[Discuss] Proposal: Open Source Hardware Score/Index

Fri Feb 27 20:48:20 UTC 2015

Hi, Nancy - not sure if you wanted to reply to just RJ and I, but CC'ing
RJ. I'm busy this weekend, but RJ's been working on a Bower-like system for
open sensor networks, so he knows way more than I about the ins and outs of
how the Bower and NPM servers work.

I'm tempted to just try cloning Bower's "registry" repo, i.e.
https://github.com/bower/registry and see if that could be modified to
work. What do you think?

I agree that we might name the file something broader, but I like the
reference to the really basic "CONTRIBUTORS.TXT" found in many FOSS
projects.

On Fri, Feb 27, 2015 at 2:45 PM, Nancy Ouyang <nancy.ouyang at gmail.com>
wrote:

> Hey RJ / Jeff / possibly other interested folks,
> I've been toying with the idea of making isitactuallyopensource.info
> which gives small projects an automatic 0--5star review for about a month
> now. For open source software, actually, but could easily be
> expanded/forked to cover OSHW.
>
> Maybe it'd be a stopgap until people adopt contributors.json (I'm not sure
> that's the ideal name, it seems broader in scope than 'credits', for me
> it'd be more like YesThisIsActuallyOpenSource.json) -- it was going to be
> some auto-evaluation of how quickly a bot can find your source and license,
> and maybe allow users/authors to fill in anything my crappy bot failed at
> finding. *
>
> Anyway, I was going to code up a prototype webapp this weekend, if either
> of you are free and around Union Sq. this weekend.
>
> Thanks,
> --Nancy
>
>
> * I'm told this is also somewhat of a problem in science / scientific
> articles, where people will claim code is open-source and never get around
> to actually posting it online (that's fine if you're too busy, just don't
> claim it's open source). The ratings I imagine would be humorously like
> Michelin's restaurant rating:
>
> 5: public domain or copyleft or something
> 4:
> 3:
> 2:
> 1: obscurely "open-source" (source code printed and buried under a rock in
> Siberia)
> 0: closed-source
>
> It's somewhat tied to the reproducibility movement, since if your software
> has undocumented dependencies, it kind of defeats the purpose, but I
> haven't thought too much beyond this.
>
>
> ~~~
> narwhaledu.com, educational robots
> <http://gfycat.com/ExcitableLeanAkitainu> [[<(._.)>]] my personal blog
> <http://www.orangenarwhals.com>, orangenarwhals
> arvados.org (open source software for provenance, reproducing, and
> scaling your analyses)
>
> On Fri, Feb 27, 2015 at 12:38 PM, David A. Mellis <dmellis at gmail.com>
> wrote:
>
>> I agree that it’s important to provide consistent guidelines and
>> standards, and make them as clear and easy to follow / evaluate as
>> possible. And I think the OSHW definition is a good standard to start with
>> (although we may want to tweak / improve it at some point if necessary).
>> Again, any suggestions on how to better communicate the definition (both to
>> hardware makers and to users) is welcome.
>>
>> I’d be curious to hear other people’s opinion on establishing other
>> standards to complement the OSHW definition. What do you all think about
>> trying to define a partially open standard or a more pure OSHW standard?
>>
>> David
>>
>> On Feb 27, 2015, at 12:34 PM, Mario Gómez <mxgxw.alpha at gmail.com> wrote:
>>
>> Hi David!
>>
>> Thanks for your comments,
>>
>> At the begining I was thinking at the score as a measure of the level of
>> compliance with the definition and not a black and white classification.
>> However in the way the OSHW definition is redacted you can practically
>> guess which situations can directly prevent a project of being "pure OSHW"
>> and that's what I tried to include in the questions.
>>
>> However I consider that the score is not the the most important thing,
>> it's the way that the score gives you insightful recommendations and a
>> system that allows crowd-validate the compliance.
>>
>> The other thing that I really did wanted with the score is that it could
>> serve for certification purposes. For example, you cannot state "my
>> bussiness complies 99.99% with the ISO 9001 requirements", I mean you can
>> say it but no one is going to take you seriously. And to be sincere, for
>> OSHW the line of what-is and what-not must be drawn somewhere. I would
>> think that for many shady manufacturers it's really good that there isn't
>> any clear line drawn yet because they can market their products as Open
>> Source Hardware without following the spirit of the OSHW definition.
>>
>> For the levels personally I think they add more confusion to the issue of
>> what is and not open source hardware, however they can be helpfull to guide
>> the designers about what they need to do if they (ever) want to release
>> their designs as OSHW.
>>
>> My point is: If we, as members of the OSHWA do not draw a line somewhere
>> and use some tool that allow us to do it in a consistent, replicable and
>> transparent way... Then someone else is going to do it and there is a risk
>> that they draw the line in a place where there is no true intent of
>> following with the "Open Source Hardware" philosophy that the OSWHA tries
>> to promote.
>>
>> Regards,
>> Mario.
>>
>>
>> On Fri, Feb 27, 2015 at 11:05 AM, David A. Mellis <dmellis at gmail.com>
>> wrote:
>>
>>> To me, it’s confusing for the required criteria to yield a score, if
>>> only a perfect score counts as OSHW. That is, a naive reader might think
>>> that 13/15 or 14/15 is a good score, even though we wouldn’t consider the
>>> project OSHW. To me, it seems like we’re better off using a checklist
>>> approach instead, i.e. these are all the things you have to do to be
>>> considered OSHW. OSHWA has some things like that already:
>>>
>>> http://www.oshwa.org/wp-content/uploads/2014/08/oshwchecklist.pdf
>>> http://www.oshwa.org/wp-content/uploads/2014/08/OSHW-May-and-Must.pdf
>>>
>>> Although suggestions are always welcome.
>>>
>>> A related approach I’d love feedback on is whether there is a
>>> well-defined and agreed on set of practices that could constitute either a
>>> weaker or stronger standard than our current OSHW definition.
>>>
>>> For example, can we imagine trying to establish a meaning of “partially
>>> open” hardware — e.g. hardware for which design files are released but
>>> under a more restrictive license than OSHW; hardware for which some files
>>> (like schematic PDFs) are released but not others (like the actual design
>>> files). This is still more open than many pieces of hardware, so it might
>>> be worth trying to recognize these efforts, even if they’re not fully OSHW.
>>> Thoughts?
>>>
>>> In the other direction, could we imagine something like a “pure
>>> open-source hardware” standard, e.g. hardware which is designed using
>>> open-source software tools, and which only uses, say, components that are
>>> standard / widely available / publicly documented?
>>>
>>> David
>>>
>>> On Feb 25, 2015, at 7:12 AM, Mario Gómez <mxgxw.alpha at gmail.com> wrote:
>>>
>>> Hi Ben,
>>>
>>> That's the idea of the proposed score, there is a set of questions that
>>> evaluate compliance against the OSHW definition. Your project must meet the
>>> required score 15/15 to be considered OSHW.
>>>
>>> The reason why is a score instead a simple evaluation of compliance is
>>> because I was thinking that it also must work as a tool for the begginer
>>> that want to develop OSHW and a guide of which changes are needed to be
>>> compliant. Currently in the way the score is designed you must have 15 of
>>> 15 points of compliance to be considered OSHW if you doesn't meet all of it
>>> well... then your project simply isn't OSHW. However you'll know after the
>>> evaluation how far is your project of getting the goal, it's not the same
>>> to get a score of 1 than a score of 14. The system later would underline
>>> the things that you  failed to comply and (hopefuly) give you a guide or
>>> ideas about what to do.
>>>
>>> After the 15 "required" points there are 7 aditional points that
>>> evaluate good practices. The idea of including this in the calculation of
>>> the score is because in some way is easy to comply with the definition but
>>> that doesn't guarantee that you are following good practices. Then again if
>>> you've got the 15 required points the extra points help you to know if you
>>> are following the best practices and giving added value to your project
>>> generating a good and accesible documentation.
>>>
>>> Also I think that the definition is pretty clear of what things prevent
>>> a project to be considered OSHW and the questions of the score were
>>> elaborated that way, following the definition.
>>>
>>> Regards,
>>> Mario.
>>>
>>>
>>> On Wed, Feb 25, 2015 at 1:55 AM, Ben Gray <ben at phenoptix.com> wrote:
>>>
>>>> Although I like the idea of an index, it seems to be enough of a
>>>> problem (even on this list) to recognise what constitutes Open Source
>>>> Hardware or not. I feel that adding an index or score could muddy the
>>>> waters even more.
>>>> However it could add to understanding if the compliance elements are
>>>> stressed and failure underlined rather than a low score given.
>>>>
>>>> --
>>>>
>>>> Best Regards
>>>>
>>>> Ben Gray - Director
>>>>
>>>>
>>>>
>>>> www.phenoptix.com
>>>> twitter.com/phenoptix
>>>> plus.google.com/+phenoptix
>>>>
>>>>
>>>>
>>>>
>>>>
>>>>
>>>> On 25 February 2015 at 07:16, Jeffrey Warren <jeff at publiclab.org>
>>>> wrote:
>>>>
>>>>> So one thing I like about the contrib.json file is that it'd have a
>>>>> BOM requirement with potentially optional things like prices, links for
>>>>> where to buy materials, etc.
>>>>>
>>>>> I had some ideas (talking with RJ Steinert
>>>>> <http://publiclab.org/profile/rjstatic> of Farm Hack) about how a
>>>>> more Bower- or NPM-style utility could parse such files... these are just
>>>>> roughly sketched out ideas -- say we called it "newt":
>>>>>
>>>>>    - newt init -- would run a text-based questionnaire to generate
>>>>>    contrib.json file
>>>>>    - newt compile bom -- aggregate/merge BOMs of nested projects
>>>>>    - newt compile bom <string> -- aggregate/merge BOMs with links
>>>>>    matching provided string like "digikey.com"
>>>>>    - newt compile price <int> -- calculate unit price for int units
>>>>>    - newt compile contributors -- compile contributors of nested
>>>>>    projects
>>>>>    - newt register -- makes searchable, tests for presence of req'd
>>>>>    docs, clones repos or zips
>>>>>
>>>>> Updated my post in the comments here, where there's also been some
>>>>> discussion about versioning:
>>>>> http://publiclab.org/notes/warren/02-24-2015/standardizing-open-source-hardware-publication-practices-with-contributors-json#c11215
>>>>>
>>>>>
>>>>>
>>>>> On Tue, Feb 24, 2015 at 7:35 PM, Roy Nielsen <amrset at gmail.com> wrote:
>>>>>
>>>>>> Hello,
>>>>>>
>>>>>> One possibility would be to require a "BOM" or bill of materials that
>>>>>> is required for an OSHWA certified design.  Perhaps something like the
>>>>>> following for an embedded board:
>>>>>>
>>>>>> * contributors.jason
>>>>>> * Project BOM - in the part descriptions - includes whether a part is
>>>>>> open source or closed source
>>>>>>                           (ie processors, complex chips, etc)
>>>>>> * Schematics list - including descriptions & if the schematics are
>>>>>> modifiable (ie, not pdf)
>>>>>> * License
>>>>>> * Hardware Design Documentation
>>>>>> * Software Design Documentation & License (if applicable, like
>>>>>> firmware)
>>>>>> * Connectors - if they are open design/interface
>>>>>>
>>>>>> anything else?
>>>>>>
>>>>>> Score could possibly be based on what of the above is available . .
>>>>>>
>>>>>> Regards,
>>>>>> -Roy
>>>>>>
>>>>>>
>>>>>> On Tue, Feb 24, 2015 at 4:15 PM, Pablo Kulbaba <
>>>>>> pablokulbaba at gmail.com> wrote:
>>>>>>
>>>>>>>  On the validation via a community or a specific group of people,
>>>>>>> maybe the initial open community can provide a seedstock to raise educated
>>>>>>> people to form a later trusted group of people that gives an ulterior
>>>>>>> certification.
>>>>>>>
>>>>>>> PD: Had to search JSON.
>>>>>>>
>>>>>>> On 24/02/2015 08:00 p.m., Mario Gómez wrote:
>>>>>>>
>>>>>>>  @jeff:
>>>>>>>
>>>>>>> That's great! It can even work both ways: If you already have a JSON
>>>>>>> you can provide the URL to automatically calculate the indicator for your
>>>>>>> project and vice versa: if you complete the questionnaire it could
>>>>>>> automatically generate the JSON file that you can include in your project
>>>>>>> as you propose that would be easy to do.
>>>>>>>
>>>>>>>  Sadly I'm a little busy this week but let me see if I can program
>>>>>>> a functional prototype so we can experiment how it could work for the next
>>>>>>> month. (I would not mind if someone else wants to help)
>>>>>>>
>>>>>>> @Javier:
>>>>>>>
>>>>>>>  I personally like the idea of the community, because if the
>>>>>>> process is straight forward, verifiable and transparent what matters is the
>>>>>>> result of the evaluation system and not the person/group of persons doing
>>>>>>> the evaluation. This is assuming that the evaluation system provides means
>>>>>>> to minimize/prevent abuses (That's why I consider important to also
>>>>>>> implementing a meta-evaluation system).
>>>>>>>
>>>>>>> However... being certified from a trusted group of people it's
>>>>>>> really important and I think that the OSHWA could be an appropriate group
>>>>>>> to do that. But let's hear more opinions, I think that it's possible to
>>>>>>> build something simple that helps people to follow the OSHW philosophy in
>>>>>>> their projects.
>>>>>>>
>>>>>>> Regards,
>>>>>>> Mario.
>>>>>>>
>>>>>>>
>>>>>>> On Tue, Feb 24, 2015 at 3:54 PM, Jeffrey Warren <jeff at publiclab.org>
>>>>>>> wrote:
>>>>>>>
>>>>>>>> I really like this idea!
>>>>>>>>
>>>>>>>>  Somewhat related is this idea from chatting with Alicia Gibb a
>>>>>>>> few months ago, of a contributors.json file which would fulfill (with
>>>>>>>> links, short descriptions, etc) all the terms of the OSH definition.
>>>>>>>>
>>>>>>>>  I finally typed up the idea and our sample format here:
>>>>>>>>
>>>>>>>>
>>>>>>>> http://publiclab.org/notes/warren/02-24-2015/standardizing-open-source-hardware-publication-practices-with-contributors-json
>>>>>>>>
>>>>>>>>  Love to hear input. Perhaps the questionnaire could generate such
>>>>>>>> a file. At Public Lab, it'd be interesting for the file to be
>>>>>>>> auto-generated from our tool wiki pages. The nice part about it is that
>>>>>>>> it's not specifying a way of browsing or aggregating projects (as other
>>>>>>>> folks are exploring that space) but specifies a standard way to make the
>>>>>>>> relevant/required information available for such projects to
>>>>>>>> scrape/consume. Also, it's easy enough to write by hand and include in a
>>>>>>>> github repository.
>>>>>>>>
>>>>>>>>  Best,
>>>>>>>> Jeff
>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>> On Tue, Feb 24, 2015 at 3:55 PM, Javier Serrano <
>>>>>>>> Javier.Serrano at cern.ch> wrote:
>>>>>>>>
>>>>>>>>> Mario, I think this is a great idea. I see this can play a role in
>>>>>>>>> the
>>>>>>>>> solution to one of the biggest problems of OSHW: how to make sure
>>>>>>>>> developers have more incentives to publish their work. Economic
>>>>>>>>> incentives in particular. An OSHW label could give (more) prestige
>>>>>>>>> to
>>>>>>>>> developers who hold it and induce purchaser-driven growth of OSHW.
>>>>>>>>> We
>>>>>>>>> are already seeing that prestige is a big element in the success
>>>>>>>>> of OSHW
>>>>>>>>> companies. A well advertised and supported label or mark could
>>>>>>>>> enlarge
>>>>>>>>> the population of savvy customers.
>>>>>>>>>
>>>>>>>>> On 02/24/2015 05:58 PM, Mario Gómez wrote:
>>>>>>>>> > The idea is that the community validates if you are telling the
>>>>>>>>> truth.
>>>>>>>>> > To prevent abuse a meta-validation system could be implemented
>>>>>>>>> were you
>>>>>>>>> > can "evaluate the evaluators" to see if their are being fair on
>>>>>>>>> their
>>>>>>>>> > evaluations.
>>>>>>>>>
>>>>>>>>> One alternative is to entrust the OSHWA with that role.
>>>>>>>>> "Community" is a
>>>>>>>>> vague term. If I have to trust someone on whether a piece of
>>>>>>>>> software is
>>>>>>>>> free software I will trust the FSF over the "community" any day.
>>>>>>>>> One way
>>>>>>>>> of doing it would be through a creative use of marks or labels, in
>>>>>>>>> the
>>>>>>>>> vein of what OHANDA [1] proposes. See also the work of the
>>>>>>>>> Wikimedia
>>>>>>>>> Foundation [2] in this regard. In this scenario, developers have a
>>>>>>>>> natural incentive to not misuse the mark, because they can be sued
>>>>>>>>> with
>>>>>>>>> all the arsenal of trademark law if they do.
>>>>>>>>>
>>>>>>>>> Cheers,
>>>>>>>>>
>>>>>>>>> Javier
>>>>>>>>>
>>>>>>>>> [1] http://www.ohanda.org/
>>>>>>>>> [2] http://wikimediafoundation.org/wiki/Trademark_policy
>>>>>>>>>  _______________________________________________
>>>>>>>>> discuss mailing list
>>>>>>>>> discuss at lists.oshwa.org
>>>>>>>>> http://lists.oshwa.org/listinfo/discuss
>>>>>>>>>
>>>>>>>>
>>>>>>>>
>>>>>>>> _______________________________________________
>>>>>>>> discuss mailing list
>>>>>>>> discuss at lists.oshwa.org
>>>>>>>> http://lists.oshwa.org/listinfo/discuss
>>>>>>>>
>>>>>>>>
>>>>>>>
>>>>>>>
>>>>>>> _______________________________________________
>>>>>>> discuss mailing listdiscuss at lists.oshwa.orghttp://lists.oshwa.org/listinfo/discuss
>>>>>>>
>>>>>>>
>>>>>>> --
>>>>>>> PabloK
>>>>>>>
>>>>>>>
>>>>>>> _______________________________________________
>>>>>>> discuss mailing list
>>>>>>> discuss at lists.oshwa.org
>>>>>>> http://lists.oshwa.org/listinfo/discuss
>>>>>>>
>>>>>>>
>>>>>>
>>>>>> _______________________________________________
>>>>>> discuss mailing list
>>>>>> discuss at lists.oshwa.org
>>>>>> http://lists.oshwa.org/listinfo/discuss
>>>>>>
>>>>>>
>>>>>
>>>>> _______________________________________________
>>>>> discuss mailing list
>>>>> discuss at lists.oshwa.org
>>>>> http://lists.oshwa.org/listinfo/discuss
>>>>>
>>>>>
>>>>
>>>> _______________________________________________
>>>> discuss mailing list
>>>> discuss at lists.oshwa.org
>>>> http://lists.oshwa.org/listinfo/discuss
>>>>
>>>>
>>> _______________________________________________
>>> discuss mailing list
>>> discuss at lists.oshwa.org
>>> http://lists.oshwa.org/listinfo/discuss
>>>
>>>
>>>
>>> _______________________________________________
>>> discuss mailing list
>>> discuss at lists.oshwa.org
>>> http://lists.oshwa.org/listinfo/discuss
>>>
>>>
>> _______________________________________________
>> discuss mailing list
>> discuss at lists.oshwa.org
>> http://lists.oshwa.org/listinfo/discuss
>>
>>
>>
>> _______________________________________________
>> discuss mailing list
>> discuss at lists.oshwa.org
>> http://lists.oshwa.org/listinfo/discuss
>>
>>
>
> _______________________________________________
> discuss mailing list
> discuss at lists.oshwa.org
> http://lists.oshwa.org/listinfo/discuss
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.oshwa.org/pipermail/discuss/attachments/20150227/835a2926/attachment-0001.html>