Web Development

Programming high functioning websites is an intriguing exercise, as it involves the use of many different programming languages and formats, simultaneously. I’ve written several pages recently with no fewer than four distinct programming languages/syntaxes: html, php, javascript, and sql. And this is not an uncommon situation in web programming. PHP is a widely used server side scripting language (program executed at the server that is triggered by the http request from the client), html is, of course, the standard language to get client side browsers to display any sort of content, javascript is the most common client side scripting language (program sent with the html to be executed by the client computer on their system, nowadays usually under high security scrutiny by the client browser program), and SQL is the language to communicate with databases.

And, of course, that’s just the web page files. To extend the system usefulness into wider reaching applications, I’ve used server side scripts to write files that interact with separately running python and/or bash scripts. I haven’t yet written an application that uses the php to write a python script which is executed by a crontab command and writes a new php file and then a bash script and executes it to open a browser and send the new php file an HTTP request, which triggers it to write a…. and so on, but it’s not outside the realm of possibility.

I might make it sound a little ridiculous that there is such language diversity in these common and commonly interacting web development languages, but it does serve a purpose.  It helps to separate the various subsystems involved in the process.  If there was only one language involved, then there would have to be some extra syntax added to separate out the client side from the server, and the server and client from the databases.  But that extra syntax wouldn’t be too hard to find an elegant solution for, especially with an object oriented language.

Rock Crushing And Plastic Recycling

Currently, there is a huge demand for mechanically crushed rock, and a huge excess of mixed plastic waste.

Mechanically crushed rocks grip together when compressed, unlike smooth river stones and cobbles that slip apart. This makes them a crucial engineering material that adds strength and longevity to concrete mixes and the engineered soil surfaces which are prepared beneath foundations and roads. Unfortunately, it takes huge amounts of energy to break the stones apart.

Mixed plastics are our poorly sorted recycling waste. In order to produce high quality recycled plastic from mixed plastic, more sorting and lots of energy are required. We in the US have a lot of this waste because China has recently stopped taking it.

Can we take the mixed plastic, melt it down, combine it with a source of existing sand, form it into shapes that mimics or improves upon crushed stone and in so doing. produce a synthetic crushed stone? I think it is an idea worth exploring. You see, the quality of the plastic is likely to be compromised because it is unsorted, but the produced synthstone can be quality sorted by attempting to press it through a steel grate with gaps that are smaller than the desired stone size. The product that breaks and is pushed through the grate is rejected and either reprocessed or used for some non-structural fill.

Such production of synthetic stone could be incorporated into existing crushed stone production process, by using fines which are washed away from clear crushed stone as a waste product in place of the sand, and the synthetic stone could then be blended with standard crushed stone.

I’m not an expert on plastic manufacturing, but my first guess approach would be to heat sand up to the minimal mixed plastic melting temperature in a large rotating barrel, then introduce chunked mixed plastic, and rotate until a synthstone dough is formed. Then load the dough into a die extruder and produce the desired mechanical shapes by slicing a die formed extrusion. I’d call this the synthstone gnocci process, if it was effective.

I never do stock picks.

Largely because I don’t spend a lot of time researching stocks. But as I have just reached a milestone in my work project utilizing NVidia Jetson Nano, and just ordered a personal Jetson nano to build a home workstation, I decided to check out the stock. The P/E is at over 50 now, after a lot of stock gains this year (no doubt resulting from the market anticipating what I’ve learned practically).

Such a high P/E suggests that the stock is over-valued, but I see incredible long term profitability for NVidia, as their GPUs overtake CPUs as the dominant tool for massively parallel computing power, whether it is cloud based or within local desktops and laptops. And with the growth of distributed home voice recognition systems (such as Alexa and OK Google), all of which require massively parallel cloud computing with machine learning, we can only expect more demand for NVidia products, and as those home assistant systems are upgraded to offer more and more capability, long term demand is assured.

Openscad vs. Solidworks

I’m familiar with two 3d modeling programs; openscad (pronounced open-skad by the truely enlightened, and open-ess-cad by the technically correct) and solidworks. In openscad, one constructs 3d models programmatically, by typing in structured text functions to define shapes and extrusions and using loops for repeated features. This makes for fast and easy 3d modeling of simple structures. The same can be accomplished in solidworks, but the shapes are defined by cuts and extrusions of sketches which are graphically drawn on defined planes. This can also be quite fast.

Overall, If all one needs is a 3d model or stl file, I would recommend openscad in cases where all of the model is easily referenced to the origin, and I would recommend Solidworks for all cases in which the features are more easily defined based upon each other’s surfaces, especially when those may need to be adjusted later and the changes need to be carried through to all connected bodies.

If one needs to create drawings from the models, Solidworks is the clear choice, with easy dimensioning, automatic section cutting, and external views, it is a huge help for mechanical drawing generation.

Dream Distro: Xen Debian Wrapped Windows (DubDubOS?)

There are a lot of linux distributions and I’ve used and tried out many of them over the years. I started with ubunto, then I got into elementaryOS, then centOS, debian, qubes, and then back to debian. I’ve tried parrotOS, openSUSE, and tails, but never used any of those as a main system. My current favorite is debian with gnome due to speed and compatibility with a huge pool of precompiled software packages.

What I want to talk about here is what my dream linux distribution would be like. What I want is a system which grants access to all of the software of windows and debian. Qubes can do this, of course, but is more secure and less fast than what I’m after. What I’m looking for is a distribution that loads a xen hyperviser with a debian dom0 system and a single windows hvm. The distro installs xen and the debian dom0 with desktop flavor of the users choice, then sets up an HVM and prompts the user to load the windows install disk, and installs that.

The key to the distro is to control software packages to keep the virtual system from breaking, and to optimize memory and cpu use balancing to insure the best performance (though this is primarily a goal of the xen developers). The distro must provide fast and simple screen switching from dom0 debian to hvm windows, and permit maximal hardware usability in the hvm and dom0. Finally, the dom0 home directory should be shared to the hvm as a mountable drive. Does the HVM have to be Windows? No, but that should be the main goal of the project, as it is the greatest need of the user.

Watt Ave Might As Well Be The Wild West

A huge road, three lanes on both sides from I80 to 50, passing right through densely populated residential and commercial areas, Watt Ave sees heavy traffic every day with the high speed limits only controlled by traffic congestion.

In addition to all this, large sections of Watt have no lighting (where it passes through residential areas) and the road largely lacks surveillance.

Given the recent eggy attack on my person from a moving vehicle, it’s obvious that having the high speed limit road running next to pedestrian walkway poses a potentially lethal risk to pedestrians. I’m in favor, therefore, of reducing the maximum speed limit to 30 mph on Watt from 50 to I80 and installing roadway surveillance with license plate capture around the intersections, or, alternatively, installing protective barricades between the road and walkways.

Sulfur hexaflouride is the news.

SF6 is known in the utility industry as a dielectric material, but what is meant by that is that it is a material that can withstand high electric fields. Air can withstand electric fields around 30kV/cm, with some variance due to humidity and pollution, and SF6 can withstand 3 times that. Another common dielectric media in the high voltage industry is mineral oil which has a dielectric strength of up to 5 times that of air.

SF6 is mainly used in circuit breakers and gas insulated substations. SF6 permits these devices to be much smaller and lighter. Air insulated high voltage circuit breakers are theoretically possible, but would be massive and would perform differently depending upon weather conditions. Mineral oil circuit breakers were commonly installed before the invention of SF6 breakers, but these devices had to be large and heavy (10-20 times heavier than their SF6 counterparts), and had to contain a lot of heavy flammable oil. These characteristics that I’m describing are for dead tank circuit breakers, which have an outer shell that is grounded. Circuit breakers can also be “live tank” which permits a much lighter design. These are more difficult to maintain and require separate current transformer instruments which add more cost, but are not nearly as much as the difference between a live tank oil breaker and a dead tank oil breaker.

Gas insulated substations have all of the equipment of an air insulated substation but can be built much smaller in areas like cities that don’t have the necessary clearances for an open air station. Something similar to GIS can be done with mineral oil, but the weight, and therefore the necessary structural strength and cost increase by large factors.

In theory, it is possible to start to move away from SF6 technologies and return to mineral oil, but replacement of all of the currently in service SF6 devices would take decades and cost hundreds of millions to billions of dollars. The most cost effective method would be to use live tank oil circuit breakers, which would require less improvements to civil works like foundations and would require the least additional oil containment and fire wall additions. The problem is that this technology is not really being built anymore, as the industry has moved to the more efficient SF6 devices.

If one wanted to go with something new, nitrogen gas insulation is a reasonable alternative, Nitrogen is only a little stronger than air as a dielectric, but if it is in an enclosed tank with high purity, then none of the variances that would occur in open air would need to be accounted for and a reasonably sized design, but still much larger and more expensive than SF6, could be developed.

Dear my kind of people:

We have been attacked by strangers in ways which have no reasonable explanation from our own actions, as many people have. Naturally, we interpret these attacks to be acts of psychosis or intentional evil, and we treat the perpetrators according to their actions.

We may consider that these attacks could be a result of twisted versions of our history, some kind of slander, but we cannot, in a stranger, tell if they are attacking because they have been manipulated by slander or because they are insane, or are otherwise motivated for evil.

My friends, we treat people with kindness and respect by default. That applies to everyone we meet. Any conflicts which arise between us and other people are entirely a result of unprovoked attacks from them against us or others, but aggressors will never know that if they approach the problem of finding the truth by attacking us. What’s more, it is our preference to avoid conflict, so we do not seek out these sorts of conflicts nor try to exacerbate or prolong them.

We just can’t help people that demonstrate themselves to be evil to us, and it doesn’t make sense for us to have to explain that part of it. There is no logical reason for us to have to say this, or for people to not treat other people as if they all also believe this same thing. It’s basic humaning, and it’s also the golden rule if you’re into that sort of thing.