There is a “Red Queen’s Race” between the accelerating productivity gains of abstractions, frameworks, libraries, and other stacks we can leverage, balanced against the fact that each one is a leaky abstraction, and the leaks interfere with each other in weird, mysterious ways.
Programming with all these tools is wonderful. Debugging problems in such am environment, not-so-much. But net, it’s a massive win.
First, I think you and the OP are correct. It’s not Barbara’s money, and Barbara’s career is driven by hiring and developing productive people. If her group is shipping good product, nobody in a growth company is sitting down and asking why a group that ships 30% more product costs 10% more per head.
That will come much later in the company’s trajectory.
One small exception to this from my n=1 experience is that all other things being equal, when hiring Barbara would rather pay at the low end of the range and then be generous with a raise, than pay at the top end of the range and have to fight to get you a raise.
That first performance review and raise set the tone for years to come, and good managers want it to be a happy experience for everyone.
—-
But that’s only if all other things are equal. Barbara doesn’t want to lose you to another employer by over-negotiating, so ask for what you want.
Let her worry about fighting for your first raise. That’s the job she signed up for.
I guess you're suggesting that adding a million lines to a log is excessive. But in my math, you can count lines like this: 1, 2, 3, ..., the-number-of-lines-that-fit-in-a-window, many.
Once you have more lines than fit in a window, you've already got "many" and now it's time to talk about tooling like searching logs.
A million lines of log feels like an excessive waste intuitively, but that's only because I was brought up on 8-bit computers with 16K of RAM. Now my watch has more power than probably all of the computers I owned in the 20th century put together, and a single image it displays might have more bits than a million lines of a log file.
---
I know this stuff isn't free, but we should at least consider the possibility that what we really need are proper and comprehensive logs/traces, and a "stack trace" is nothing more than an abbreviated execution trace that just happens to be relatively cheap to collect when something goes wrong.
The equally interesting implication of your comment that "we don't get a stack trace from an ordinary loop" is that maybe we should have logging and introspection the behaviour of loops, and the fact that we have them for function invocation but not for loops is combination of a historical accident and a lack of interest in rethinking how languages are implemented.
I have been to excellent public facilities in the US. Excellence is not evenly distributed anywhere, but the discrepancies of available public services between rich and poor neighbourhoods, between predominantly white and predominantly black neighbourhoods have been stark in my (n=1, granted!) observation of the US.
I think that the reality for the underclass in the US is that in theory, public services, voting, parks, community centres, education, jobs, and so forth are equally available, but that is not how it works in practice.
I have a feeling that as a Canadian, I really cannot cross the chasm between our society and America's, but I'll try.
Here in Toronto, the libraries are open to the public. Borrowing material or using their online facilities requires membership, still known as "having a library card," and for that you need to live, work, or study in Toronto.
(Update, courtesy xatt: There are also paid memberships available for those who don't qualify for free membership, approximately CAD10/month: https://www.torontopubliclibrary.ca/using-the-library/your-l.... But physical access to the library is free to everyone.)
I also think there are other allowances, such as if you care for a relative who lives there, but the gist is:
1. Anybody can walk in and use the physical library, and;
2. There are additional resources for those who have a connection to the jurisdiction.
There are similar things around community centres. During public swim, everybody can swim. During certain programs, you may need to be a member of the facility. I am aware that some physical parks have special programs, and you may need to register with the city and be a resident to use them.
But the park itself is public, and anybody can go there. You just may not be able to drop into an outdoor yoga program.
So clearly, we have a two-tier system, and the broadest tier is literally "everyone." The second tier is the size of a municipality.
In a certain very generous sense, Toronto's libraries, community centres, and park programs are similar to "gated" parks for a neighbourhood, as they have some features/programs that require membership and/or some connection to the municipality.
But in another, they are manifestly different. Neighbourhood parks are usually set up to exclude "the wrong people." They're about constructing gated communities, with all the classist and racist implications associated with these structures.
Whereas, public libraries are exactly what they say on the tin: They're for the public. Rich, poor, from near, or far.
I believe it's somewhat correct, based on some random diving training.
There is a thing where you deliberately hold your breath, which starts with taking a big lungful of air, which presumably is at normal pressure and contains 21-ish percent oxygen.
Even untrained individuals can hold their breath for a while like this.
There's another thing, which is what happens if the air you're trying to breathe either is at lower pressure, or has lower oxygen proportions. Either way, each breath has fewer oxygen molecules.
If you haven't prepared for this and try to continue breathing, you can quickly become short of breath, and then cognitive abilities become impaired, hampering your attempts to rectify the situation.
I wonder if the problem is that if cabin partially depressurizes by the time the masks drop, you may no longer be able to take a deep breath and hold it while taking your time putting on a mask.
I acknowledge that I do not have training in dealing with cabin depressurization. As luck would have it, my sister has worked for Air Canada for more than 30 years, I will ask her about it.
---
I see some other comments about the oxygen actually coming out of your blood into your lungs at low pressure, and the air coming out of your lungs because the cabin is lower pressure than your lungs.
Those make sense, and had I trained for summiting 8,000m peaks instead of tech diving, I would have known that!
"Wait until I find out?" Don't patronize me, I am keenly aware of automobile deaths, and furthermore, it's a complicated problem to solve for many, many reasons, some of which are caused by greed, lack of oversight, recklessness, and other social issues.
Nevertheless, in my lifetime cars have gotten safer by several metrics, most notably by distance traveled.[1] And they did not get better by thinking that driving around in deathtraps without seat belts and swigging whisky behind the wheel was acceptable.
They got better when people like Nader and MADD stood up and defied arguments like yours and loudly complained that yes, the system was BROKEN and needed to be fixed.
And then other people went about trying to fix automobile safety. Imperfectly. In fits and starts. And sometimes entirely wrongly. But nevertheless, progress happened when people attempted to fix things they acknowledged were broken.
They may not have obvious solutions, but we recognize that they are broken and we try to find solutions. We then find places the solutions don't work and we iterate or even replace the solutions with new solutions.
All systemic problems have complex sets of interactions and unexpected consequences. Racism, automobile deaths[1], and yes, air transportation.
We still don't shrug our shoulders and deny that the causes of deaths reflect a broken system, nor do we refuse to attempt to find solutions just because it's damn hard to make progress.
[1]: Sometimes, making the vehicle safer makes drivers more reckless. Making cars safer in practice is not as obvious as it appears in theory.
I make software that is broken ALL THE TIME. And what do I do then? Find the faults and then fix them. Just because software is not perfect does not mean that I sit like a dog in a house fire saying, "this is fine."
Killing 300+ people is broken. And our obligation is to admit that it is broken, and fix that thing. When we find another thing, we admit that the system is broken there as well, and go fix that.
Trying to diminish the seriousness of 300 deaths with mealy-mouth terms like "imperfect" is pure spin. If a plane, staffed by pilots allegedly trained on how to fly it, crashes from pilot error... TWICE, the system is broken. terribly broken.
See also: Security, physical or digital.
Terrorists flew planes into the World Trade Center. The system was broken. We fixed some of it. We put on some secuity theatre too. But we didn't say, "It's impossible to fix everything, so the system is fine." We didn't shrug it off as "imperfect."
We always begin by being truthful with ourselves about the fact that we have discovered that the system is broken. And if the consequences of its broken-ness are unacceptable, we fix it.
"Imperfect" is a word that should only be used for acceptable faults. Like, "The seating in economy is imperfect." It's too close together, but we don't have planes falling from the sky because they're cramming passengers together.
---
And now a footnote: Please avoid ad hominem arguments like "please name one system you've..." If the speaker's argument has a logical fallacy, point it out. If the speaker's argument is sound, it doesn't matter whether they write faultless software, or even whether they write software at all.
It's the argument we are discussing, not the person making it.
There are two completely separate issues here, and they need to be disentangled to have a productive discussion.
First, there is the question of the right-to-repair and to what extent Apple is hostile to the existence of a viable second-party and third-party repair ecosystem, neutral to it, or actively supports it.
Second, there is the question of the relationship between the costs of the product and the costs of the things needed to repair the product, like replacement screens.
The two are not entirely orthogonal, because many strategies for reducing the costs of the product or increasing the utility of a working product have consequences with respect to the costs of components and the ease of third-party repair.
An example discussed on HN many times is the choice between user-removable batteries and building a battery in. When it's built-in, engineers have more options for miniaturization or increasing battery size and product life. The consequence, of course, is that batteries are now expensive to replace, and the batteries themselves are harder to get in a third-party ecosystem because there is little standardization: Every device might have a different battery optimized for that device.
If batteries were standard sizes as they are for many consumer devices with removable batteries, engineers have less flexibility to increase battery life, reduce weight, or reduce device size.
The same reasoning extends to replacing screens and keyboards. Supporting a vibrant and viable third-party ecosystem means making engineering compromises that have a distinct effect on the product's price and competitiveness in a world where every review discusses device weight, size, battery life, &tc.
The very best thing for a third-party repair ecosystem is to have fewer device-specific parts, fewer proprietary connectors, fewer components hard-wired into place, fewer components that change from device version to device version, &c.
The more different parts there are, the more that are introduced for a model or two and then discontinued, the harder it is for the viability of a third-party ecosystem with affordable options.
I feel that there's a deep and challenging tension between repairability and the immediate, out-of-the-box product value. In some industries, consumers value the ability to wrench their own product higher than in others. In bicycles, for example, there is a great deal of conservatism around engineering.
For all the bragging about new technology, the bicycle industry doesn't really advance very quickly. What do we have now, twelve-speed rear clusters? Whereas when I raced in the 80s, there were seven speeds back there. Whup-dee-doo, where are the internal gearboxes? And the other big innovations are disc brakes and electronic shifting? Compared to the advances in telephones, this is nearly nothing. When I had a mechanical seven-speed rear mech, I also had a physical phone hard-wired into my car. That was "mobile communications" back then.
I'm not defending Apple, it's up to everyone to decide for themselves which choices they think Apple should make. But we should accept that any choice for making things repairable has an impact on the out-of-the-box value of the device itself.
And if we want right-to-repair to be a viable and profitable business choice, something has to change about the marketplace itself, namely:
1. Get consumers to repair their devices far more often than they replace them, and;
2. Regulation.
I prefer regulation, personally. Yes, it's the bogeyman of "regulated marketplaces." But it also creates a level playing field, so that repairable devices do not have to suffer product reviews complaining about their price, weight, size, lack of differentiation from other devices sharing common standards, &c.
Obviously, Apple stole mice and bitmapped graphics and GUIs from Xerox Parc. But a huge part of their pitch for MacOS in the early days was integration between apps on the platform.
Consistency of UX meant that there was lower friction for adopting new apps. Integration between apps such as being able to copy and paste text between apps from different vendors was a revelation by the consumer microcomputer standards of the day.
If I friend/follow a lot of people, and I have a simple timeline, I see what they're thinking about.
But if there's an algorithm selecting which posts I see from my social circle, and it happens to shade one way versus another, there are no ads, no synthetic content, but I have still been manipulated.
FB experimented with this quite famously a while back, they used sentiment analysis to see if they could manipulate people into being more or less depressed.
Without informing users, obtaining their consent, or following any of the other rigorous requirements for doing psychological research that we demand from medical proactitioners.
I question the words "sucked in." Influenced is more appropriate.
Before VisiCalc was invented by Frankston and Bricklin, nobody wanted it. Many people had the problem it solved, but nobody knew it existed, so they didn't go looking to buy it.
Advertising and public relations (see PG's essay "the Submarine") influenced people to buy it. A lot of those folks were people whose businesses improved after buying it, which is why "spreadsheets" went on to become one of the most important product categories of its age.
I think the difference between "influenced" and "manipulated" has to do with informed consent. An ad, clearly marked as such, influences. Advertorials masquerading as independent opinion, paying influencers to use a product to generate faux social proof, manipulating social media algorithms to make it appear as if a preponderance of people you know share a particular viewpoint...
That's all manipulation because the recipient is either completely unaware of what is going on, or dark patterns are used to make it difficult for them to discover what is going on.
But using keywords to target people and then giving them an ad that is clearly labeled as an ad? I think that's just influence.
If it's a national security issue for Tiktok to operate in the US, it's a national security issue for facebook to operate in Canada.