07/10/2021 -- Took a deep dive into Deep Learning's mathematics to get to the bottom of the issues. And, now will report this. As an aside, Alan Kay, pure mathematician and developer from way back - Smalltalk, weighed in on some current affairs with regard to what John McCarthy wrought. Stay tuned.
12/15/2020 -- With respect to Baez, forgot to add this: https://ncatlab.org/nlab/show/HomePage. Example all around of things that we would like to discuss, however the focus on category theory and computationally modes thereof more than inspires.
12/13/2020 -- Was actually doing code yesterday; had it in mind to come here for an update; got too busy with the issues; no one was watching except myself; so, my clock is not under any control; even reminders don't get the attention when the domain problem is heavily grabbing the focus; ...; oh, yes, on the interplay of Haskell (and its ilk) and Lisp (and its kind with all of the wannabes), the former is exactness enforcing while the latter is free as a bird (with the 'exact' being motivated by alg top's use).
12/11/2020 -- Lots going on. This (https://johncarlosbaez.wordpress.com/2020/10/19/epidemiological-modeling-with-structured-cospans/) got several things going. One was a stepping back to the original work around 2000 where the focus was intelligent support for design/build/maintain in which the main thrust was model properties that were significantly meeting several criteria. The bonus was that the computational aspects covered issues relate to manifolds (in the sense as adopted by algebraic topology). Baez has worked in category theory for years. His group has used Julia (language) and like it.
09/30/2020 -- Took a moment to assess the interest in Spinoza which seems to be apparent, everywhere. Found this: https://ethics.spinozism.org/.
09/18/2020 -- Will get back here if I remember. Turns out that Spinoza was a computational brain, the first one from 400 years ago. More later.
08/18/2020 -- Need to get back to how the current modes force closure inapropriately while ignoring the real limits of decidable modes, meaning not as strong as expected.
08/10/2020 -- ACM Comm articles are much relevant this month. One's on extending SQL to handle linear algebraic modes, natively. Nice. Another is one entity matchin which is a fundamental operation/requirement in data science and just about any computing using databsase. Like their use of in premise, cloud, and mobile to separate out the three major modes that we see now. There are others that we mmight look at. Another article considered the issues of repeatable, replicable, and reproducible in terms of CompSci as its failure influence science, in general.
07/10/2020 -- Notice, only one post per month. Will be picking up once some responsibilities that were picked up on the spring are met. Will do an app, no doubt, related to the theme of history of the U.S. and how it relates to today and tomorrow.
06/19/2020 -- Last day of Spring of 2020. Went fast for many reasons. Our look is to the future: https://thomasgardnerofsalem.blogspot.com/2020/06/the-future.html. And, how we can leverage the knowledge over 400+ years.
05/12/2020 -- The portal (to truth) has a nice, clean, simple feel. Will continue that, perhaps with some more js. Given that, we will be reviewing our presence in general with respect to the message to convey which will be reassessed after ten years of work.
04/15/2020 -- So many things to discuss, analyze. The wealth of possible research topics has never been so great, especially those of the inter-disciplinary type that will be so important to a sustainable future.
03/18/2020 -- Time flies for the older crowd. Do we take jostlings, like COVID-19, better? Remains to be seen how well we handle this disrupting event. This will be different, going forward. One benefit will be the opportunity to look from a lifted view.
03/09/2020 -- One inspiration: https://github.com/nayuki?tab=overview&from=2007-12-01&to=2007-12-31. Several aspects of this work appeals. One example: https://www.nayuki.io/page/project-euler-solutions. From: https://www.nayuki.io/page/about. Who knew that Comp Sci (the Nayuki's coder's college work, 2007-2012) could be so much fun?
03/06/2020 -- Missed all of February. Busy across several spheres. Actually, dived into the issues related to machine logic and support of this via assembler. I know, there's deeper, yet. Some think that patterns tame things down there; after all, physics is just that, in a sense. Those higher-order patterns several steps above the machine are far removed from anything physical, many times, except for the consideration that it's the lower that supports, and might even give life to, those dancers and their dances on the wider stage. One new thing is the cloud. So, considering issue there brings in newer dynamics albeit without anything being real revolutionary. ... Truth engineering is the focus. AI plays a role as does a lot more. ... Oh yes, with the mobile devices morphing along during the past decade, the technical aspects became a little more interesting, due, in part, to increased capability. Take the 'tether' as I have used (like a virtual leash - albeit the nose rather than the neck) it since Job's gift started getting attention (2007/8 timeframe). Lots and lots of differences, with the cloud rolling along symultaneously enabling the stack's evolution. Take AI. There are still lots and lots of things that we don't know about ourselves, since science cannot handle these type of things (not yet, will), that are going to change the landscape of computing at some point. How and when will be discussed. ... Oh yes, in that context, we are really seeing a new type of 'tower of babel' coming to fore.
01/26/2020 -- Need to describe the notions behind AI (ML or otherwise), in the following vein. Hats? Yes, right now, I working with an advanced system, as a user and not a developer. Helping out. As I use it, obvious tweaks come to mind, however I don't drive the requirements. Too, the system works well, for the most part. It's the domain that difficult, sometimes. What we don't want is for the computer aspects to add to the difficulty. Guess what? It has, too many times. So, rather than bite the proper bullet, the industry has rushed after mathematics as the savior, especially one aspect related to statistical means. Then, that led to the eventual machine learning examples which are merely a subset of what we need to think about. Hats, again. These change according to the modes and more that are of interest. Automation seems, to me, to be a peanut butter attempt. So, we'll pick this up, atain.
01/24/2020 -- In the larger context, we're within six years of the 250th of the U.S. start if you use the Revolution as initiated by the Declaration of Independence. That is, 2026. Of course, now, we seeing the Mayflower/Plymouth folk celebrate their ancestors arrival upon these shores. Massachusetts will have several of these. Four hundred ago, in 1626, there was a transfer from Cape Ann to the interior. That resulted in Salem, MA. We're within five years of King Henry III re-issuing the Magna Carta. This is a brief wander off of the road of futurism in order to look back and ponder.
12/29/2019 -- Services? A way to support your supposed client. Too, it opens the door to mischief. Rather than jumping into the technical details, I decided to look at the state of things. Okay, peer-to-peer is as old as the hills. Client to server is about as old, perhaps a little newer. But, the cloud allowed 'serverless' computing. Or, the illusion of such. Underneath, there are farms. And, one has to enter via a known protocol. The query needs to be formed, sent, and the results handled. Some work has been done on standardization. One hopes that this will allow us to extend the framework's conditions, in all senses.
12/27/2019 -- Oh yes, looking at 'services' which are offered via API. How many ways to entrap? Sorry about that, but the relaxation along the timeline has accumulated lots of negative potentials. Along with looking at this in terms of the 'floor' or 'wall' as mentioned many times, will be taking a closer look at jekyll's abilities. GitHub uses it, too. That's been on the plate since August; looks like there might be time available, finally, to get to it. Balancing a lot of hats might be fun, but it plays havoc with time. Time? Cutting corners? Want to really discuss this as it ought?
12/25/2019 -- Saw some outtakes from a conference in Madrid (on the telly, didn't have specifics) in which they quoted a couple of guys. One was enthused about AI. Almost giddy about the possibilities. The other cautioned that some things will be decades away. The other day, one researcher talked of the current AI being used to add numbers. What? Does that not seem wasteful? However, if we tried to determine rules for addition from such (say addition), that would be interesting to see. Otherwise, we could do a comparison of the hugely expensive (not just money, time, resources of various types) AI with our native ability. As arithmetic was worked out long ago in our evolution. This is one example (https://www.pluralsight.com/guides/deep-learning-model-add) where the researcher admits that it's not the ideal use of aNN, however we could learn. Yes, I might add, we have to also bring in intuitionism. And, I mean, in a sense not really considered to date (as far as I can see; why? too simple until we had this new reality to cope with?).
12/20/2019 -- Per usual, lots of hats being worn, tasks balanced, and slowly moving toward teleological ends (yes, multiple). No days off. After all, while we're alive, the heart and lungs and more work, constantly. So, was looking at backend alternatives, not wanting to queue up with one of those huges affairs. You know, all of the social media, and other, firms offer services (which is one manifestation of API'hood which goes back to RPIs). Actually, did a couple of those a few years ago while toying with Codecademy. Well, what do I see? Jekyllrb.com. Jekyll was one of the first bits of software that I looked at, but it was in the context of having a blog supported by files. I was happy getting into GitHub for several reasons, one of which was the development focus with people working in modes outside of the restrictions of GUI. Well, I did my own little website here on GitHub using my own method and put the Jekyll stuff on standby. So, what did I see, yesterday? Lots of companies using Open Source methods are using Jekyll (https://en.wikipedia.org/wiki/Jekyll_(software)) to boot. Interesting examples. Okay, so that then gets me to think of looking backend'ing here on GitHub, partially with Jekyll examples as a learning tool. That's on the plate where the fork is about ready to address thing for eating, so to speak. Oh yes, and templates are in the focus, too (https://jekyllrb.com/tutorials/convert-site-to-jekyll/). Okay, another example (https://www.smashingmagazine.com/2014/08/build-blog-jekyll-github-pages/) which discusses going WordPress to Jekyll. I see that I have been hacking my own thing the past decade. Remember, my bit: https://tgsoc.org.
12/18/2019 -- I spent a great deal of my computing years with highly-configured (high-end - Unix/RISC) workstations that were networked (geometric and numeric processes in support of design/analysis of large systems). Of course, network, early on, meant LAN, however we had T1 lines with several sites across the country (U.S.). So, the whole basis for the cloud/full-stack'ing, etc. is familiar. I was there early. In the case of the work for TGS, we're at that point, again, where I can dig in. My front-end focus (https://tgsoc.org) was due to knowing of the capability of most clients (not thin, at all). You know, thin client might be there, but, for the most part, mobile devices aren't tiny things (except for presentation). Well, yes, they can be. So, I'll continue with some front-end data handling. But, the direction is going back to the server end. Looking forward to the new focus. Of course, what do I see (well, take note of as I knew this for a while)? API access being offered everywhere. Why? Playing around? Folks, we need to discuss all of this crap. I don't see the necessary types of review being done. But, I might be missing something. We'll end up with a TGS API (preferrable, for now, to an app (we'll get there). So, I'll be blogging along while looking. Hosts have to juggle lots of players; same goes for the cloud'd variety. The server I am using doesn't like node.js (issue of tuning). I don't want to use Python (yet, lots of reasons, one of which is to see how far js can go). ... To be continued.
12/13/2019 -- About to the point where the portal to truth's front page is mostly generated from data via JS. The point is that a minima approach has a place in the new world. Complexity for its own sake sucks.
12/06/2019 -- Kode Vicious (George V. Neville-Neil) had a nice look at what he calls 'Koding Academies.' This was in his regular column in the ACM Communications (Dec 2019 issue). He used plumbing which I have, as well. These training sessions might help one be a front-end plumber, say going out to fix leaking pipes, etc. However, doing construction which entails planning through actually doing would require more training. So, not bad as an analogy. I liked that he mentioned html/css/java/... where '...' included a few more of the means to doing software for web sites (and their activity). The '...' list python. Interesting that he mentioned that. And, he notes that many would not include the group that he listed in computer languages. Lots to discuss there.
12/03/2019 -- 1/2 way through the recent The Atlantic. They've improved content as well as restructured their look. A balanced look at red/blue which applies to computing. Has to do with 'floor' (and walls). They quote the PhD Cambridge (from Idaho - who wrote of her hayseed folks). Yes, this: the difference is those who sit in the back with Uber and those who drive. I have not Uber'd. Don't really intend to. One old progressive mind, on being cajoled by his daughter to Uber (verb) and having talked to the driver, said to me: that is no life with future expectations. Of course he knows holding a full pension plus the accumulation allowed by pre-tax investment over the decades. Yes, indeed. All of those magic carpet riders (riding with their diapers as mom/pop carried them along to somewhere else than where they were) need to relook at themselves.
11/30/2019 -- The 400th ought to help get a necessity started. Namely, northern New England's importance in establishing what became the U.S. And, we are talking fifteen generations or so, even more when southern New England is brought in the equation.
11/23/2019 -- Time runs away from us. Trying to control by overloading, or even overlording, does not work. However, one method which has seriously been overlooked is handling cross-temporal matters. Saw interest in https://en.wikipedia.org/wiki/Time_crystal as something to look at in this regard.
11/20/2019 -- Well, it's here. GitHub can be used to represent the floor/core in the context of the spiral of content and configuration in the various spaces with which we are familiar. Needless to say, there are many more spaces; and, I am not referring to those of the abstracted realm behind the doings of data science (machine learning, et al). That is of the configuration side, presenting interesting views to below-the-floor stuff. No way does it capture the essence, as we need to do. ... Now, floor/core? I have a lot about this on Quora and will be coordinating (with myself) on getting this stuff more organized. Briefly, I need to introduce 'psychether' and how it pertains to truth engineering which is a serious requirement being overlooked in the rush down the roads to perdition.
11/19/2019 -- Coming soon. It'll parallel the old file. Perhaps, we'll drop text here and do the other modes in HTML/CSS/JS.
For prior posts, see https://jmswtlk.github.io/GitHub-ing/blog.html.