another little 4-layer board

This is worth reading.

Given that people are making chips with hundreds of CPUs, it's time we had a new OS architecture. Something reliable for a change.

John Larkin Highland Tech Glen Canyon Design Center Lunatic Fringe Electronics

Bill Sloman snipped-for-privacy@IEEE.org wrote: |-------------------------------------------------------------------------| |"[. . .] give them what they think they want, rather than | |what they need." | |-------------------------------------------------------------------------|

Ex-bosses who professionally pretend to be engineers override engineering advice by me that what they should want is different than what they say. They used to prefer liars who used to tell them what they want to hear instead of what they needed to hear. They went out of business.

|-------------------------------------------------------------------------| |"[. . .] Plausible nonsense has been responsible for quite a lot of death| |and misery over the years," | |-------------------------------------------------------------------------|

Alas true.

|-------------------------------------------------------------------------| |"when the US educati I started asking my students to examine seven different elements of an article. If the information checks out on each of these points, it has a high likelihood of being accurate. Still, passing the test is not a guarantee that it’s fact.

1. Copyright: I always ask students to check the bottom of the webpage to see if the information has been submitted for ownership.
2. Verification with multiple sources: Students must double check the information on a few different web pages. Like in a trial, the more corroborating witnesses, the more likely the truth will be discovered.
3. Credibility of source: I tell them to check if the source has been recently created. Sources that have been around for a while can show reliability over time and be tested by hindsight, whereas recently created sources don’t carry much of a track record.
4. Date published: I always ask them to check how recently the page was updated to see how current the information is and whether anything has changed.
5. Author’s expertise and background with the subject: Students should check if the author is someone who has dedicated time and effort to learning this subject. For example, a university professor typically has increased credibility versus a hobbyist.
6. Does it match your prior knowledge: I ask them if the information matches up with what they have learned before.
7. Does it seem realistic: I tell students to use their common sense. Does something seem authentic or probable?

Of course, even Bedley admits that this method isn’t fool-proof. Websites can falsify copyright dates, for example. They can also be written by someone with higher degrees but who, upon investigation, is not a credible source.

Take Dr. Jason Reza Jorjani, who received his philosophy degree from Stony Brook University. He presents his beliefs as research, and according to some of his colleagues, he is a charismatic person. However, just because he has a PhD does not necessarily mean that he is a credible and reliable source of information. [. . .]

References Allcott, H., & Gentzkow, M. (2017). Social media and fake news in the

(S.

John Larkin snipped-for-privacy@glen--canyon.com wrote: |----------------------------------------------------| |"Has anyone designed a new OS in the last 35 years?"| |----------------------------------------------------|

Do BeOS or Inferno or new Commodore-Amiga OSes or Plan 9 or video-game consoles' operating sysyems or a lame new IBM thin computing offering or new versions of Acorn RISC OS count? (S.

Don Y snipped-for-privacy@foo.invalid wrote: |----------------------------------------------------------------------| |"In industry, the attitude seems to be to just resign yourself to the | |capabilities of whatever you are "stuck with" -- because deciding what| |you actually *need* is too much effort (easier to see what you can do | |with what you've "got"). This is not true in academia. But, how many| |practitioners are actually aware of current -- and past -- research | |and trends... they likely consider themselves "practicing experts" | |(despite the poor empirical quality of their codebases)." | |----------------------------------------------------------------------|

I used to work as a researcher in a putative university, under a so-called tutor who was surprised that I installed Plan 9 (which he heard of only because of me), who ordered me to use Microsoft Windows.

Two putative researchers in a putative university interviewed me. They acted as if Microsoft Windows and Linux are the only operating systems.

A researcher at a research conference was surprised that I called GNU/Linux "GNU/Linux" instead of "Linux", because he is accustomed to researchers who dare to call GNU/Linux "Linux". (S.

John Larkin snipped-for-privacy@glen--canyon.com wrote: |-----------------------------------------------------------------------| |"Given that people are making chips with hundreds of CPUs, it's time we| |had a new OS architecture. Something reliable for a change." | |-----------------------------------------------------------------------|

Who will buy it? Who will refuse to buy Microsoft megascoria? How many posts became made to news:alt.destroy.microsoft in the last 3 years? (S.

Sure. US manufacturers become non-competitive, the US economy goes down the tubes, and Canada takes over the northern state of the US and Mexico takes over the southern states. Electing a fantasist like Trump as president is one of the aspects of the mechanism.

Magicians don't have to explain what they are doing. Neither did the people who designed the legacy electronics that I got to clean up. By the time I was doing the clean-ups we had design reviews for main stream projects, and I'd got into habit of documenting my designs as if they were going to be reviewed.

You don't draw particularly easily intelligible schematics. They should be designed in a way that helps the viewer work out what is going on, with signal paths basically running from left to right and from bottom to top, with the exceptions carefully labelled and high-lighted by extra white space around them.

None of it stops you building a dog house, but it should make you more careful about scaling up the design. You do need to think about what you are doing.

That would depend on what you were doing and how "practical" the expected results. If the goal is to create something of immediate application, then there is pressure on using an "established" set of tools soas to allow for others to replicate (or deploy) your results.

OTOH, Computer *Science* deals with addressing and exploring fundamentals that can advance the field by noting deficiencies in it or the way by which it is practiced.

You can introduce FETs as alternatives to BJTs and, in a couple of product cycles see them widely accepted. In a relatively short timeframe.

Trying to retool software systems to exploit new technology is considerably harder owing to the scale of the problem. (Imagine retrofitting FETs in place of BJTs in every EXISTING product under the assumption that redesigning all of them to use FETs would be very time consuming and labor intensive)

Add to that the observation that folks routinely notice new *devices* in trade magazines and can/do request samples to "play with". What publications are being consulted for new software technologies? How easy to "tinker" with them in a real application?

To most people, that's likely the case. Forget the fact that Android/iOS run on more machines than any other OS. Or, that "none of the above" runs in the dozen CPUs that are used in your desktop PC (mouse, keyboard, disk drive, NIC, etc.). Or, the 50 ECUs in your vehicle. Toaster, microwave oven, stovetop, refrigerator, television, HiFi, DVD player/STB/VCR, etc.

Sadly, it is a small application segment that drives a huge part of the technology base. Notice how much Linux and Windows have been bastardized to address various features that aren't inherently part of their actual designs but "bolt-on" components that bastardize the design (the cathedral vs. pyramid argument, up-thread).

If you *need* features X, Y and Z, why bolt them onto something as an afterthought instead of designing with that in mind, from the start?

It's the same naive idea that leads to products with s***ty security because security was bolted on and not inherent in the original design (so, likely plagued with latent bugs).

If you're doing the same thing in the same way, over and over, why are you surprised that your quality and process aren't improving? Who's making the technological decisions -- a bean counter? (Do you even know WHY your designs are weak?)

It is very easy to QUICKLY become "outdated" in engineering fields. This is particularly true if your education was "current" and, thus, outdated before degreed.

For folks who "rise through the ranks", there is less time and drive to remain current with the technology that their hires are expected to master. Its rare to find an employer/supervisor who is "up" with the skillsets of those he employs/supervises (though one can *pretend* as long as not called to account)