daly

Quoting Dr Griffin (Jan 8, 2022):

We are clearly seeing with this Omicron wave vaccines really do work to protect people against severe disease.

(Review of a published paper): Risk Factors for severe outcomes among persons aged greater than 18 years who completed a primary covid vaccination series..

The authors looked at 1,228,664 persons who completed primary vaccination for Covid-19. They defined this as 2 MRNA shots or 1 J&J shot. They reported that severe Covid-19 associated cases were rare and only 0.015%, death occured in only 0.0033%. Risk factors for severe outcomes included age greater than 65, immuno-suppression, and about 6 other underlying conditions.

All persons with severe outcomes had at least 1 risk factor and 78% of them who died had at least 4 risk factors.

So I want to try to put this fear in perspective. Prior to vaccines we were seeing 10-20% of people who got infected ended up in hospital and we had a case fatality rate of about 2%.

So we went from a 1 in 5 chance of hospitalization to about 1 in 10,000 with vaccination. Chance of death from 1 in 50 to about 1 in 30,000 with vaccination.

Sadly it appears that he has no expertise in Virology. He gets the medical facts wrong. Perhaps he should spend time listening to TWIV (This Week in Virology). Or perhaps he could listen to Virologists in THEIR area of expertise. I fear the Supreme Court suffers from the same defect.

People seem to be confused about the vaccine. It works like a smoke alarm. It helps your body DETECT and RESPOND to the virus. It does NOT PREVENT a virus attack, just like a smoke alarm does NOT prevent fires. The result of early detection is that your body can reduce the damage caused by the virus.

This is a scientific question. Visit TWIV (https://www.microbe.tv/twiv/). Today (Jan 8, 2022) there is a blog post by Dr. Daniel Griffin (TWiV 850: COVID-19 clinical update #96 with Dr. Daniel Griffin). He comes on once a week and gives details of the science and updates for clinical treatment. This is his 96th weekly update.

Vincent Racaniello (https://www.youtube.com/channel/UCyFgCoP4ovsHbt92vM4zN2A) has a free online course on Virology, which I can highly recommend. His 2 volume textbook is also available.

If nothing else, a proper education will enable you to recognize when lawyers are confused about the medical facts.

Whether a mandate is legal is a political question. I have no opinion on that.

Whether a mandate is good virology is a scientific question. From what I've learned over the last few years, I believe it is. Don't take my word for it. Get educated.

No, "do your own research is NOT that simple".

Research is hard. It is time and life consuming. It is also more fun than, as the saying goes, a "barrel (lab-full?) of monkeys".

The idea is that each SRP unit makes 2 copies of itself and then moves off to perform another task, such as finding raw materials.

By this "power of 2" you get exponential growth.

At the mechanical level, 3D printing and robots allow self-construction.

At the biological level, plants like sunflowers can take up things like zinc, copper, and radioactive elements. (https://pubmed.ncbi.nlm.nih.gov/22567714/) If we can figure out how this is done perhaps we can find a genetic basis. If so, we could engineer ants that would naturally mine materials.

If nothing else, at least we can cover Mars in paperclips :-) (https://en.wikipedia.org/wiki/Instrumental_convergence)

In any case, I'm happy to see that NASA is interesting in terraforming Mars. Maybe my research will find it to be a practical application.

Unlike Bitcoin it is used by all countries. I'm not sure that all of the distributed Bitcoin journals can handle the volume. Imagine your Bitcoin journal trying to clear a (U.S.) billion transactions a day.

I disagree with your premise. I write open source. Tooling isn't a problem.

I want an 8.5x11 device with large-size microUSB. Nothing fancy, no markup, no net connection, huge battery. Simple page turning (could be touch or button). Simple file handling. Memory of where I stopped reading.

You could sell these things in volume, the first venue being libraries and bookstores.

Students with school library access could use them to carry all of their papers. Profs could use them for research.

Sci-hub would be a perfect research match.

https://www.amazon.com/Overshoot-Ecological-Basis-Revolution...

One key problem is the need for "fiducial oversight". That is, a grant organization needs to know there is a professional accountant who manages the funds and receipts.

Rich Hickey can do the open source world a favor by sponsoring a couple full-time funds management "fiducial oversight" professionals (aka an accounting firm). These 2 or 3 accountants could handle grant fund management for a large range of open source projects.

Block grants could be given and distributed to multiple open source projects based on their grant proposals.

Universities have this. The only grant I got was at a university. The provost took about 55% "for support" but I did get my research funded.

Perhaps somebody who knows somebody who knows Rich could pass this along.

After 20 years of maintaining open source software on multiple projects and getting $0 dollars of support I applaud the attempt to reward these professionals.

Building and maintaining "personal infrastructure" to do open source "at scale" is expensive in time (github, continuous testing, test suites, pull requests, email, DMs, phone support, etc.) and equipment (development computer, router, network connection, AWS, backups, etc.)

In addition, if your project gains traction you'll get invited to give talks. These can sometimes be overseas (yeah, I know, nobody travels anymore). The time to prepare a talk and the travel cost is another personal expense. The most I ever got from a talk was a free hoodie. But I paid for the round-trip airfare, rental car, 3 days at a hotel, and most of my meals.

On the project I've maintained for the last 20 years I estimate that I've spent about $3000 per year. That should be the "baseline" for supporting an open source project.

Having spent a decade or more in "cyber" related activities I think it is vital that the director have a deep understanding of how attack and defense ACTUALLY works. This is a very complex, fast changing topic and you need to be able to distinguish the money-seekers from the actual effective proposals.

You'd need a lot of time with the military to understand their issues. You'd need a lot of time with industry to understand their issues. You'd need a lot of time with the blackhat community to understand their issues. You'd also have to fight "stove-pipes" that try to keep their information secret (e.g. 0-days).

You would almost need to create a "cyber immune system". It would have fast response teams (i.e. antibodies), "reserve units" able to help overwhelmed fast response teams (i.e. macrophages), "tactical units" that have expertise in specialty areas, such as air-gap attacks, cache-poisoning, and row-hammering. You'd need a wide network of organized reporting (i.e. a lymphatic system). You'd need specialty consultants on subjects like LoRa exfiltration (i.e. t-cells) and a large library like the one at CMU with a collection of malware (i.e. memory t-cells).

The scale of the problem alone makes it difficult.

Sure it is sustainable, at least as long as I can work on it. I've been at it 20 years so far. I still work on it as a research project.

I was also the maintainer of Magnus, an open source Infinite Group Theory system. That project died when I left the job. The new person who took over stopped working on it. So that project wasn't sustainable.

So, to your question "Is Open Source Sustainable?"... that depends on the maintainer.

What YOU mean by "sustainable" might be different from what I mean, however.

Are you asking if maintainers / projects can get funding? The answer is no (modulo a few exceptions). I've tried DARPA. I've tried direct company requests. I've tried a contribution fund. I've tried direct requests to funding organizations of various sorts. I've tried the Google Summer of Code (got a yes on day 1, got dropped on day 2).

One of the issues is that funding, such as grants, require "fiducial management". That is, somebody has to handle funds and collect receipts. I asked a large company to set up a very small organization that had some accountants who would manage the grants for open source projects, including mine. That also failed.

Suggestion: Set up an open source account management firm with fiducial accountants. Ask large companies that use open source to give grants to the firm to support the projects they use. Make the funding public, including who contributed and who got funds.

Are there costs involved in running Axiom? Sure. I maintain my own servers (paid for by me). I pay for the network connection monthly. I have (in the past) gone to conferences. I have prepared and given talks at conferences worldwide. I have Axiom t-shirts, cups, glasses, a hoodie, and stickers. I have created and distributed CDs. I have self-published a book. I have done "remote help" for several people for free. On average Axiom has costs me about $3000 / year, not counting my time. So, doing the math, it has cost me $60,000 over 20 years.

Open source (and Free Software) is NOT free-as-in-money.

Russia needs Ukraine. China needs Taiwan. Both locations are perfect "control areas of the battlespace". Any military commander would recommend an attack.

The real issue would be direct force-on-force confrontations between Russia/US and/or China/US. So far the superpowers have avoided this and worked through proxy fights. Even in Syria the US was VERY careful to avoid Russian troops.

However, in Taiwan, there are US service personnel. I don't know but I suspect there are also US "advisors" in the Ukraine. Direct superpower attacks won't end well.

The situation will get wildly out of control VERY quickly.

Given that Wolfram Research is tightly held, what happens when the company ends? Does it "punch a hole" in computational mathematics? Does all of your code stop working?

This isn't idle speculation. I used to use a Symbolics machine. Symbolics Inc. failed. The Symbolics software still exists but you can only buy a copy that runs on DOS. I used Common Lisp on Symbolics. Common Lisp is a standard. There are open source implementations so all of my code survived.

Will your code survive when Wolfram Research dies?

(Full Disclosure, I'm the lead developer on open source Axiom)

You'll see that we can't feed 9bn people.

My reasoning is that I don't understand why I would move from a subject I know really well (programming) to one that I don't know at all (management).

Plus I know I'm neither good at self-management nor good with people so wisdom seems to dictate remaining a programmer.

Note that all of the former programmers all claimed they wanted to either return-to or continue-to program. One of them even had it written into his contract that he could program "in his spare time at work" (who has that?). Not one of the managers returned to programming later in life.

Programming is hard, a lot of self-inflicted pain, and the ultimate source of frustration. Few people want to return to that once they have "escaped".

Know thyself.

Spy communication from inside a nuclear facility.

However, retirement is amazing. I spend most of my days on my own projects, which usually involves taking several courses on youtube. For instance, I'm working on a problem that requires a camera so I'm watching "The Ancient Secrets of Computer Vision" (https://www.youtube.com/watch?v=z5WSV6CXsxs&list=PLjMXczUzEY...). Nobody would pay me to learn how to compute the robots grasping pose of a 3D printed object from its GCODE description.

Retirement is great fun and the perfect opportunity to spend 18 hours a day, 7 days a week (modulo random naps :-) ) hacking on whatever software that strikes your fancy. Enjoy it while it lasts.

The Power Grid: Last Week Tonight with John Oliver (HBO)

https://www.youtube.com/watch?v=qBpiXcyB7wU

The database included several newly developed "stored procedures".

Time elapsed... and it was nearing the time to ship the code. So we tried to populate the database. But we could not. It turned out that the stored procedures would only allow a single record in the database.

Since a portion of the "business logic" depended on the stored procedures... well, things got "delayed" for quite a while... and we ended up having a major re-design of the back end.

Fun times.

The original Amazon Kindle DX was so very close.