GitHub'ing
Index by images
1,0

See GitHub'ing (and At GitHub) to watch progress.
For each of these, see the README.md and Rlog.md files.

We'll post in this area for a while as we organize our thoughts about next steps and a lot more. Such as, meet deadlines. Such as, now the priority is getting The Gardner Annals, Vol. V, No. 1 into a draft form. Then, establish a process that is related to all content (not just code). We'll see. The below will be push down not unlike when I used Concrete 5 for our What's New for an extended period.

I'm going to start a list of issues to discuss (ala Quora or similar) related to software as it is now (a product of what it was) and how it will be. Huge task? Not really. It's necessary in terms of truth engineering.

1. Open versus cloaked. Of course, the former is seen here on GitHub to the extent of public access; one can see all of the details. Public implies, too, repositories that can be seen by anyone. There are private ones, however those working thus are using the same tool set which is an important element to understanding. The latter? Any of those who offer services, say web presence development, through means that are opaque. I bounced off of those in 2012 (this is documented and will be part of the summary).
2. Affectation. What? That related to computing? Has to do with expectations and their mismanagement which is a common theme. Somewhat, this issue was softened a bit throught the disruptive assumption. Yes, top-down inforcing of change for no reason other than to line the pockets of some, albeit with glowing reviews on some sides while others (many) suffer (we'll point to more than the gig economy's failings of the populace).
3. Posture always comes before the act. Somewhat, like the spark of the neuro-world.
4. There's a reason that the old guy lags behind those rushing forward. For one thing, the maddening crowd rushes after what's hot at the moment. Some get rewarded. Guess what? Many get punished for things outside of their control and knowledge. Truth engineering has a whole new realm to poke into, now, courtesy of this article (Surviving Software Dependencies) by Russ Cox at Google. Case in point is the Equifax fiasco. But, there are many more. Russ makes good points about being safe. Guess what? They may use, carefully; but, they rewrite; they, as in Google. Roll-your-own. I would say, like we had this saying in the 60s, 'God grows his own' (very much an analog). Here is an example of code that is pulled that Russ uses as an example: sindresorhus/escape-string-regexp. Russ mentions an inhibitor to doing things right, which I have written on (post at the truth engineering blog). Cost. Yes, if we did this stuff right, we wouldn't have the excess power for things like trying to show that DL is AI. We would be proving, minute to minute.
5. ...




--- blog ----

09/21/2019 -- Got reminded, today, of a paradox that is apropos: Moravec's paradox. TGS, Inc. uses the domains associated with people and their lives as the framework for evaluating AI and other advanced methods.

09/20/2019 -- Finally, digging into OpenAI. Have read of it, seen references, and such. Now, looking into it with my own eyes. AI --> AGI, thereabout. More later. ... Forgot, this is Musk's deal. Time to do a scan of the different views. OpenAI uses 'human will' which is a nice thought. I'll write more elsewhere. ... Oh yes, OpenAI's thing writes so well that they don't want to release it. Still, William Blake. Keep that in mind. ... Final thought, tonight: other views. Nice demo of what the Internet brings. Also, OpenAI said that they would be more cooperative than competitive. I just used it as a search term.

09/18/2019 -- New infrastructure. Too much fun.

09/15/2019 -- Saw that there had been discussion about a 'version manifest' and its use, such as for showing what is in a binary (executable). Truth engineeing is the context. That techniques deals with judging software in its various modes. Then, we need to deal with general knowledge, people, and a huge gamut of stuff. ... Post this post, got email notice of a build failure. First one. Didn't do anything different, so environment (wherever) or spook or ... Okay, worked. Only difference is that I lazily took the default comment on the change. Now, I know about possible errors and debugging.

09/12/2019 -- Would be interesting to see what 'AI' is being done by GitHub within the domain of it responsibility.

09/07/2019 -- Need to put in a reference to what got me, finally, to GitHub. It was reading this (Artificial Intelligence: A Modern Approach), again, and seeing the particulars with respect to aima-javascript which uses a blog nestled at GitHub. That was one month ago. I got lost in running around GitHub. The focus is (will be) this: ML/DL is not AL which is the larger domain. The brain is not the key thing; your gut is more involved with your survival. ... So we need to see how older AI fits in (lots of ways) plus there is the demonstration of little javascript in action (aimacode). Their code was at aima.cs.berkeley.edu/code.html. -- incidently, still hacking HTML, looking for a good editor

09/02/2019 -- The future? Here is a nice thought (Will Git ever be replaced?). Data and code were split for various reasons, however, as many themes have shown, these two are inextrictably woven. I have written, too: code is a mere side-effect of logic. It is that code and data represent jointly; it is that which is 'represented' that we are sorely ignoring.

08/29/2019 -- Watched the video, today, announcing Actions (Live Events). Will be commenting on several of the things discussed, eventually, but, initially, my great joy in running into GitHub (or, rather, opening the door to the experience) was seeing all of these capabilities integrated and accessible. Granted, there are private repositories which I will have, to boot. But, the openness plus the file format carrying forward just sits well with what I have been doing (or thought that we should do) for decades. Too, the whole framework reinforces my thoughts that JS has more to it than was recognized by those doing 'real' code. This work demonstrates the future, in many ways. Also, want to know? I immediately see how to use this experience to talk truth engineering and to discuss underlying issues that go way beyond what computationally oriented minds like to consider. ... later ... run these on your own hardware (or VM or ...) for free (31:04)? now you're talking ... it just keeps getting better. I, definitely, will be taking this discussion to Quora. Example: Is JS useful in artificial intelligence?.

08/28/2019 -- Earlier, I signed up for testing the workflow manager being developed by GitHub (GitHub Actions) and got notice of access, today. After a quick look, we will be getting to this in the next couple of days. So far, and it has been a couple of weeks, GitHub has been great. Having the 'open' option for poking around (and, perhaps, contributing) cannot be over emphasized. Software, as a part of life, does reign large. Needs hardware, though. Too, interpretion belongs to us, the mere humans.

08/23/2019 -- Lots of distraction, for the coming week or so. Have to get the summer ended with proper enthusiasm. So, may be quiet for a bit.

08/22/2019 -- Answer before reading Bejan: Why should I study economics? Bejan via NIH: The constructal law of design and evolution in nature. Two of my answers referring to Bejan: Why is economics split between macroeconomics and microeconomics? (classical) and Why do a lot of Go developers vehemently dislike JavaScript developers? (computing) Comment after reading Robert Smith: https://www.quora.com/share/Why-should-I-study-economics-10.

08/20/2019 -- I am using hidden variables for parameters. Will switch the template. However, the other day, on getting a file from the existing site and doing a modification, the menu handler didn't work on GitHub's pages. So, I hacked the code; on second thought, remembered the parameters. So, just changed that. Difference? Commenting out and changing a couple lines of code versus ... one parameter value. BTW, tackling understanding of DL will be grasping with parameters. We forgot the long trek from wires to abstract'd methods. All we have to do is recover that process.

08/17/2019 -- Have been reading discussions about JS and AI. Some see it as a travesty. Myself, if we can build a common platform for people to experience AI's effects, that is a step forward. Not unlike, from what I have seen so far, the motives talked about long ago with the Lisp Machine, albeit we were way too early with regard to communications and other technology. ... In this regard, I am working on getting the code organized, GitHub'd, plus get some of the auxilliary capability used, especially those things related to project management and collaboration.

08/16/2019 -- Looked at JS; yes, it's been extended to be general, including supporting mobile devices. Some wags say, mimicking browsers. Whatever. Looked at the code management situations; several other alternatives, however some of these actually just extend GitHub. ... Truth engineering is it, folks. Not hacking code. That is not the essence. Time we learn what differentiates us from whatever AI comes about. Okay? ... During this period, I'll get a process started after thoroughly reviewing all of the ins and outs of the important issues and packages offered, including psyching out motivations. Yes, we're non-profit; with, I might add, a new interpretation (about time?).





... little floater down here to avoid truncation ...