[FRIAM] Automata with FFT

[Steve Smith]

On 9/27/22 7:40 AM, glen wrote:
> Ha! The way you phrase it makes the guilty verdict sound more serious 
> than it was. But the point about problem vs solution space is sound, 
> especially given the *forward* map implied by "design". And thanks for 
> the summary of Alexander and motivations for the publications, all 
> stuff I didn't know.
>
> But I'm piqued by the assertion that the link to geometry was tenuous. 
> Salingaros' progression from master of geometric algebra to pattern 
> language guru *must* have something to do with a geometric gestalt in 
> or *around* pattern languages. 
I had some early correspondence with Salingaros as I tried to get my 
head around the broader implications of PLs and Alexander's *A*PL in 
particular and I felt the same then.  He definitely has a geometric bent 
in his perspective.
> I admit it could only be circumstantial, peri-pattern-language. But if 
> we look at Penrose (building physical "machines" as a kid), it seems 
> clear that embodied-in-4-dimensions can have occult implications for 
> the brain farts in later life. 
an early philosophical reading I did on "computer graphics", probably 
written in the early 70s (I cannot find a proper reference and have 
looked for decades now) stated that the (anthropic argument) reason we 
live in a 3 dimensional world is that  is the lowest dimension in which 
an arbitrary network/graph can be layed out without edge-crossings.  
This has huge implications for Edwin Abbot Abbot's Flatlanders as well 
as Bob Forward's Cheela (Dragon's Egg, neutron-star-surface creatura).
> So if Alexander's 1st love was art, which is inherently geometric, 
> wouldn't his adoption of the application of pattern language be a 
> reasonable correlation?
Most visual artists would probably argue that their work is not 
inherently geometric.   I'll elaborate in the next slab of lard.
>
> Maybe there's some ambiguity around the word "geometry"? 
I think both visual (flat, wall-hanging, etc) art and architecture are 
intrinsically embedded in a *spatial* context which is most commonly 
(but not exclusively) described geometrically. 
Painting/Sketching/Drawing/Photography/Printmaking/etc. adn.  tend to be 
made up of 0,1,2D marks on a surface for sure, but the *import* of those 
marks is quite often much more relational. Composition is the easiest 
perhaps to express/grok... and yes, the *geometric* composition (what is 
next-to/above/occluded-by, etc) is the expression, but the *relations* 
being expressed are in an entirely different domain, and the *mapping* 
from one to the other is essentially metaphorical, though the 
metaphorical source domains (e.g. religious, political, sociological... 
) are often layered and rich in their structure, but definitely more 
"topological" than "geometric"?
> Sorry for being skeptical from my completely ignorant position. But 
> can you explain why you claim the connection is tenuous?
I hope DaveW gives his own contrasting/complementary answer and 
hopefully others here have their own trick dogs in this circus?


- Steve

>
> On 9/26/22 19:47, Prof David West wrote:
>> Alexander was a Janus: a mathematician at his father's insistence 
>> when he wanted to be an artist. An architect by compromise.  Face two 
>> was a Taoist mystic infused with hard core Catholic fundamentalism.
>>
>> His Ph.D. thesis—which became his first book, /Notes on the Synthesis 
>> of Form/—was an attempt to define a mathematical science of 
>> [architectural/industrial] design. But in the same book, he stated 
>> that optimal design arose from a "non-selfconscious" process, 
>> embedded in myth and ritual and culture.
>>
>> /A Pattern Language/, was part of a trilogy that included /The 
>> Timeless Way of Building/ and the /Oregon Experiment/. /APL/ was 
>> written by committee and edited by Alexander (although he took all 
>> the credit) to fulfill a government grant. His mystical side was 
>> front and center in /TTW/; and the Oregon Experiment was a case study.
>>
>> Alexander transcended Patterns and his last major work—/The Nature of 
>> Order, vol 1-4/—centered 15 generative properties that have little to 
>> nothing to do with patterns and is far more mystical and Catholic-God 
>> focused than his earlier work.
>>
>> Ward Cunningham and Kent Beck brought /APL/ to the attention of the 
>> software community as a workshop at OOPSLA (ACM conference on Object 
>> Oriented Programming Systems and Applications). The 'Gang of Four' 
>> authors of /Pattern Languages of Programming/ participated in that 
>> workshop. A year after their book was published a mock trial of the 
>> GoF for "heresy" was staged and they were found guilty.
>>
>> Perhaps the most significant error made by the software community was 
>> seeking patterns in "solution space" rather than "problem space;" the 
>> latter being where most of Alexander's work was focused. The software 
>> patterns community looked at written programs to find multiple 
>> instances of similar bodies of code and attempt to discern a 
>> generalized problem that they solved (albeit with contextual 
>> idiosyncrasies).
>>
>> There are hundreds of thousands of software patterns published, but 
>> maybe three or four that actually capable of being applied in 
>> multiple contexts—of actually being considered "true" patterns.
>>
>> The connection to geometry, both in Alexander and in software 
>> patterns, was never more than tenuous. A majority of the patterns in 
>> APL (e.g., "Dancing in the Streets," "Sleeping in Public") had 
>> nothing to do with geometry or any other mathematical formalism. Even 
>> patterns like "Light from Two Sides" are geometric in the only the 
>> simplest sense.
>>
>> The math in /Notes/ was algebra, not geometry. Only in his last major 
>> work NO, can you find properties that are overtly geometric, e.g., 
>> "centers" and "alternating repetition."
>>
>> more upon request
>>
>> davew
>>
>>
>> On Mon, Sep 26, 2022, at 5:13 PM, glen wrote:
>>  > I'd appreciate you (and SteveS) throwing some words at it. In
>>  > particular, since software patterns are *supposed* to be linked to 
>> the
>>  > geometric patterns of architecture, *where* or *how* has it gone 
>> wrong
>>  > in extrapolation? Did Alexander go wrong in his extrapolation? Or did
>>  > others [mis]interpret?
>>  >
>>  > (I've purposefully left the Subject the same because it definitely
>>  > relates to Chan's morphology based taxonomy and my argument with my
>>  > meso-biologist friend about "species diversity" versus "phylogenetic
>>  > diversity".)
>>  >
>>  > On 9/26/22 15:35, Prof David West wrote:
>>  >> I am a patterns and Alexander expert. glen's uncertainty / mild 
>> antipathy is spot on. Software patterns are an oxymoron.
>>  >>
>>  >> Strong words, but happy to back them up with dozens of papers 
>> written/presented and hours of discussion.
>>  >>
>>  >> davew
>>  >>
>>  >>
>>  >> On Mon, Sep 26, 2022, at 6:29 AM, glen wrote:
>>  >>> Very cool! Thanks.
>>  >>>
>>  >>> In particular, our property abuts "the ravine", which is a 
>> semi-wild
>>  >>> place. The permaculture categories might help me orient my own
>>  >>> intuition (that everything in the ravine should be left alone) 
>> with my
>>  >>> neighbor's (clearing the whole area and reintroducing natives). 
>> He owns
>>  >>> the majority of it. So, c'est la vie ... or perhaps "telle est la
>>  >>> mort". (Don't blame me. I don't know French.) One thing this 
>> zone 0-5
>>  >>> model might permit is modularity. That blog post implies such 
>> with the
>>  >>> inverted garden interface. But it seems like there could be 
>> pockets of
>>  >>> zone0es in wild areas and pockets of zone5s in urban areas,
>>  >>> particularly in sprawling cities like LA or Houston. Growing up in
>>  >>> Houston, where every square inch of semi-abandoned land seemed 
>> rapidly
>>  >>> reclaimed by the swamp, is probably the source of my skepticism 
>> with my
>>  >>> friends' diversity doctrine.
>>  >>>
>>  >>> There's a lot to digest in the biophilia links. I have to 
>> confess, I
>>  >>> haven't given pattern languages much attention. It always seems
>>  >>> motivated by geometry, which fails for me. Of course, I'm familiar
>>  >>> enough with software patterns. But that's always failed for me 
>> as well.
>>  >>> They seem too ephemeral, unstable ... i.e. not real, convenient
>>  >>> fiction, and *perfect* opportunity for gurus to blind others 
>> with their
>>  >>> gobbledygook mouth sounds. I guess it reminds me of category 
>> theory,
>>  >>> too abstract for my ape brain. But maybe some of his earlier 
>> work on
>>  >>> Clifford algebras might motivate me? I could start here, I guess:
>>  >>> https://link.springer.com/chapter/10.1007/978-1-4757-1472-2_41 
>> <https://link.springer.com/chapter/10.1007/978-1-4757-1472-2_41>
>>  >>>
>>  >>> Thanks again.
>>  >>>
>>  >>> On 9/24/22 10:29, Steve Smith wrote:
>>  >>>>
>>  >>>> On 9/24/22 9:49 AM, glen wrote:
>>  >>>>> Such efforts seem so inherently metaphorical it's difficult 
>> for me to approach a concrete conversation. For example, I have a 
>> couple of biologist friends, one meso (bugs) and one macro 
>> (ungulates), who thought I was being contrarian when I challenged 
>> their assertion that biodiversity in urban areas was *obviously* 
>> lower than that of natural areas like forests. Of course, I admit my 
>> ignorance up front. Maybe they are. But it's just not obvious to me.
>>  >>>>
>>  >>>> This may seem a little tangential but the realm of Permaculture 
>> Design has a suite of truisms on these topics, though they are 
>> articulated in their unique language which can be a little hard to 
>> translate sometimes.  I think the permaculture community represent a 
>> fertile laboratory for doing *some* experiments as implied by Glen's 
>> questions.
>>  >>>>
>>  >>>> A good example which gestures toward the Chan work at least 
>> morphologically is maybe worth a scan if not a full read here:
>>  >>>>
>>  >>>> 
>> https://aflorestanova.wordpress.com/2016/04/08/zones-in-permaculture-design/ 
>> <https://aflorestanova.wordpress.com/2016/04/08/zones-in-permaculture-design/> 
>>
>>  >>>>
>>  >>>> Permaculture's 5 zone quantization doesn't preclude a 
>> recognition of there being continuous gradients in many dimensions 
>> from a locus of "technological closed-loop" (zone 0) and "biological 
>> closed loop" (zone 5).
>>  >>>>
>>  >>>> There is a *lot* of talk in the literature about the interfaces 
>> around zone 0, 1, 2 techno-structures creating localized ecozones 
>> that harbor diversity (desired and undesired == vermin) which I think 
>> provide some good anecdotal evidence about biodiversity in transition 
>> zones and acute technological interfaces (e.g. roofs, walls, corners, 
>> posts, fences, etc).  Permaculture is a domain of recognizing and 
>> exploiting "happy accidents".
>>  >>>>
>>  >>>> It is also worth noting the diversity spike that happens in 
>> estuarial contexts...
>>  >>>>
>>  >>>> A more formal study of Urban/Architectural design with an eye 
>> to *health* (human-centric view) is the domain of Biophilic Design 
>> <https://www.terrapinbrightgreen.com/report/biophilia-healing-environments/ 
>> <https://www.terrapinbrightgreen.com/report/biophilia-healing-environments/>>. 
>> Nikos Salingaros is a hard-core Mathematician at UT-San Antonio who 
>> addresses abstractions of Complexity 
>> <https://en.wikipedia.org/wiki/Nikos_Salingaros#Complexity 
>> <https://en.wikipedia.org/wiki/Nikos_Salingaros#Complexity>> and 
>> Pattern Languages <https://en.wikipedia.org/wiki/Pattern_language 
>> <https://en.wikipedia.org/wiki/Pattern_language>> as well as 
>> Architecture and Urbanism.  He also has some interesting opinions 
>> <https://en.wikipedia.org/wiki/Nikos_Salingaros#Philosophy 
>> <https://en.wikipedia.org/wiki/Nikos_Salingaros#Philosophy>> about 
>> post modernism as well as Dawkins Atheism.
>>  >>>>
>>  >>>>
>>  >>>>>
>>  >>>>> Since then, they've presented (meso and macro) arguments that 
>> justify their position. It does seem obvious that urban areas trend 
>> to more adaptable animals like coyotes and raccoons and less so to, 
>> say, deer. The bugs are more interesting. Meso guy found some 
>> articles that show "species" diversity in urban areas is roughly the 
>> same as natural areas. But phylogenetic diversity is clearly lower in 
>> urban areas. That seems counter intuitive to me. It's a cool result.
>>  >>>>>
>>  >>>>> My main point when I originally expressed skepticism, though, 
>> was about microbial diversity. Is it possible that bug-layer and 
>> microbe-layer (including what lives in/on large animals like rats and 
>> humans) diversity makes up for lower diversity in large-layers?
>>  >>>>>
>>  >>>>> I *feel* that projects like Chan's could help with this 
>> question since it seems prohibitively expensive to sample and test 
>> enough microbial populations of urban and wild areas, especially if 
>> we include intra-animal populations. I'm just not sure *how* they 
>> could help.
>>  >>>>>
>>  >>>>> On 9/24/22 03:38, David Eric Smith wrote:
>>  >>>>>> It’s funny; I know Bert.
>>  >>>>>>
>>  >>>>>> One of our colleagues played a role in bringing him out to 
>> work at Google in Tokyo.
>>  >>>>>>
>>  >>>>>> A mathematician (Will Cavendish) who has part-time support at 
>> IAS
>>  >>>>>> https://www.ias.edu/scholars/will-cavendish 
>> <https://www.ias.edu/scholars/will-cavendish> 
>> <https://www.ias.edu/scholars/will-cavendish 
>> <https://www.ias.edu/scholars/will-cavendish>>
>>  >>>>>> is also interested in the mathematical dimensions of this, 
>> though I have only a glancing exposure to how those two together are 
>> trying to frame the problems.  Because Bert has come at it more from 
>> the ALife/engineering approach, and Will’s interests run more in the 
>> direction of proving capabilities of broad classes of systems, often 
>> interested in their aggregation as categories  (and also about the 
>> role of simulation as a replacement for proof in systems that produce 
>> complicated enough state spaces), it should be a productive and 
>> interesting collaboration.  I don’t know how engaged others are in 
>> the Google group on this specific project, because I am too far 
>> outside that loop.
>>  >>>>>>
>>  >>>>>> Eric
>