Hacker News new | past | comments | ask | show | jobs | submit | drfrank's comments login

Elon wrote: https://twitter.com/elonmusk/status/1649523985837686784

> 3 months ago, we started building a massive water-cooled, steel plate to go under the launch mount.

> Wasn’t ready in time & we wrongly thought, based on static fire data, that Fondag would make it through 1 launch.

To me, it sounds like people that would have been considered experts (and probably still are) made assessments than turned out to be incorrect.


Some times you just have to try. This was way bigger than anything that's gone before. I'm sure they'll have plenty of data now to make the launch pad even better than it would otherwise have been.

Videos of the debris storm were pretty mind boggling to watch - it's difficult to comprehend the sheer amount of energy expended.

I am truly looking forward to the learnings published from this.


Naively, it seems to me that there should at least be a warning somewhere if there are declarations after a REJECT-THE-REST.

I'm not familiar with whatever language this is, but wouldn't such a construct always indicate something was being ignored?


It seems this is a one-time opportunity for some engineers to catch up with inflation. It doesn't seem that Microsoft is adjusting it's salary ranges.

Put another way, starting to work at Microsoft, this year, will still be 11% less attractive than it was two years ago.

Microsoft seems to recognize (but not acknowledge) that it's not competitive in compensation, and seems to be reiterating its commitment to its current strategy.


Has an insurance market already arisen?

I.e. can I buy insurance for some percentage of my transaction against a bug swallowing the whole value of my transaction?

Seems like an opportunity to let an economic market drive more rigorous dev and test practices by giving shops with more rigorous practices lower fees.


yes, its pretty vibrant for the past 2 years. claims get made in advance and paid out unceremoniously with no fanfare (just like outside of the crypto space) but there could certainly be more fanfare and articles around that experience, since you can tell by this thread that people who should know about it don't know about it. You can find people in the discords of most communities who have the claims after an exploit who are like "yeah they paid" and move on, and you can look at the insurance protocol's contract to see it paid out for your own verification of that.

not sure if I've seen it for NFT collection contracts, but they're pretty robust and actively cater to the DeFi category of services.

here is a list of insurance protocols that also use a token for any number of reasons. its a whole category on the "marketcap" sites.

https://www.coingecko.com/en/categories/insurance

there are likely many well capitalized insurance protocols without a token but I don't know about them for that specific reason, which is a fun irony when wondering "but does this need a token?"


Yes. I'm not going to name drop for fear of it coming off as shilling, but I'm familiar with a couple DApps that are two-sided marketplaces for insurance on other DApps.

If you lose your money to a smart contract bug in the insured DApp, the insurers will reimburse you.

If you really trust the code of a certain app, you can earn revenue as the insurer on the other side of the trade.


If you want to influence the direction of the sequel to Star Control II currently under development it seems that you can do so on reddit: https://www.reddit.com/r/uqm2/


TFA implicitly conflates "leadership" and "management", suggesting that the author is oblivious to the distinction.

Even when the author describes dividing the responsibilities of "leadership" among multiple people, they describe dividing both leadership and management responsibilities between those people.

This is an authoritarian model that makes the local despot the bottleneck for the team: Decisions end up being routed through the "leader", individual growth and development and team value is constrained within the scope of the "leader".

These constrained teams have limited potential, and so, because of the leader's limited focus and the departure of team members who feel constrained, the organization develops cracks and holes in responsibility and ability to act on opportunities.

Managers who subscribe to this authoritarian model promote authoritarians, capable technical contributors who want more control over the part of the system they and their peers have been working in.

The best managers don't have a hard time finding talent for open positions on their teams: They already have networks of former employees and peers that they can use. Indeed, the employees of the best managers recruit for their teams as soon as positions open up.

But virtually all of the job postings you'll see publicly (especially for replacement positions or incremental growth) are described solely in terms of the project and the expected technical skill requirements rather than or in addition to the attributes that make a team more than a collection of extra appendages for a "leader". E.g. team values and non-values, the existing roles and expertise of the other team members and how the open position will complement those, the team's norms around work-life-balance and communication, etc, etc.

I think that the average manager at a tech company is not a good manager, and that the larger the company the lower the average (smaller companies just fail with bad managers). But by far most open positions at any time are positions reporting to bad managers.


One interesting note is that some people actually want an authoritarian leader because they don't want responsibility or more things to think about. I was talking online about how much I delegate to my team and I got responses along the lines of "the tech lead or manager should do all of that, why should I be doing their job."

>The best managers don't have a hard time finding talent for open positions on their teams: They already have networks of former employees and peers that they can use. Indeed, the employees of the best managers recruit for their teams as soon as positions open up.

It's also fairly easy to tell from talking to a team if the manager is actually good or not. If many people don't accept offers to your team then you may want to do some soul searching.


> some people actually want an authoritarian leader because they don't want responsibility or more things to think about.

It's easier to leave work at work that way.

1. Show up 2. Do work 3. Get paid 4. Leave

Your brain is freed to think of important things rather than work minutia. Of course, jobs like this often are mind-numbing if you're not the type to think with one half of the brain and do with the other and don't pay particularly great (in general).


> I was talking online about how much I delegate to my team and I got responses along the lines of "the tech lead or manager should do all of that, why should I be doing their job."

Well, that's because there's a difference between autonomy (e.g. "I'm going to let you approach the task in your own way without micromanaging your efforts.") and responsibility (e.g. "Your job is to do X and Y and Z"). What most people want from their leaders is some degree of autonomy, not necessarily extra responsibilities (unless there's a promotion attached...)


I wonder if there's an underlying cause for that attitude. I could see, for example, people working in a culture of high accountability and low trust trying to push decision making uphill lest they become a scapegoat. Meanwhile, places with a better office culture might feel more comfortable exercising autonomy and might value individuals taking the initiative more.


Agreed. Managers are empowered by organization and position. Leaders are empowered by inspiring and creating followers. They are diametric opposites in terms of mechanism and direction of power. Followers GIVE the leader power. Managers impose by force of position and control of organization power/money/etc.

The number of direct reports that function well as managed or led employees or any other organization hierarchy is about 5-7 but at most 10-12. This is a biological, neurological and psychological limit. It's also related to information theory and the UI-limits of the brain. It's related to chunking but also related to other limits of cognition and the finite number of hours in a day intersected with the minimal required cognitive focus/attention required to guide or manage an employee.


The distinction here is between Politician and Technician.

At the DevLead level one is a technician - diving into the complexities and trade offs and making a decision that will meet constraints that are the end point of a political process the devlead probably was not a part of

The politician you seem to be thinking of is someone (possibly the same someone) who is active in the political process that ends up deciding the constraints the technician works under - internal and external to the organisation

So this will be some of email threads or chats amoung some or all other devleads, it will be "being helpful" answering requests, assigning resources or time to help other politicians, or worse architecture diagrams, or it will be speaking at conferences etc etc etc. In the lingo it's becoming an "Authority". But one is only an "Authority" by being accepted by a "constituency" - maybe "all the devs in the company who think we ought to start using git" is your leadership position (ok maybe that's a decade old now but it's an easy example). Betray the constituency and lose the leadership position.

Neither are better than the other. Some political processes are better than others (in a democracy we perhaps appreciate this). Open decision making is likely to lead to long term better decisions - even if it is messy.

Linus Torvalds is an obvious leader example - using one political process (this is what I think - essentially using email as Thomas Paine used pamphlets)

But yeah - "managing" 5 people is reasonable - by the time you hit 12 it's a full time job listening to the moaning and they stop being a devlead. Only some can be a politican and a manager at same time.


Could you please describe your ideal, non-authoritarian model for management and leadership?


Manager poaching their former reports will become virtually unhirable in 2-3 cycles. Who wants the Hameln Rattenfänger inside their ranks?

Having lots of peers is not enough anyway; first you need to have a budget to lure those peers.


> This is an authoritarian model that makes the local despot the bottleneck for the team: Decisions end up being routed through the "leader", individual growth and development and team value is constrained within the scope of the "leader".

This isn't an absolute conclusion, I don't think. I can speak for my own organization -

I lead 2 functionally-aligned teams, meaning they own internal product/systems for two different partner orgs. I also have a technical program management function which owns initiatives related to the above when they span more than a single team or are company-wide initiatives. I also have an engineering team with its own EM who roughly cover the scope of whatever my organization is doing - could be single-function, could be company-wide.

In my model, decisions from Principal/Director/Manager team members don't wait on me. They can if they want extra eyes or extra air cover, but otherwise, my job these days is less about setting technical direction and more about unblocking via executive alignment, budget, people, etc.

So far you might be thinking this is uninteresting and just means that my individual managers are the local despot, but I've done two things to (so far) eliminate that -

1. The TPM is responsible for getting us through the design stage, and they manage no one (or maybe other TPMs). This means they herd the cats and work with engineers and SMEs to get aligned on whatever design will make us functionally and technically successful.

2. This one is more on my qualities as a leader, but I also ask the managers on my team to err on the side of delegating more and letting their staff impress them, and being there as a resource when they need help. I think this is called servant leadership, but I don't read a lot of management stuff.

Team composition (experience and horsepower) plays a large role in whether this model can be successful and how hands on you'll have to be, but in trying to distill a decade+ of experience into a couple of paragraphs, this is generally how I like to run things. YMMV of course, I don't hold myself out as an authority, just offering a counterpoint.

> The best managers don't have a hard time finding talent for open positions on their teams: They already have networks of former employees and peers that they can use. Indeed, the employees of the best managers recruit for their teams as soon as positions open up.

If you adhere to the standard 1 year non-solicit, this isn't true right away. Hiring in my first year at a new company is one of the hardest parts of my job. But I'm probably a bit of a rule follower when it comes to the non-solicit - moreso than others, I've observed.


What does "TFA" mean?


The f**ing article. A piece of internet slang that goes back many years, possibly originating on Slashdot and referring to the article that we're discussing here.


May have come from RTFM (manual) which according to this wikipedia link may have originated in 1979 https://en.wikipedia.org/wiki/RTFA


The "Featured" Article


The Freaking Article, in other words, the linked article whose thread this is.


What was "the fasting condition" in the study? They started a fast with sugar? They ended a fast with sugar? What was the interval between the ingestion and the test?

How much sugar was consumed?


If you're having trouble concentrating on a ketogenic diet it's likely that you're not eating enough vegetables, and might be eating too much protein.

Too much protein can take you out of ketosis, but people think they can just eat lean hamburger three times a day.

If the fat/protein ratio is right, the inability to focus can be a symptom of electrolyte deficiency, particularly sodium, potassium, and magnesium. You should be getting potassium and magnesium from vegetables, but supplements can be a useful diagnostic tool: If you take 400 mg of potassium and find, an hour later, that you're thinking more clearly, that's data.

Don't overdo it with electrolyte supplements, though; too much sodium can increase the calcium in your blood enough to give you kidney stones, if you're lucky, and worse effects if you're unlucky. Too much potassium can mess with your heart.


Eating more protein doesn't take you out of ketosis, that's just a myth; it will just lower the amount of keytones that are being produced. If only are protein or nothing at all and you would still be in ketosis.


> Take regular breaks from the medication.

I'm not sure this is universally necessary. When my pre-adolescent son forgets to take his medication in the morning it's immediately obvious, so it's clearly still having pronounced effect. He's been on the same dosage of the medication continuously for five years.


I’ve tried following the “take a break once in a while” advice and it does not work for me. I started to write daily and was able to realize being inconsistent with medication led to inconsistencies for me.

I was hesitant to take any medication. I was concerned it would make me anxious or similar.

Instead, my anxiety was greatly reduced. I had the thought, “is this what normal should feel like?” I don’t like the idea of being dependent on any one tool, but I’ll use a tool if it helps.


A spherical Whipple shield might be able to capture debris. A 1 mm thick aluminum shell with a radius of 100 m would mass 810,000 kg. If SpaceX's Starship can achieve launch costs of 10/kg, that's $8 M per sphere.

But passive objects are only as likely to provide value as the ratio of the number of the total volume of the sacrificial objects to the total volume of assets, so you'd want a bunch of them.

Since the debris of interest is really just the material traveling in orbit, a cylinder might be more efficient, if some primitive attitude control could be included efficiently.


Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: