Many of those comments are from Adafruit (user "ptorrone" who never types capital letters), so you can get their side of this story.  None are from SparkFun.  Most are just random uninformed opinion.  But some comments appear to be from people who may have knowledge of the drama.  Several cite sources, so if you don't have time to read through 500+ comments, maybe visually scan for links.

Hopefully you can make some sense of all this.  If you do, I'm curious to hear.

Thanks Paul.

While I usually design PCBs for many of my microcontroller implementations for clients, it doesn't make sense for smaller projects where the production quantity may only be a few dozen over the product life.  I've designed a few projects around the Teensy (mostly 3.2 & 4.0), and also used them for controlling some test jigs where I needed multiple serial ports but didn't want to include something like a RPi.  The big advantage of the Teensy to me has been the excellent (& early) Arduino integration which I am sure took a lot of work, and the hardware capabilities which were, and likely still are, some of the best in the Arduino space.

I've been buying from Sparkfun since around 2008 when I started my current consulting business, and Adafruit and PJRC since around 2012.  I also know some people who had past business relationships with Sparkfun, as well as being involved in an early Makerspace.  While I have not been privy to anything confidential, I have been following the companies with interest so I have seen a bit of the stuff going on that has led up to this.  This includes some drama in the last few months that leaked out via less popular social media (possibly in the stuff linked above by others).

Phil Torrone (PT) from Adafruit seems to have some minor but valid complaints, but from what I've seen he has a history of making a big deal of things in public that probably shouldn't be.  Limor from Adafruit has always been positive and cordial from what I've seen, but I imagine being an early & prominent female figure in a misogynistic "industry" (electronics and makers) would have its challenges.  I won't mention names, but it relatively easy to find out who it is/was at Sparkfun that PT has complaints about.  For Sparkfun this recent drama was apparently the final straw.

Sparkfun and Adafruit obviously do compete, but have coexisted in a mostly friendly manner for 15+ years or so.  They have a lot of overlap in their products, which is mostly open source anyway, but each has their own implementation and documentation style which has lead to some favouritism amongst makers for one or the other.  I can see that leading to people taking sides due to unintentional biases.  Teensy is one of the few products which they were both reselling that was not fully open source, and also which was separately documented and supported.  So PJRC got caught in the middle after Sparkfun took over manufacturing and then had enough of the drama caused by PT.

I think it is sensible that Sparkfun and PJRC have kept their comments to a minimum.  I think that anyone who does some digging will likely see this reflects badly on Phil more than any other party.  I am not saying he is the only one at fault in the ongoing disputes, he is just the one who was making it public and caused it to unnecessarily blow up.

Anyway, this is just my opinion.  I am an observer from the other side of the planet, so I do not know any of the cast personally, and I certainly cannot know/remember all the history or facts.

On wilfreds point, there are a bunch of partly open source things that have been very successful.  One that immediately comes to mind is Nordics reference designs and software stacks, where the core Bluetooth code is proprietary but most of the other stuff is relatively open.  Similarly, almost all silicon implementations are proprietary, so as Dave's video points out there is almost always a gray area - so the OSHW classifications are inherently messy.

[tszaboo]

Much as I would love to make everything 100% open source,

Why make any of it open source if you aren't, or feel you cannot, make it 100% open? With software if there was a closed binary blob that had to be linked into the final executable it would really make the rest unpalatable. Making a commercial product partially closed seems rather pointless if others cannot reproduce it and take it in other directions. Commercial or open seem kind of mutually exclusive.

Did you even watch my video?
Just the act of releasing say schematics for a product can be hugely valuable for repair or modification.
You know, the whole Right To Repair thing.

Not that I know, but how many really successful software or hardware projects are there that aren't fully open? I'd expect the answer to be none.

Any successful product that has a schematic available for starters.
The Teensy being discussed of course is very popular, if it wasn't popular, no one would care.

Exactly, releasing schematics makes it that people can understand what's going on. While you can keep ie. gerber files closed, so copying is not as straightforward as pressing some buttons. That cog was a very wise video, too bad it wasn't becoming a standard.
 

The OSHW industry is pretty tight nit here.
Founder of Adafruit LadyAda a.k.a Limor Fried is married to Phil Torrone who founded Hack-a-Day, editor of Make magazine, they have two kids (one just a few weeks old).
Founder of Sparkfun Nathan Seidle is married to Alicia Gibb who is director of the OSHW association.

After reading this: https://oshwa.org/resources/a-resolution-to-redefine-spi-signal-names/ it seems like they are less about open source or hardware, and more about SJW crap like tilting at windmills.

I expect that the whole drama will come down to "microagressions" or "misgendering". 

[EEVblog]

That cog was a very wise video, too bad it wasn't becoming a standard.

Quite a few people used it (and still do), but all the influential people (Sparkfun, Adafruit, OSHWA) wanted nothing to do with it, so it didn't get much traction. And it's not because they didn't know about it, they made the choice that it wasn't "pure" enough for them. The excuse was that Open Source Hardware should be 100% open, or nothing, no if's no buts. Yet my cog concept directly copies the Creative Commons concept in software, and that's hugely popular. And ironically, the Open Source Hardware logo and the OSWHA uses Creative Commons

[tszaboo]

That cog was a very wise video, too bad it wasn't becoming a standard.

Quite a few people used it (and still do), but all the influential people (Sparkfun, Adafruit, OSHWA) wanted nothing to do with it, so it didn't get much traction. And it's not because they didn't know about it, they made the choice that it wasn't "pure" enough for them. The excuse was that Open Source Hardware should be 100% open, or nothing, no if's no buts. Yet my cog concept directly copies the Creative Commons concept in software, and that's hugely popular.

I don't think it's purism, it's more like a business decision. They want everyone else to publish everything, to make it easy to copy, and let them work for free.
While they make stuff "open source hardware" it's the type of open source that's not very useful. I look at some of their Github repos, BOM isn't published, gerbers are not published. The supplier information isn't in the project, which is a big deal for something like an LCD. If they adapted the cog, it would be clear that they are withholding just enough information make it difficult for everyone to clone the board.
It's the same story when they give you some software, and you don't get a binary, just instructions on how to compile it on some Linux distro. Malicious compliance.
And TBH there is nothing wrong with that, if you say that this is a business, and we want to make money. But they paint themselves as the white knight of free information. (Adafruit. Sparkfun is actually honest about it as far as I see)

[EEVblog]

On wilfreds point, there are a bunch of partly open source things that have been very successful.  One that immediately comes to mind is Nordics reference designs and software stacks, where the core Bluetooth code is proprietary but most of the other stuff is relatively open.  Similarly, almost all silicon implementations are proprietary, so as Dave's video points out there is almost always a gray area - so the OSHW classifications are inherently messy.

I came up with that concept and did that video because the industry was flooded with people/companies/projects using the words "open source" or open source hardware" or "open hardware" without everything being 100% open. Always has been an issue right back to the magazine project days long before the term open source was even invented, and before the www came along.
Back then people would publish a project and give you every conceivable bit of info (even more so than today, who does a "Theory of Operation" and Block Diagrams any more?), except they'd keep the micro binary file proprietary or only sell code protected chips, or they wouldn't publish the PCB, so you had to buy it from them.

Anyway, the OSHW Associations response was to this was the gatekeep the whole thing using a registration system and their own "approved" logo through their association.

Which is fine of course, they are free to do that and people are free to use it.
They don't however acknowledge that any other form of "Open Hardware" exists.
And there has even been discussion that some registered projects aren't sharing what they are supposed to share.
And there has been community dispute over the exact terminology, with them trying to claim "Open Source Hardware" as the "enforcers" of that definition.

In the end, after a decade of debate, I think things are pretty much the same as they always have been. People will do what they want to do, and trying to gatekeep it doesn't really work. No surprise.

Anyway, this has nothing to do with the Sparkfun vs Adafruit debate, just an interesting non-sequitor.

[EEVblog]

That cog was a very wise video, too bad it wasn't becoming a standard.

Quite a few people used it (and still do), but all the influential people (Sparkfun, Adafruit, OSHWA) wanted nothing to do with it, so it didn't get much traction. And it's not because they didn't know about it, they made the choice that it wasn't "pure" enough for them. The excuse was that Open Source Hardware should be 100% open, or nothing, no if's no buts. Yet my cog concept directly copies the Creative Commons concept in software, and that's hugely popular.

I don't think it's purism, it's more like a business decision. They want everyone else to publish everything, to make it easy to copy, and let them work for free.
While they make stuff "open source hardware" it's the type of open source that's not very useful. I look at some of their Github repos, BOM isn't published, gerbers are not published. The supplier information isn't in the project, which is a big deal for something like an LCD. If they adapted the cog, it would be clear that they are withholding just enough information make it difficult for everyone to clone the board.

Yep, common as mud. The "pure definition" and their own registration system say you have to release the BOM, but what does that mean?
Some people do detailed BOM's with exact manufacturer part numbers, distributor details, price, and 2 alternative parts and other proprietary info etc
Most just do some generic thing spin spat out from the schematic.
And many as you say don't even bother.

[wilfred]

Did you watch my video?
Just the act of releasing say schematics for a product can be hugely valuable for repair or modification.
You know, the whole Right To Repair thing

I remember I watched it years ago and if I recall it was in the context of your multimeter. I remember thinking the various grades of open was a self serving gesture to signal the virtue of being open-source even though there were parts you wouldn't or weren't allowed to release. Hence I thought it wasn't truly open source. And I remain of that opinion.

I remember back in the 80's buying devices from Tandy/Radio Shack that included a schematic in the back of the manual and indeed that would help with repair. But no-one was touting it as open source as a result. I will vociferously argue that trying to claim something is open when it isn't entirely open is akin to trying to claim being a little-bit pregnant.

[jpanhalt]

I will vociferously argue that trying to claim something is open when it isn't entirely open is akin to trying to claim being a little-bit pregnant.

I think it is pretty clear there are different opinions on that.  Dictionaries contain word definitions and in some cases uses.   Very few, if any have every word in a language.  Yet, they are all dictionaries.

Take your extreme definition of "open source".  Do you include source code used for the COG chips in the display?  What about die pictures and schematics for IC's that you use?  I suspect not.  How is not including that critical information any different than when a developer excludes information for a small part of a project, but makes that part available for a reasonable price?  Just view that part like your display or IC's.  It can still be duplicated.

[Siwastaja]

I came up with that concept and did that video because the industry was flooded with people/companies/projects using the words "open source" or open source hardware" or "open hardware" without everything being 100% open.

Not everyone remembers it, but that was pretty dark ages. "Open source hardware" was labelled on anything. Heck, people advertised freaking Raspberry Pi, which is probably, after Microsoft, the #1 textbook example of extremely closed ecosystem (e.g. with actual ****ing DRM preventing 3rd party add-on module development!), as some sort of "open source hardware" project.

Your cog idea was very much needed, albeit maybe a bit too complicated with all those tiny markings meaning different things, so one would need to look up every time what they mean, in which case just free-form description does the same and is even more flexible. - i.e., honestly say what you offer and what you do not. Although the problem at the time was exactly lack of honesty.

I think we are much better now. The initial OSHW craze passed so now every crap company does not feel like they absolutely must use the term, and the remaining OSHW projects tend to be more open, including actual design files not just partial PDF output.

[floobydust]

Hopefully we're past the extremists defining what open-source is and should be, and what that logo means.
Because, I think we'll be rollin' backwards, after Qualcomm acquired Arduino. I fully expect proprietary silicon stuffed in there, for profit and shareholder value 

Regarding the hissy fit and fight, I perused some archived posts and was surprised it was going on for over 2 months.
It apparently exploded after one party doxxed the other, CC'ing some shitty email to old employers, partner's employers, relatives etc.
Wreck a person's life and reputation, in the name of social justice. Thing is, you won't change/fix/correct anyone by doing that, other than you get a blip of dopamine for being "right".

[EEVblog]

Did you watch my video?
Just the act of releasing say schematics for a product can be hugely valuable for repair or modification.
You know, the whole Right To Repair thing

I remember I watched it years ago and if I recall it was in the context of your multimeter. I remember thinking the various grades of open was a self serving gesture to signal the virtue of being open-source even though there were parts you wouldn't or weren't allowed to release. Hence I thought it wasn't truly open source. And I remain of that opinion.

So what is your solution to the problem of people claiming things are "open source" when they don't give you absolutely everything?
Just wave your hands and the problem magically goes away?

My solution is to clearly document what is available and what is not.

[EEVblog]

Hopefully we're past the extremists defining what open-source is and should be, and what that logo means.

Impossible, because their solution (strict adherence and formal registration) is completely binary. No gray area for the complexities of the hardware business is allowed, and that's the problem I was trying to solve, because I didn't see anyone else trying to solve it, they just endlessly complained about it.
But nobody really talks about it any more, so it's all kinda, meh.

Wreck a person's life and reputation, in the name of social justice. Thing is, you won't change/fix/correct anyone by doing that, other than you get a blip of dopamine for being "right".

Unfortunately very real.
Back in the day "they" tried to cancel me. When that was unsuccessful they went after people for just being associated with me.

[EEVblog]

I came up with that concept and did that video because the industry was flooded with people/companies/projects using the words "open source" or open source hardware" or "open hardware" without everything being 100% open.

Not everyone remembers it, but that was pretty dark ages. "Open source hardware" was labelled on anything. Heck, people advertised freaking Raspberry Pi, which is probably, after Microsoft, the #1 textbook example of extremely closed ecosystem (e.g. with actual ****ing DRM preventing 3rd party add-on module development!), as some sort of "open source hardware" project.

Your cog idea was very much needed, albeit maybe a bit too complicated with all those tiny markings meaning different things, so one would need to look up every time what they mean, in which case just free-form description does the same and is even more flexible. - i.e., honestly say what you offer and what you do not. Although the problem at the time was exactly lack of honesty.

I think we are much better now. The initial OSHW craze passed so now every crap company does not feel like they absolutely must use the term, and the remaining OSHW projects tend to be more open, including actual design files not just partial PDF output.

Yes, it was quite the "gold rush" and seems to have naturally dissipated a lot. The debate is considered old hat these days, and there seems to be a lot less people caring about using the OSHW logo any more, even though I'm sure the problem still exists if anyone cares to look into it.

It seems that the OSHWA has now switched focus to "Open HealthWare"
https://oshwa.org/announcements/oshwas-new-open-healthware-certification-how-we-got-here-and-where-were/
With 4 of the 7 team members dedicated to it?

[jpanhalt]

So what is your solution to the problem of people claiming things are "open source" when they don't give you absolutely everything?
Just wave your hands and the problem magically goes away?

My solution is to clearly document what is available and what is not.

1) Why are false claims of "open source" a problem?  Whenever I buy from SF or Adafruit, I always check what's available before buying.  I don't pay attention to any claim of open source.

2) Your solution might work in theory but only kicks the can.  There will still be arguments about whether the letter assigned to each cog is accurate.  Like in my analogy to IC's and displays (and other components), what can be ignored as not being open source but still label that cog as open source?

EDIT:  It just occurred to me that projects presented with Assembly code might not be considered open source as they may fail to meet a standard such as, "Can be duplicated by a person normally skilled i the subject matter."  (That's a standard applied by chemistry journals for the experimental section of a paper.) 

[wilfred]

Did you watch my video?
Just the act of releasing say schematics for a product can be hugely valuable for repair or modification.
You know, the whole Right To Repair thing

I remember I watched it years ago and if I recall it was in the context of your multimeter. I remember thinking the various grades of open was a self serving gesture to signal the virtue of being open-source even though there were parts you wouldn't or weren't allowed to release. Hence I thought it wasn't truly open source. And I remain of that opinion.

So what is your solution to the problem of people claiming things are "open source" when they don't give you absolutely everything?
Just wave your hands and the problem magically goes away?

My solution is to clearly document what is available and what is not.

I'm not involved directly in the open source world. I'm just looking in from the outside. It is my lay persons view that people claiming something is open-source when it truly is not fully open are behaving deceitfully and should be called out for it. Open should mean there is sufficient detail for a skilled person to reproduce the  article and make changes without constraint for want of missing details. Less than that then thanks for the details provided. But don't try to call it open.

Clearly documenting what is and isn't available is fine but that can still be done without calling it open.

The problem seems to be with a product that has commercial value that concealing part of the design seeks to protect. I do not accept redefining open is any sort of solution. This is a very semantic argument about just what the open in open source means and I think for the sake of simplicity it is better to leave the ambiguity aside.

I have doubts about the relevance of this but I'm reminded of the riddle I first heard in the 1960's Batman TV show. Q: When is a door not a door? A: When it is ajar. Would you still call the door open? You can't pass through it, but it isn't closed. It depends on the semantics of how you define open.

[5U4GB]

If they adapted the cog, it would be clear that they are withholding just enough information make it difficult for everyone to clone the board.

Never attribute to malice what is adequately explained by "well it works when I do it".  It seems to be more or less standard for smaller open-source projects that the first few hours or using them is trying to replicate whatever unique setup the originator had that made the thing actually buildable and working.  Even ones where they think they've documented the silly-walk appropriately invariably miss out a pile of steps that surely everyone knows so there's no need to list them.

[floobydust]

Projects get posted and seem to be more a "look at what I made" in order to gain attention - yet the project docs are actually a trainwreck.
Majority of the projects on hackaday.io have missing docs "Files: 0". You won't be making them. What's the point then? A bunch of noise, that site reminds of MySpace lol.

[jpanhalt]

Ironically Hackaday was founded by Phillip Torrone, as mentioned earlier.  I agree with that comment as most projects I have been interested in are not nearly open source.

[AuricG]

(Please note that while I use "you" below, I don't mean you `wilfred` in particular, but instead it refers to "The Reader" and I'm just trying to give another point of view for your riddle.)

I have doubts about the relevance of this but I'm reminded of the riddle I first heard in the 1960's Batman TV show. Q: When is a door not a door? A: When it is ajar. Would you still call the door open? You can't pass through it, but it isn't closed. It depends on the semantics of how you define open.

I think this is very relevant "example", because the point actually changes and depends very much on the perspective!
But it also shows why I completely agree with Dave's (EEVBlog) point on why the word "open" (hardware) can't be defined in terms of black/white or true/false (at least not in the hardware world).

The riddle's meaning will immediately change as soon as you have a stake or "skin-in-the-game" if you like.
If door (or something) is not open, then it must be closed?! Right?
If this is true, then if the door to your house is ajar, why do you bother closing it? Why not leave it ajar since it's the same thing as closed? It's definitely not _fully_ open, so that means it's closed, right?
I think it's probably because now you have a stake and want to make sure you (and your property) remain safe.
And you know that depending where you live, even if you leave the door ajar over night or many nights in a row, it does not mean someone will break in right away or something horrible is going to happen.
But I doubt you would consider a front door to your house that is ajar closed or sleep well at night unless it's fully closed.
The same goes for open door, how much open door is open? 45 degrees, 60 degrees, 180 degrees, door off the hinges? You don't need the door to be _fully_ open to enter, however if you are moving in a fridge... every bit counts and the more the better. Unfortunately same goes if someone is looking to clone a product or make a profit.

Open should mean there is sufficient detail for a skilled person to reproduce the  article and make changes without constraint for want of missing details. Less than that then thanks for the details provided. But don't try to call it open.

I get your point of view and you have every right to it (and as a hobbyist I agree with you completely). Unfortunately, a lot of "business" often use "open" hardware/software as a marketing keyword and to gain attention/publicity.
But we are talking here about open hardware, not open software! So in this case we also have to define what is "skilled person" and what does that mean? What tools and materials they have access to?

I understand why for example CERN releasing every possible design/document for their particle accelerator as open-source and open-hardware does not bother them much. Good luck building those in China for cheap and selling them on eBay/AliExpress.
Okay, maybe a bad example because this is "academia" and sharing and peer review is "necessary", but the point on cloning and selling those designs still stands.

But at the same time, someone making a product (not a half-baked weekend/hacky project but an actual "polished" product) and they decide to release complete schematics (maybe even with pcb layout) and firmware source code. But they don't provide gerbers, pick and place, BOM and other files needed for (mass) manufacturing.
Now you can can understand how it works probably learn something new/cool from the schematic/layout and also make modifications in whichever way you want or integrate it into something, because they provided the documents for you to know the "ins and outs".
To me this is very open and what I would like consider the threshold for "open" hardware project (but again, I'm just offering my point of view).
But I would love to hear why this is not "open hardware"?

This might be a very unpopular opinion but it's my own and we can have different ones and still be friends.
To me, this is not the same as "open software" and what really grinds my gears is that for (a very vocal minority of) hobbyist people, the issue is not about:
"Hey buddy, I would love to share my time and expertise so I can help you with this schematic/layout or I can help you optimize the BOM or add a firmware feature, that's what Open is all about".
But instead it seems to me that (again, vocal minority is making) the situation more along the lines "This is open source world buddy, hand over _everything_ that you have and that you know or else we will bully you into submission! And don't even think of hiding something, because cavity search is part of the process. Why? Well because that's what open means buddy and you made a mistake of calling your project open on the internet!".

Don't get me wrong, as a hobbyist, I love "open", and the more open stuff you give me the better. That's very greedy of me, but it's in human nature and also I have zero skin-in-the-game. I can either gain something or things remain the same (for me).
But at the same time, we as a community have to realize that this is not sustainable in the long run (take everything and give nothing back).
And we should protect our community from those who want to fan the flames by "s*it-talking" and bullying people/projects because they shared "more than enough" of their project but didn't share _absolutely everything_.
I also personally think that some "hobbyists friendly corporations" are very happy to "fan the flames" because it works in their favor. (Also let's get this straight, their cause might be aligned or born from the hobbyists, but they have bills to pay and mouths to feed. Absolutely nothing wrong with that but they are not poor mom and pop shops, they are corporations/companies that have to make money in order to survive).
Because if I'm one of these "hobbyist friendly corporations", I'm mostly making money from manufacturing (eval/dev boards). "Hey look, we made a Pico/Arduino/led/sensor eval board. Here are all the files that you need to make your own. Oh it's exactly the same as 500 other boards on the market from different shops, but this one has our company logo and custom standard (proprietary?) connector, and it's open! Keep in mind we shared everything because we care (although we probably mainly just copied somebody else's design)".
Now please keep in mind that I'm not saying that they are evil or anything like that, but at the same time... maybe they are also not all that altruistic and open hardware saints as they would like to portray themselves in our "open" world .

Unfortunately I've seen multiple projects and know people who burned out and gave up on "open hardware" and "open source" projects because vocal minority of people were "bullying" them and requesting to make the project meet their own definition of open.
And again, not because they want to contribute to that project in any shape or form. No, it was just because they saw the word "open" and this gave them the idea that it's their birth right to demand everything that was ever made or touched that project to be handed over...

We as a community have to realize and accept that "open" is a very broad term. Door that is ajar is not closed, it's partially open, this is the definition. Is it ideal or would it not be better if the project was fully open? Absolutely! At least for me. But this also depends on the perspective and how much "skin-in-the-game" you have...
But for the open hardware community, if there are no other options, partially open is still much better than fully closed, at least for me.
So I'm hoping that maybe (one day) we as the community can agree that "open" is an acceptable broad term that can be used for describing a project and that "open" does not imply "fully open", as much as a lot of us would love that to be the case.

[EEVblog]

Open should mean there is sufficient detail for a skilled person to reproduce the  article and make changes without constraint for want of missing details. Less than that then thanks for the details provided. But don't try to call it open.

I get your point of view and you have every right to it (and as a hobbyist I agree with you completely). Unfortunately, a lot of "business" often use "open" hardware/software as a marketing keyword and to gain attention/publicity.
But we are talking here about open hardware, not open software! So in this case we also have to define what is "skilled person" and what does that mean? What tools and materials they have access to?

AFAIC if a company say releases a schematic for repair/mod/whatever, then they are (at least partially) "opening" the product, and should be celebrated for doing that, and be allowed to at least say they are "opening" the product.
This is why many people prefer to use the term "Open Hardware" instead of (the "purest") "Open Source Hardware", and I have no problem with that distinction if you want to make it.
But I guarantee that if you go to official "registered" OSHW projects (over 3000 of them), you'll find a huge disparity in what each project provides.
There isn't even consistency in the licensing.

The OSWHA themselves even tell you in the checklist that you don't have the open absolutely everything 
https://certification.oshwa.org/process/hardware.html

Have you clearly indicated which parts of a product are open-source (and which aren’t)?

Have you applied an open source license to your hardware?

After certification, remember to:

    Label your hardware with a version number or release date, so people can match the physical object with the corresponding version of its design files.
    Use the OSHWA certification mark logo on your hardware. Do so in a way that makes it clear which parts of the hardware the logo applies to (i.e. which parts are open-source).

(my bold)

But at the same time, someone making a product (not a half-baked weekend/hacky project but an actual "polished" product) and they decide to release complete schematics (maybe even with pcb layout) and firmware source code. But they don't provide gerbers, pick and place, BOM and other files needed for (mass) manufacturing.
Now you can can understand how it works probably learn something new/cool from the schematic/layout and also make modifications in whichever way you want or integrate it into something, because they provided the documents for you to know the "ins and outs".
To me this is very open and what I would like consider the threshold for "open" hardware project (but again, I'm just offering my point of view).
But I would love to hear why this is not "open hardware"?

As Louis Rossman would say "Schematics or GTFO".
If every company did this this would be HUGE.

[wilfred]

Open should mean there is sufficient detail for a skilled person to reproduce the  article and make changes without constraint for want of missing details. Less than that then thanks for the details provided. But don't try to call it open.

I get your point of view and you have every right to it (and as a hobbyist I agree with you completely). Unfortunately, a lot of "business" often use "open" hardware/software as a marketing keyword and to gain attention/publicity.
But we are talking here about open hardware, not open software! So in this case we also have to define what is "skilled person" and what does that mean? What tools and materials they have access to?

I understand why for example CERN releasing every possible design/document for their particle accelerator as open-source and open-hardware does not bother them much. Good luck building those in China for cheap and selling them on eBay/AliExpress.
Okay, maybe a bad example because this is "academia" and sharing and peer review is "necessary", but the point on cloning and selling those designs still stands.

A "skilled person" has suitable skill and expertise required to take advantage of a complete specification. I do not even think a PCB design is necessary but that does make it easier for someone to reproduce the device if that is the aim of putting an open source design out there for others to make use of and adapt. That's how open designs can evolve and better meet the community needs. I look upon it very much as an exercise in Darwinian evolution.

But there is always the question as to, in the PCB example, what CAD program the design is in. If it is an expensive commercial program then it makes life more difficult. If it is in an open source program like Kicad that would have wider adoption and probably appeal to a purist. Which I am not. 3D printing probably has similar issues.

I'm assuming that some hardware would need a PCB design that embodies highly specialised design elements in which case not supplying a PCB design would go somewhat toward considering the design is not fully open.

None of it is clear cut but it does go to the point you raise about the motives for declaring a design open. A commercial design attempting to masquerade as open in order to advance a marketing campaign or to signal some dubious virtue is for me the crux of the matter. The valid use of open leaves all that sort of questionable conduct behind. The motives are crystal clear in that the design is freely and wholly shared for the common good.

[brucehoult]

If they adapted the cog, it would be clear that they are withholding just enough information make it difficult for everyone to clone the board.

Never attribute to malice what is adequately explained by "well it works when I do it".  It seems to be more or less standard for smaller open-source projects that the first few hours or using them is trying to replicate whatever unique setup the originator had that made the thing actually buildable and working.  Even ones where they think they've documented the silly-walk appropriately invariably miss out a pile of steps that surely everyone knows so there's no need to list them.

I can't speak for anyone else, but I always make sure my published instructions / build script work from a fresh OS install -- something very easy to check these days using docker or similar.

[EEVblog]

None of it is clear cut but it does go to the point you raise about the motives for declaring a design open. A commercial design attempting to masquerade as open in order to advance a marketing campaign or to signal some dubious virtue is for me the crux of the matter. The valid use of open leaves all that sort of questionable conduct behind. The motives are crystal clear in that the design is freely and wholly shared for the common good.

Any "open" is better than closed.
As I said, all the "official" registered OSHW projects are not equal in openness.
What pure OSHW really means is "Take whatever we give you, and the license enables you freely use it without legal concern for your commercial business".
What that "whatever we give you" is varies a lot.

[AuricG]

None of it is clear cut but it does go to the point you raise about the motives for declaring a design open. A commercial design attempting to masquerade as open in order to advance a marketing campaign or to signal some dubious virtue is for me the crux of the matter. The valid use of open leaves all that sort of questionable conduct behind. The motives are crystal clear in that the design is freely and wholly shared for the common good.

The crux of the matter is that a very vocal minority of people don't want to differentiate between "fully open" and "open". Instead they chose on their own to police the internet and harass people because they assume open and fully open have to be the same thing and there is no other way.
I believe that for hardware products, this is very naive approach and mostly pushed by either people who never designed and published anything really worth making and selling, or by companies that live from making (cloning) stuff and selling it for profit.

I have seen many examples of "fully open" projects from these "We are fully open bro, so much that we don't even have the kimono on" corporations that are mentioned in this thread where it's impossible to make heads or tails looking at the manufacturing files.
No we can argue that this opaqueness is benign and probably because of someones incompetence, lack of time, have a personal life, coop student did it etc... or maliciously doing this because they know competition is also taking a look at these files.

Let's be real and really blunt for a second; Adafruit, Sparkfun, DFRobot, SeedStudio, Microsoft, Google, clone shop from AliExpress... they all are business companies and _have to_ make money in order to pay bills.
Is there anything wrong with that? Absolutely, 100% not!
But I think it would be very naive to believe Adafruit or Sparkfun or any other "fully open" companies that you can think of are not trying to protect their business and are sacrificing their profit for the greater good of "fully open" community.
Adafruit and Sparkfun both have 100+ employees. What do you think would happen if they can't pay salaries, pay rent, pay suppliers or pay bills?

Below is my blunt/harsh view on this whole elementary school drama.
Adafruit and Sparkfun are business corporations and I believe they should be treated like that, not like they are some mom and pop shop employing five people.
From my perspective, PTorrone is going around the internet panicking that they might lose sales and trying to get "sympathy" from people/supporters that would inevitably recoup or boost some sales.
He is whining about not being able to sell a "closed source" product, after selling it for years...
Why is a "fully open" company even selling closed source products? Isn't that against what you believe in/stand for in the maker community?
Oh, right, you are "supporting" makers and the maker movement... but also making a nice profit from that thing you stand so firmly against.
Nothing wrong with making a profit, but don't BS and brainwash people how you and your company are the messiah of the open hardware world, fighting to make everything fully open _for the greater good of the community_.

And at the same time he is blatantly _marketing_ his new product on _every_ popular forum/site on the internet, even though this new product is completely unrelated to Teensy, and not even out yet.
I don't think anyone would get away without at least being banned on Adafruit/Sparkfun or any other forum. And definitely there would be an angry online mob defending Adafruit and saying how this behavior is unacceptable and everyone should boycott this new product.
Yet somehow people don't comment on this or don't want to in fear of poking the angry online mob or maybe just because they like the taste of that cool aid.

Teensy has existed for how long now? And coincidentally Adafruit, after years of selling the product for profit, just now realized that "fully open" solution is what is needed, just as they were cut off. How convenient.
The maker community was basically mostly unaffected by this change. Teensy still existed, you can still buy it, you can still make projects with it. Only thing changed is that _only_ Adafruit can't sell it anymore and they might loose some revenue...
And what is their solution? Are they actually investing time and money into building a competing or a better product? In my opinion, absolutely NO.
They are just regurgitating one of their existing products into a new form factor and hoping people will take it and start buying.
I think they are leeching off by blatantly marketing it online and calling it "working name freensy" and "fully open" replacement, when even on paper it can't match Teensy.
Obviously we could exaggerate and say that the goal for this "freensy" is to blink an LED, then even an ATTiny or NE555 would be considered a valid replacement.

Bottom line is that this event... whatever you want to call it, has made me lose some respect that I had for both of these companies and at the same time challenged my very naive look at how some companies "align and present themselves" to the maker community.

[nctnico]

And at the same time he is blatantly _marketing_ his new product on _every_ popular forum/site on the internet, even though this new product is completely unrelated to Teensy, and not even out yet.
I don't think anyone would get away without at least being banned on Adafruit/Sparkfun or any other forum. And definitely there would be an angry online mob defending Adafruit and saying how this behavior is unacceptable and everyone should boycott this new product.
Yet somehow people don't comment on this or don't want to in fear of poking the angry online mob or maybe just because they like the taste of that cool aid.

IMHO it is not about fear of poking the bear. It is blatantly obvious that Freensy thing is not going to see the light of day ever. It is just childish anger directed into designing a PCB and blabbering about it on internet. This whole situation is bad for Adafruit. Who wants to to business who somebody who behaves like that?