10 Responses to “Ask Nanodot: Open Sourcing Nanotechnology”

1. redbird Says:
   September 1st, 2000 at 1:35 PM

   General comments

   Before I start, just let it be known that I have only had time to skim through the draft in about five or ten minutes. Later I will go back and read it all, but especially since it is just a draft I don't want to spoil the final version for myself.

   I found the use of the word common misleading. Again, I may need to read more closely, but common, in my mind, is too closely associated with the Tragedy of the Commons. Remember that since IP can be reproduced freely by anyone (i.e. no need for a field for ideas to live in), it is not so much a common as a network of individuals propagating memes. Maybe meme space would work better? (but then that requires the description of memetics)

   Secondly, maybe the notion that nanotech can mean the end to scarcity in most senses (so long as the universe in finite, there is some ultimate scarcity thanks to entropy) should be brought up. It might be helpful for people to see that ideas already lack scarcity by showing how matter can, too. Again, sorry if you mentioned this and I missed it.

   Finally, I just want to write that I think this is going to be a great paper. I know that we've been having these sorts of debates on nanodot, so hopefully you were able to use them to gain some insight (even if they eventually turned into intelectual flamewars). I think that some of the best arguments I've read (or posted) in favor of open source have come out of responses to post strongly in favor of proprietary control of nanotech.

2. RobertNansel Says:
   September 1st, 2000 at 4:50 PM

   Open Sourcing MNT & Homebrew STMs

   Here's a couple pages on building homebrew STMs: Simple STM Page Homebrew STM Page

3. BryanBruns Says:
   September 1st, 2000 at 9:40 PM

   1. Thanks 2. Antipatents

   Thanks for the link to the Simple STM Project. I hadn't seen it. Looks like the guy knows what he's talking about. Amazing that he's aiming for under $100. – Bryan

   In connection with one of the other questions and issues in the paper, Slashdot has a new discussion related to opening up potentially patentable intellectual property "What happens when patents meet antipatents."

   The idea comes from a new proposal (September 1) on Transparent Patents. Rebecca Hargrave and Carl Malamud suggests a database of antipatents: "Antipatents are simple, a registration mechanism for your open-source inventions." The essay has good ideas and lots of links to further sources. An essay on "Patents vs Antipatents" describes their proposal and seems to have been the source for Slashdot, but the original article isn't that long and has more content.

4. prion Says:
   September 2nd, 2000 at 7:41 AM

   Once again

   Please make a distinction between "Open Source" and "Free Software" as RMS defines it. They are just not the same thing, and they are more and more confused together.

5. BryanBruns Says:
   September 2nd, 2000 at 3:39 PM

   GPL, LGPL, and other open source

   The paper includes a section on licensing, discussing different kinds of "Open Source" software, with bullet points on the key characteristics from the Open Source Definition and a table that tries to highlight the differences between:

   1. public domain
   2. BSD, X11, MIT, Python and other unrestrictive licenses
   3. MPL, LGPL and other licenses which require disclosure of modifications but which permit incorporation into "combined works" with closed software
   4. The GPL which only permits use with other "Free Software."

   Licensing is a complicated and confusing subject which I'm still trying to understand, and I would appreciate comments and corrections on any points that I may have misunderstood or not presented clearly.

   The (draft) recommendation in the paper is that the Free Software Foundation's Lesser GNU Public License (LGPL) may be the most appropriate license for nanotech, allowing (modular) combination with closed commercial software while promoting disclosure of modifications.

6. redbird Says:
   September 2nd, 2000 at 7:18 PM

   The nature of the LGPL

   The LGPL is never appropriate for anything (disclaimer: I, the classic hypocrit, have licensed modules under the LGPL)! A strong statement, true, but let me explain. The LGPL is very much a transitional tool, and should be considered as such. It allows for the meme and benefits of free software to propogate, but does not allow for true freedom. Only the GPL provides true freedom (at least, in terms of current copyright laws) by not promoting non free software. Until enough people beleive in the GPL, the LGPL and other, week open source licenses will serve to transition them. I don't know how far reaching in chronological terms this paper is supposed to be, so for now the LGPL may be fine, then move to the GPL, and finaly go to having no copything law at all (which is something nanotech developers can't diretly change, but a lofty goal). This, too, depends on how free you, Bryan (after all, it is your paper), want nanotech, software, and just about anything affected by copyright to be.

7. BryanBruns Says:
   September 3rd, 2000 at 12:01 AM

   LGPL as a useful compromise

   The LGPL is a compromise, which may be one of its key virtues. In an earlier draft I'd thought of weighing more towards the MPL. However the LGPL has the advantage of being compatible with the GPL, so that intellectual property (IP) under the LGPL can be used under the GPL. By contrast the MPL is not compatible with the GPL in that way (or at least not until the MPL is either revised or dual/triple licensed in accordance with the intentions that Mozilla has stated but not yet implemented)

   The LGPL, as you know, allows combination with closed commercial software. I agree with the ideas of the Open Software Initiative and others on the advantages of a license which is friendly and attractive to commercial involvement. Many companies and many individuals do not want the required "true freedom" (they sometimes call it "tainting") of the GPL, and find the GPL exclusive and unattractive. And that is not an accident, but the consequence of the deliberate design of the GPL (what you refer to as "not promoting non free software"). This issue is a big part of what led to the Open Software definition. I suggest that a more inclusive strategy is also suitable for nanotech.

   Furthermore, my paper tries to address both software and hardware, in the near to medium term . I think that, at least for a good while, nanotech is going to involve the same kind of mixing of open and closed IP as occurs in computers. This is true almost by definition, given the current and expectable roles of software, proprietary hardware and other technologies in nanotech development. If mixing open and closed IP is likely to be inevitable for nanotech, then the same sort of practical arguments which underly the creation and use of the LGPL would apply for nanotech. That includes the pragmatic benefits of being able to work within existing systems and the priorities for use of scarce resources (particularly programmer time in Free Software/Open Source communities).

   However, if only to stimulate continued discussion on what I think is a significant point, it's worth noting that the LGPL not only might facilitate a transition toward the GPL, but possibly also movement the other way. It allows what could be useful competition between open and closed modules, while laying out groundrules about having to disclose APIs, etc. To me, this seems another potential advantage, albeit one not likely to be highly valued by those opposed to commercial software. There could be choice and competion among various plug-ins and among other components, depending on what works best.

   The LGPL satisfies the Open Software Definition, while being tolerable to (though, as you clearly express, not the first preference of) Free Software proponents. The debates about licensing are confusing and discouraging to many, so some consensus could be useful. Using the LGPL could also reduce the potential incompatibilities of a proliferation of different open source licenses with slightly different terms, e.g. different companies' open licenses. Personally, I have some views about why the simplicity of unrestrictive licenses may be preferable to the hassles and putative benefits of the ones which require disclosure, and I may add a bit on that in a later draft. What seems much more important though, would be at least trying for a rough consensus on one clear, widely acceptable standard, and so, IMHO, the LGPL seems the best candidate for licensing intended to share intellectual property for nanotech.

8. BryanBruns Says:
   September 4th, 2000 at 7:58 PM

   LGPL as Cinderella

   LPGL as Cinderella, or, The Quest to Share the Source

   Her family knew she was the Lesser one.
   declaimed long lecture on the virtues of the Elder,
   and why none but the Greater one deserved
   to be your only true choice.

   As for the others,
   the unFree ones who claimed to be of Open mind,
   why they knew she was of that fervent kind,
   and so not suited in scrupulous specification their wares to bind.

   But in the latter days, minstrels told of how,
   the camp of Moz the Red came to laud the sisters' beauty.
   (And if some thought the Lesser fairer,
   why perhaps this was only whispered,
   lest such heresy disrupt the wooing.)
   And much was spoken of how all ought have license
   to wander at liberty, among the inner realms of the Open and the Free.

   And then the minstrels relayed the tale
   that even the keenest of the delving elves
   now pledged fealty to the Eldest sister of the Free,
   jousting with the gnomish usurpers for favor in each release.

   And all the while the trolls, as was ever their wont,
   found much of which grumble,
   and didst incessantly strive
   to veil the lands in fuddish confusion.

   Still, many were the requirements,
   exacting and essential,
   which inspired the quest
   to release the Source for all to share.

   Naysayers said not too narrow,
   never restricted so neatly
   as to fit only the tight confines of the Free,
   in haughty exclusivity.

   And the minstrels told of muttering that many who
   raised the banner of the ram had motives mercenary,
   pursuing plunder from lands beyond the Free,
   complaints that, in outland endeavors, aid in arms and armor
   was only advanced upon golden tribute to the few.

   Fashion flaunted its fickle fads
   against broadly styled designs,
   frowning on excesses of generosity,
   forsaking the wisdom of serpentine serendipity.

   All agreed on glass, transparency
   at the invoking of origination,
   when the glory of source
   was unveiled in splendid glare.

   And most felt clarity must bind each contributor
   in the same enchanted spell, for all who dare
   in the open to declare, to forestall foul deceiver,
   and the rise of selfish rivals who might not share.

   Many said that only a shapely vessel would do,
   no bloody chunks ripped out, to strew
   in butcherous abandon, greedily.
   No, though works might be combined,
   they should be linked entire, modulated
   with visible display of appropriate proper integrity.

   And long the searching, so many lost hope,
   falling by the wayside in baffled despair,
   with faith too frustrated to renew.
   The minstrels sang of how clashes and flames
   drove off many who might have joined the quest to share.

   Many seekers clung to clumsy convention,
   quarrelled over creaky compromise,
   or carved their own crude likenesses,
   caring little what conflict might ensue.

   But still some strove, until one day,
   after long questing, some came to see that the Lesser one,
   so long disparaged and neglected in her shabby niche,
   might be the fairest of them all.

   And thus might be told
   the tale of how she was found,
   the Lesser Princess of Gracious Liberty.

   Bryan Bruns, September 5, 2000

   Coda
   I fear in time's harsh glare she may not prove
   half so fair as I might dream,
   but in this light, for now, she outshines all others.
   And tis a pleasant tale to tell.
   For who can yet know, how far her domains might grow.

9. BryanBruns Says:
   September 8th, 2000 at 5:00 AM

   Update on answers so far

   First, thanks to all who have made comments: on the NanoCAD list, on Nanodot and directly. I'd like to share some of what's come up, and note what still seem to be some gaps. I've also posted a revised version of the paper, at the same URL:
   http://www.cm.ksc.co.th/~bruns/open_mnt .htm

   > 1. … pointers to good comparative reviews of [molecular modeling]
   > software, especially if they discuss licensing.

   The Kachinatech list is a useful starting point (which was already cited in the paper). While exploring lists and links works okay, it would be nice to know if there are any FAQs or introductory surveys with an overview of what's available. <http://sal.kachinatech.com/Z/2/index.shtml>

   > Similarly, I'd be interested in any reviews of the
   > current situation with open hardware and patent licensing.

   No comments on Open Hardware and open patent licensing beyond the links already in the paper. Slashdot had a discussion on the Transparent Patents article by Rebecca Hargrave and Carl Malamud which argues for improving availability of "prior art" and other approaches to improving review of patent applications. They also recommend arrangements for publishing "antipatents" which disclose information to prevent subsequent patenting. Similarly to the approach of open sourcing MNT discussed in the paper, they recommend a strategy that is not premised on abolishing or radically reforming the current intellectual property system. <http://voice.media.org/essays/patent.html>

   > 2. Are there industrial consortia currently active in nanoscience R&D?

   No consortia seem to be working on MNT-related efforts. No pointers so far to consortia working on other aspects of nanoscience.
   An interesting example of collaboration to support open source software is the Open Source Testing Lab recently announced with support from IBM, HP, Intel and NEC. <http://slashdot.org/article.pl?sid=00/08/30/0115259>

   > 3. What is the cost of an AFM?

   Expensive. "a DIY proximal probe microscope can cost 2 k$ upward. Batteries Not Included/Some Assembly Required" and better equipment costs far more.
   I'm hoping for some cost info from a packet that ThermoMicroscopes has sent.

   > What are the prospects for "homebrew" hobbyist access to MNT equipment?

   John Alexander's Simple STM Project aims to build a simple scanning tunneling microscope that can image individual atoms for under $100. It looks like he knows what he's doing, and this suggests that at least some kinds of hobbyist access may be likely, with all that could eventually imply. <http://www.geocities.com/spm_stm/>

   > 4. Would clearer ideas about potential system
   > architecture for MNT be useful, or is it too soon
   > for such discussion?

   [This question has stimulated interesting posts on the NanoCAD list.]
   Ideas about MNT systems architecture seem to be in an early stage of development. At the molecular level there are still disagreements about how and when it would be suitable to abstract to simpler approaches than detailed quantum calculations. At higher levels, massive parallelism and complexity pose major challenges beyond current software systems architectures. In terms of discussing possible architectures it would be useful to distinguish between fixed factory-type versus floating vat-type (or Utility Fog) designs. The distribution of computing capacity among different levels is an issue (which could be linked to to Merkle's proposal for a broadcast architecture). There is recognition that levels and modularity will be necessary, but so far few proposals analogous to such (once upon a time controversial) software abstractions such as high level programming languages, the Unix kernel or the internet transport layers. The ways in which architecture evolves could have major implications for whether MNT systems architectures tend toward being monolithic and proprietary, or towards open systems with modularity and levels encouraging competition and combinations of open source and closed source components.

   > 5. Have discussions about the Foresight MNT
   > guidelines addressed the possibilities of more
   > formal self-governance structures, e.g. along the
   > lines of the internet committees and requests for comments?

   No replies so far on this. The Guidelines invite comments. Two comments have been posted using the CritSuite commenting software, (though I couldn't access the comments, due to technical problems). http://crit.org/http://www.foresight.org/guidelines/current.html

   > 6. What could be done to make open source efforts
   > in molecular modeling software more successful?

   Better tools for collaboration, such as interacting on designs and coordinating contributions of software components. Maybe some kind of competition.

   As part of trying to promote more "design ahead" to be able to take advantage of molecular assemblers once they arrive, Robert Bradbury has proposed a Nano@home project, including options to encourage resulting designs be open source.

   Other topics:

   Two comments on Nanodot seemed to favor Free Software GPL licensing rather than other open source licenses. As discussed in the paper, the modularity and disclosure requirements of the LGPL might offer a compromise solution acceptable to many involved in MNT software initiatives.

   For those interested in more on the discussion, I encourage looking in the NanoCAD archives at the threads on "Open Sourcing Nanotechnology," "Collaboration" and "Vat manufacture."

   Thanks again – Bryan

10. BryanBruns Says:
    September 9th, 2000 at 3:31 PM

    Alternate URL (mirror)

    The local ISP which hosts my website is planning to relocate their office soon, and I will be out of town for a month and not in a position to do much if problems occur. In case it is useful, the file for "Open Sourcing Nanotechnology" is also available at:

    http://www.gedanken.org/bb/open_mnt.htm

Leave a Reply

Name (required)
Mail (required)
Website

Authenticate this comment using OpenID.