Scubabear68

Google must have thought the court was a customer.

A good developer should ideally push their organization out of the ambiguous zone. Give feedback to whoever is writing requirements on the gaps.

Otherwise, if you are in that ambiguous zone regularly, either you are pioneering some super new innovative area, or your organization isn’t very good at software development as a whole.

One org I was at, the product managers constantly complained that developers couldn’t read their minds and automagically “know” all the non functional requirements, validations, etc. They were used to working with super senior devs who knew the space well, and could not adapt to devs who were more junior and also new to the industry vertical.

I have over 30 years experience in a variety of roles in software delivery, been doing architect-y type things for the past several years. I never received my baca degree, and it never hurt me materially.

Work whatever network you have. Build one from past colleagues on LinkedIn if you haven’t. Start applying for remote work. Look for contract gigs if time is that short. Get your CV out there and keep applications flowing to companies.

Actually it is amazing how few generations it takes to go back that far. Assuming 15-20 years between generations, that’s 450-600 generations.

Using that ruler, go back 6 generations and you are at circa 1900.

What amazes me about C++ is that the complexity is deliberate. I started learning it voraciously in the early 90s, read Design and Evolution of C++ and many other books. Over and over again the theme was to be broad and deep, multi-paradigm, kitchen-sink approach that should have as much power as possible, with the bizarre caveat that it had to be backwards compatible with C (mostly).

I have never understood this POV, but to understand C++ O think you need to grok this key point deeply.

I LOL’d a bit at JBoss, only because how radically history changed from App Servers ruling the world to practically disappearing over night into the maw of Spring boot (and it’s ancestors).

It was my privilege to know Dr. Tullis Onstott, who discovered the deepest known Nematodes miles below the surface over a decade ago. Absolutely fascinating how life finds a way to persist and survive in extreme environments.

https://en.m.wikipedia.org/wiki/Tullis_Onstott

Always good to read Yegge. But as with Joel back in the day, with this kind of post you have to separate the meat from the marketing.

It seems clear that LLM’s are going to be decent as assistants, but the 5x claim is BS and the marketing part of the post. Most of software development isn’t coding, it’s testing and experimenting and bug fixing and refactoring and documenting and interacting with other people and often companies.

If we get to the point where an LLM can get assigned a Jira ticket, make a bug fix, slack the PM for clarification on the ticket, debug the code, write regression tests, etc, then that will be something.

Right now? It’s a much improved Clippy with a very big set of knowledge behind it. When it’s embedded in the SDLC, that will e something. If we ever get there.

I love the last paragraph of this article:

“In the same house, in the 1980s, Nonna Fiore once served some English guests lasagna, per my uncle’s request. The lasagna was cooked from frozen, her story goes. Life was busy and, anyway, she had no qualms about serving a supermarket ready meal; people could only dream of such a luxury during the war. None of the guests suspected that she hadn’t made it from scratch and everyone was delighted, her Italian son included. She reminds me of this, then looks up at me and winks”.

I agree with the premise, but keep in mind we weren’t always so concerned about injuries and deaths on the job.

Many of our greatest works were built on pools of blood that would not be tolerated today.

I think a good argument could be made that CS programs are too science oriented and too foundational in nature.

Most developers probably do not need data structures, OS, compiler type courses, but instead would benefit from higher level, engineering type courses that reflect modern application design and development.

Note: I am 3 decades out of school, and there have been some moves in this direction I can see from interviewing juniors. But not enough.

The ideas of lines of research being suppressed is a rich topic in science fiction novels. Usually hidden aliens or secret factions on Earth actively steer scientists away from certain topics, usually that will lead to break throughs in star drives or weapon-of-mass-destruction designs.

I do a lot of technical due diligence for companies in a bunch of different contexts (private equity pre-acquisition, post acquisition, management change over, etc). I have found that tech debt is highly correlated to the business goals and tech organization structure, and has to be evaluated in those contexts.

Startups you want to see with fairly high tech debt. Tech debt there is a very often a sign of shipped code that is meeting some customer need. You do need to watch out for insane tech debt that may be hard to undo later on (running everything in an ancient proprietary language for example).

Startups with low tech debt often are devs navel gazing without a lot of revenue coming in. This is surprisingly true almost all the time.

Very large companies will tend to have a mashup of all kinds of technology that is never fully untangled. A very large Corp with little tech debt is exceptionally rare. If it ain’t broke don’t fix it is the rule.

Mid sized corps and late stage startups seem to be the sweet spot. Fairly low tech debt systems that is providing high value is still possible. This is where you have enough maturity to know what “good” looks like, and enough developers, momentum and budget to implement the vision.

College made sense for a very long time because a lot of books were hard to come by for regular folk, and the cost was reasonable.

They were great places for young adults to complete the transition to “adult”, to experiment and figure how who they want to be.

These days the costs seem to far outstrip the benefits, and the Internet makes so much more accessible at a young age.

Parents and kids are finding it hard to justify hundreds of thousands of dollars just to “figure things out”, as we used to.

This is true for autodidacts. I am one, and the same is true for me.

It is absolutely not true for most people. Most people need a human guide to show them the way. I am living this now with my two kids, both of whom need instruction (and also help from mom and dad).

Very good point. We had to do that with one family member who just does not respond well to men, as one example.

An example of a variant on this, when my wife and I started the journey to adopt in our State (NJ), we were required to give permission to speak to our doctors, and they went through everything. I mentioned something once to my doctor about cutting down on alcohol and that became a big deal in our interviews with State workers.

The most important thing around therapy is to realize that therapists are people, and some are good at it and some..not so good. I have seen it work, and seen it be a total disaster, and who the therapist is seems to be a large percentage of that.

I know nothing of Ironman training. I have trained to be a (slow) marathon runner, and I am really shocked at how low his running miles were. He said he was running between 2 and 12 miles for each run. He mentions a bone bruise that I guess severely limited his run training. Only two tiny runs between March and October! I would not think his bike and swim training would be able to compensate for near total lack of runs. He must have a fabulous baseline of fitness to complete a marathon and the rest of iron man like that.

Gamifying charity. It is sadly very common. Who knows if there even is a mystery $20k donor, or if it’s just a figment of the MIT marketing department?

I am honestly surprised there is no “2x match for the next 24 hours” or something similar (which again is completely unverifiable).

For shops light on DevOps-fu, Confluent hosted Kafka is popular for just this reason.

Too bad they didn’t do this at least six months ago. The federal government really needs to take a hard look at the ratings agencies.

Hoping interest rates won’t go up is hardly “notionally hedged”.

“If you get any kind of loan from SVB, you’re required to keep your cash with SVB”.

Really? How would this be enforceable or even discoverable?

An important stand out quote to me here: “ It’s fair to ask about the underwriting discipline of VC firms that put most of their liquidity in a single bank with this kind of risk profile“.

I really don’t understand why these firms didn’t use at least two banks for their deposits. Surely these tech firms have heard of single points of failure being problematic?

I have often wondered how celebrities and other “rich” people dealt with cash assets, this is interesting in the unlikely event I ever strike it rich.

The flip side of the coin is how do rich people select accountants and wealth managers that they can trust? I have heard so many horror shows of celebrities getting ripped off by their accountants. It seems to happen even more to poor people who suddenly become rich - lottery winners seem unable to hang onto the cash. And the Murdaugh murder trial revealed networks of lawyers and fiduciaries playing fast and lose with client money (one example was a trust manager loaning Murdaugh huge amounts of cash from a trust for two minors who mom was killed in an accident, and the girls got a multi million judgement).

Are there trust networks? Or word of mouth and luck?

With all the noise from the FDIC seizure it’s hard to find fundamentals right now of SVB (I’m on mobile at the moment which is rough to search for SEC filings from). How big are they?

How significant is exposure outside of startups and tech companies?

“HOW DO PEOPLE NOT STEP-STEP-STEP”.

For me it’s because that is slow, tedious and exhausting. Often you need to cover a lot of code surface area to find a bug.

Worse, the bug may be something reported in production in a giant pile of code.

A good set of logs will go a very long way to at least narrowing down where the bad behavior is coming from, and is generally scalabale.

The debugger is useful, but generally only at the extremes of “new to the code, totally clueless” and “really advanced bug that I need to bring out the big guns for”.

The logging approach clicks in really fast the minute you’re talking about a multiprocess program, being microservices or just plain old SOA.

The requirement that you have a public IP for your database and a third party system is sucking in our data are giant red flags for industries with any sort of confidential or proprietary data.

A huge strength of NocoDB is that it is run as just another service accessible your DB that you own.

I can see Virtual Columns being interesting for reporting-type purposes. The mentioned use case of “notations” on data owned by someone else is also interesting. But again - do I really want someone else hosting my data? I don’t know.

Is the target here then basically shadow IT adding onto other people’s data sources that they can’t get the devs to change directly?

With “traditional” RDBMS systems, putting a lot of code in the DB lead to a lot of scaling issues where you’d need gigantic machines with a lot of RAM and CPU: because the DB was doing so much work. It was expensive and clunky to get HA right.

In more modern DBs being distributed horizontally, this approach may see a rebound. The big “but” is still costs, in my experience in AWS as an example, managed Postgres Aurora was surprisingly expensive in terms of monthly cost.