sponaugle

There is no power over fiber.

"I never understand why anyone needs anything faster than 1G".

i just upgraded my home network from 40g to 100g. g, not m. Anyone who does video knows the pain of 10g.

"(running fibre in the house isn't practical yet)"

What does that mean? I assume you are saying most builders don't offer fiber?

Running fiber in a house is entirely practical and easy to do. Cheap as well.

I ran fiber when I built my house a few years ago. I have MANY 12 fiber runs going to my IDF/MDFs, Server room, the gate, the dmarc, etc. I also ran SM fiber to every AP location. Out of 20 miles of pulls about 1 mile was fiber.

This is VERY true - especially for home use. You could have many db of loss and still have a perfectly functional 100g network.

I have a $1000 fusion splicer from Amazon, and it is fantastic. It is very straightforward to use, and for my home use it is perfect. If I did spicing as a job this wouldn't be the tool to use, but for occasional splices in my homelab it works great.

https://www.amazon.com/dp/B0CW2HGVT2?ref_=ppx_hzsearch_conn_...

"While I agree with the sentiments, there is a societal need for gatekeepers in some professions. The engineer title often comes with legal liabilities. It’s not necessarily just about talent. Of course it often becomes misused as well."

I fixed this for you.

"I disagree with the sentiments, there is no societal need for gatekeepers in most professions. The engineer title rarly comes with legal liabilities. It’s not necessarily just about gatekeeping. Of course it often becomes misused as well."

That is a good estimate.

You can see US non profit compensation online - https://datarepublican.com/nonprofit/assets/

Many of the presidents/vps make in that exact range (~$200 - $300k), although there are other like:

https://projects.propublica.org/nonprofits/organizations/363...

Feeding America, which pays their CEO a tad under $1m.

It is ridiculous that your registrar was going to act on your domain without any kind of verification of jurisdiction. Can anyone just email a registrar and claim there is some injunction and have something removed?

"The uplink, in front of a directional VSAT antenna, this is very significant."

This is reasonable, and I fully can understand that direct transmission at reasonable power going into a sensitive receiver can be problematic. In practice I suspect it would require some particularly aggressive intent to aim it such that you see that condition - and of course at the point you are at intent the 'rules' are optional.

I fully understand the risk of highly directional radiation - but that is why I was asking if there was some actual specific testing done. I believe the software in the Starlink (and the beamforming) is designed to only transmit when it knows where it is pointing ( it uses receiver information and ephemeris data to know where and how to transmit ).

I would suspect that if the dish was instead pointing directly at someone with no direct satellite reception there would not be high transmit power.

Again, of course it could be catastrophic in unheard of ways, but is there any evidence of that as a real outcome, potential and likely outcome, or even likely theoretical outcome?

I can understand it takes time for legislation to catch up, although that is far more a 'choice' than a requirement.

I'm an electrical engineer - I don't understand how what you said can be true at a wide and effective scale. Of course any device can interfere with something else - But what kind of evidence exists that starlink terminals, and in particular the pattern steering and bands use, actually cause interference in a meaningful way to activly used marine system used on these kinds of crafts.

Starlink used Ku and Ka bands, so we are talking about 10-12Ghz and 37-42Ghz downlinks, and 14Ghz and 47-51Ghz uplink. This is highly directional radiation, and even with beam forming and reflections is not likely to have any specific high interference risk. Clearly the downlink transmission from the satellite is exceptionally low power at the ground, even with the beamforming used in the satellites.

I would love to learn that this is a real problem with actual tested data.

The wikipedia page for the Tristan da Cuhna island is really interesting. It is amazing that a small population was able to survive there over such a large period of time with just occasional ship access (sometimes years apart).

https://en.wikipedia.org/wiki/Tristan_da_Cunha

The entire population was evacuated in 1961 due to the eruption of the mountain on the island, but most people made it back to the island in 1963.

Also interesting that just recently got Starlink internet.

Very cool.

Interesting - I suppose it depends on how you count things. I have more than 100 smart light switches in my house, so if you add in smart plugs, sensors, cameras, etc I certainly have more than 300 'smart devices'. That has no effect on if the house is somehow less or more horrid, at least from my, my wife, and my daughters point of view ( I asked them ). Perhaps all three of us are weird however. ;)

Ha! It is a funny thing.. one time I was trying to remember something specific about a calibration table in a Subaru ECU and I googled it and found a great answer.. and sure enough it was something I posted years before. Ah the beauty of forgetting!

"I count my remaining years in much shorter spans. I do not look back and say, I should have spent more time working and traveling."

I specifically waited to have children until a bit later because of work/startup travel. Time goes in one direction and there is a giant locomotive at the end waiting to smash into you.

"That's not everyone of course, but that level of CTO with his hand in the bits just seems constantly dangerous."

The biggest danger of a CTO being too deeply involved at a technical level is it does not scale. Knowledge and deep understanding helps you scale, but actually writing production code that has to be maintained puts a burden on scale.

Of course at a startup that might be different, especially in the beginning, but the faster you get someone working for you that is even better than you at the job the faster you will be better at your future job.

"However, what I struggle with a lot is that there's absolutely no manual for this role. Everything is my problem, yet nothing is my problem."

For me, that last sentence is the most empowering. The essence of the CTO role is the ultimate responsibility for 'technology', yet you are often not the actual leader that controls the technology. It is one of the hardest concepts to make work because it relies on your own deeper understanding of the technology AND the people that are trying to use it. You are constantly trying to predict what the outcome will be so you can decided what to influence, which is like pushing on slushy water.

"Failures are concentrated (blameless post-mortems mean that it's the culture or environment that's wrong, guess who is responsible for the environment!) and successes are amortised (credit should go to the team who are hands on, of course)."

That is a really good compact statement about the crossover between credit and responsibility - and it really pushes hard that post-mortems are important. I have said more than once at a post-mortem - "Just to start us out, all of the failures are my fault, so lets leave the arrows pointing at me and dig in". Sometimes that is the only way to get people to open up and focus on the truths and not worry about the arrow. Just remember that managing upward is a different task.

"someone who is technically very curious."

By far the most useful quality in a CTO, but I'm a CTO and have said that exact thing so I'm biased.

I have met hundreds of CTOs in my career so far. It is surprising how quickly I can tell if the person is curious or not, and how often that one attribute ends up being a marker of someone who is extraordinary.

I'm in year.. let's see.. 24 of being a CTO. I always enjoy reading what other people see as important enough to write about. The conversation about shipping code is something most technical CTOs feel, and my solution to that 'need' has been to continue to explore my deeper technical 'must be building' in other parts of my life.

I have a small Youtube channel that follows a few different technical paths, and I keep another array of deeper technical challenges in the development pipe always ready to go. I still write code, design hardware, build cars, and look at stars, but they are all self driven projects.

I find the result is a tighter focus at my core job of technology strategy and the bigger decisions but also not a loss in my core technical abilities. To me the T in CTO means technical, and I think it is my responsibility to deeply understand the technology that our teams use. That understanding however can come from personal learning just as much from being in the production loop. Be a nerd.

Food for thought, at least for me.

2? I have 12 APs in my house. 5 ghz propagation is terrible through insulated walls, especially those with metal film and 5/8's drywall. Add the green stuff and a closed door and APs multiply.

I like it. I use Ubiquiti gear for most of my homelab and home networking stuff with the exception of the 40g and 100g stuff. It is both easy to use and provision as well as easy to upgrade… and of course it works.

Sometimes people do really ridiculous stuff with their homelabs, like this guy: https://youtu.be/-b3t37SIyBs?si=t_h49fv594dChYf9

Tongue-in-cheek on that of course. ;)

Nice... I did that same problem as it was a numberphile episode. Very cool! Fortunately I have over a PB of local storage at home which helps out.

I have a local copy of 10 trillion digits of Pi on the math cluster, so no web access needed.

Indeed, the comparison length is 90% one hit, 99% two hits, etc.

My current method is O( (n/h)^2 ) where h is the hash size, which is really O(n^2) for n>>h. I'll have to think about this to see if it would be better.

I believe they do.

https://blogs.sas.com/content/iml/2015/03/12/digits-of-pi.ht...

I need to noodle on this. Hmm.

1f:ca7:5ee5:b12d::ca7:eat5:b12d

While Pi is not proven to be a disjunctive number, nor the stronger condition of being normal, it is generally believed to be normal. That being said we don't have a proof of Pi being normal, nor disjunctive.

I am not familiar with how a proof of that would be constructed, as clearly numerical or computational measurements could never be conclusive.