[DanielBMarkham]

Let's make it an even 4!

Just turned 50 yesterday. Wrote my first contract program in 1982 when I was 17.

Today I teach teams how to rock-and-roll. I also keep coding, but mostly as a hobby. It has nothing to do with programming -- I simply have too many irons in the fire to do all of the things I love as much as I want to.

[lisper]

I guess I might as well pile on. Turned 50 last year. Got my first paycheck for writing code junior year of high school (writing BASIC code on a Commodore PET!) Working on my fifth startup now.

[SCHiM]

I can't imagine all the technologies and skills you guys have acquired.

If I may derail the conversation a little bit, may I ask if/how you use the things you've learned today?

I mean, is it things like paradigms that have stuck with you like OOP/functional/w\e, or do you always structure your exception handling in a certain way, no matter the language? Are there skills/technologies you've been using since you've started? For example little scripts of basic that you've never let go off that automate things like setting up build servers and the like?

[blisterpeanuts]

SCHiM: "is it things like paradigms that have stuck with you..."

For me it's always been about problem solving rather than simple fascination with some particular technology. If I can solve a problem with a keyboard macro in Emacs, great. If it requires Perl or Java or javascript, so be it.

I try to use the tool that's appropriate. If I can solve a problem quickly and move onto something more interesting, one-and-done.

If I have to stop and learn something new (e.g., WKWebView in iOS ObjC) to get my task done, so be it, and I'll put in some late nights to get there because at my age, I'm a bit paranoid about looking bad so I try to give the good folks at HQ no reason to doubt me. I spend a lot of time on Stack Overflow and Youtube doing concentrated learning.

But the real thing is what others have also mentioned: properly defining a problem that needs to be solved, proper communication, keep good records, and try to maintain transparency, honesty, and pleasant comportment at all times.

Honestly the older I get, the less I notice people's age. If someone half my age knows something I don't, then the way I see it, they have something to teach me. I've been to many conferences and watched many youtube tutorials where the teacher was very young (from my wizened perspective) but the information is why I'm there and that's all I care about.

What do I think about young people? (you didn't ask but I'm saying it anyway)

I love young people. They have so much spirit, so much energy and creativity. I keep hearing critical (snarky) things about millennials this, X-gen that. But I don't see it. The young people I've been around (for a while I was back in school full time, surrounded by 20-somethings and a few 30-somethings) were a joy. Fun, humorous, inquisitive.

Everyone has his faults, not least myself, and I believe as we get older we become more tolerant of others' faults and shortcomings. In fact that may be the single hallmark of growing older (apart from physical issues).

[lisper]

No more BASIC, but I still code in Common Lisp whenever I can. And I'm using some library code that I wrote when I was in grad school 25 years ago.

Once you grok Lisp, everything else is easy. You come to realize that the vast majority of what passes for "new technologies" is really just a re-discovery of something that exists (or is easily implemented) in CL. That makes it a lot easier (if a tad frustrating at times) to keep up.

EDIT: CLOS, and generic functions in particular, are a HUGE lever that no other language has co-opted yet.

[ScottBurson]

Ha! I'm still using a piece of Lisp library code that I wrote as an undergrad 35 years ago :-)

[bandrami]

You come to realize that the vast majority of what passes for "new technologies" is really just a re-discovery of something that exists (or is easily implemented) in CL

So much this. (From a soon-to-be geezer who's staring down 40 soon.)

[munificent]

> no other language has co-opted yet.

Check out Dylan, Julia, and my old hobby language Magpie[1].

[1]: http://magpie-lang.org/multimethods.html

[lisper]

I guess I should have said: no other mainstream language.

[PuercoPop]

Julia has generic function but they are not defined in Julia itself nor is dispatch mechanism can be extended in Julia itself. CLOS is all about meta-circular semantics so I wouldn't include Julia in there (at least not fully)

[typedweb]

Same with conditions and restarts.

[codemac]

GOOPS[0] in guile is definitely hugely inspired by CLOS, which also has generic functions with multi-method dispatch.

I highly recommend guile to anyone who wants the battery included nature of common lisp. I know that racket is also a great choice here, I just have little experience with it.

[no-name-1231]

Racket's object system is largely inspired by Java. So is its statically-compiled module system (in which you can't change anything once it's been compiled, unlike a real Lisp), and Raco build tool. Programming with objects in Racket feels a lot like programming with Java objects in Clojure.

[mgkimsal]

I'm not quite in the same age range, but north of 40, and have been doing this for > 20 years.

The tech changes - I don't use BASIC or Z80 machine code day to day. The skills that I can speak to, and the skills I see from others with this level of experience, are problem solving and communication, and secondarily the confidence that comes from having made mistakes. Few projects fail miserably solely due to technical issues and skill; in my experience failures come about because of poor communication (including documentation, but also speaking, reading body language, office politics, etc).

[madez]

We younger folks are incredibly curious about what you've learned. Can you elaborate on the topics you mentioned or recommend other sources? Can you give tips on best practices?

[mgkimsal]

Hrm...

Not sure I have any specific sources to share (others might) but after working with all types of hardware, languages, business realms and what not... you start to see patterns. Human behavior patterns, mostly; the same sorts of needs were there 30 years ago as are here today. People need stuff done yesterday, don't know how to describe what they want, oversold a client on something, etc. The numbers may be bigger, the gigahertz faster, and the memory far greater than people could have imagined, but the problems of translating what people say in to what they want via code is ... fundamentally still the same.

A 16k RAM module was ... $200 in my early days. Then a few years later I got 512k for only $150. Today I can get a 16 gig USB stick for $10, at at a local grocery store checkout aisle no less.

No doubt technology has changed and become more ubiquitous. Peoples' expectations of what's possible and what's normal are somewhat different today than 10-20-30 years ago, but the communication problems are still largely the same. Who's in charge of a project? What are the parameters? What defines "success"?

I'm not sure there's any real silver bullet here - most of the "mythical man month" stuff (from even before my time!) is still largely true (speaking in generalities because I can't think of every single line right now).

Learn to figure out how to communicate better - emails, phone, IM, paper, phone, f2f, etc - they all require different skills, and have different impacts on different teams.

Happy to answer more specific questions here or via email - [e-mail] - not sure if I'm just rambling now or not.

[narrator]

Back in the early 90s, I was doing mac68k programming. Very large Inside Mac volumes were pretty much the only documentation for anything - no stack overflow, no internet, no searchable docs. Almost everything was closed source, and many people wrote their own data structures. When programming Macs, if there was any kind of null value dereference or pointer problem the computer usually crashed and you had to reboot.

The main thing I learned was to be able to look at code and determine if it was going to actually do what I wanted it to do without having to run it through the debugger a couple times.

It's important to spend time thinking about and reviewing code when edit, compile, and test cycles start to get long. There are a lot of instances where that still applies, such as with multi-threaded programming, integration testing or long running jobs.

I also learned the value of having muscle memory for APIs. Sure you can look things up in a jiffy, but actually rote memorizing stuff that is used often can speed development up.

[mark-r]

Interesting the pattern in this thread - I'm 57 and started in 1973!

There's plenty of languages I've used over the years that I have no use for today, or hardly even remember: Basic, FORTRAN, assembler, Pascal. None of those languages even had exception handling, so it's hard to say that it influenced my thinking on the subject, except by forcing me to be familiar with the alternatives.

Today it's C++ and C# that pay the bills, with as much Python as I can sneak in. Next up in the queue is Javascript, for which I'm admittedly overdue, but that one's entirely on my own time.

[knice]

Just curious, as you moved from language to language do you think there were any patterns as far as the timing of your moves? Did you jump into new languages as they emerged or wait for them to reach a stable mindshare within your peers? Or perhaps waited until you needed to pick up a new language in order to pay the bills?

[mark-r]

Basic was what I started with, as did a lot of people from that era and after, just because it was simple and available. FORTRAN I tried on a lark, then got serious with it when I moved to the University mainframe. That's where I picked up the first of many assembly languages, after finding out it was the only way to take full advantage of the OS and achieve the best efficiencies. It's also where I learned Pascal, which was an up-and-coming teaching language, and it soon became my go-to language for many tasks. Between Pascal and assembly I was good through about 1990, when my job required me to use a Unix server, and of course the natural language there was C. That led to C++, first on my own time then at my next job in 1997. There was a short period where I found myself applying OO concepts to assembly code! C# was a recent addition for a job-related task where it was an obviously better choice than C++.

Python is the interesting one of the bunch. I was exposed to it about 1995, when a coworker selected it as an embedded scripting language for our product. I didn't pay much attention to it at the time, as it was outside of my immediate responsibilities. It was selected again as an embedded scripting language at my next employer around 2003. This time I paid attention, and came to love it. It's the language that lets me turn my thoughts into results the fastest.

Thanks for asking the question, I've never stopped to think about my programming history in this much detail.

[knice]

Thank you for a thoughtful, detailed response. It seems like a mix of job exposure and natural curiosity led you from language to language.

Throughout this thread, I've noted a distinct lack of dogma about everyone's evolution as a programmer. While I imagine there was some of it during the popular phase of many of these languages, it's nice to see mostly pragmatism in everyone's career journey.

[ericssmith]

Here is the first program I ever entered into a computer.

1 8 + A 5 + 0 0 + 6 5 + 0 1 + 8 5 + F A + A 9 + 0 0 + 8 5 + F B + 4 C + 4 F + 1 C

I remember how much like some kind of incantation it was. I still get that feeling. And I've had the good fortune to work some fun, powerful, and interesting magic with those incantations.

I suspect that most younger people haven't yet seen their tech choices slip into decay and disuse. And they hang on tightly, hoping to expend less effort in learning as time goes by and they master that tech. It's a vain hope. The essential part of the experience is the underlying creativity, the joy of getting to the next ledge, and really understanding what a marvelous thing a computer is.

[waqf]

Apparently this is KIM-1 code (https://en.wikipedia.org/wiki/KIM-1).

[edpichler]

I fell as you guys. I am very excited with technology since I started. There is so much new and interesting things to play and read that I have no time to do it all.

I can clearly that each year more and more interesting things appear. The speed is also increasing, in way I can't digest it all. It's difficult what to choose, to prioritize is the key to deal with it.

I always get myself wondering how much new things I will see next years.

I'm 30 now, and I started with 17.

[kinleyd]

I'll be 52 this November. Have been coding mostly as a hobby since '83, with peaks in programming intensity at different stages of my life: sometimes to scratch an itch and sometimes doing some contract/consulting work. Still try to hit the computer whenever I can, but my health's been letting me down lately - nothing too bad, but it's been affecting the focused time I need to do programming right.

So now when I sometimes think of my mortality, the thing I regret is the programming I can't do when I'm gone. Or maybe there is programming after death? I'm hopeful. :)

In the meanwhile, the free time that's opened up from not programming as much is being wisely invested. I'm working hard on my guitar, and before I go dammit, I'm going to shred like a pro!

[gcr]

Happy birthday! :)