30-something, tired of working

[ICouldBeTheWalrus]

Thought I'd stop lurking and post on here.
I'm in my early 30s, married, living in California. For work, I do computer programming (AKA software "engineering"). On one hand, many would not complain about being paid quite good wages for what I do.
On the other hand, I find there are many ways it doesn't mesh with my values. I don't like the marketing / new-features driven nature of this industry, I like good craftsmanship (perhaps this is why I've recently gotten interested in hand-tool woodworking.) I really wish there was more time for gradual improvement of products, rather than learning the same hard lessons over and over as organizations throw out the old for the new and shiny. I find myself changing jobs every 3 years, usually because I've gotten bored of the particular problem domain, having become one of the resident experts. And ultimately, it seems that no job can let me explore enough of my wide interests to be worth spending 40+ hours a week at. Also I just find it plain tiring to work for other people and feign some sort of loyalty to the corporate entity. Furthermore, it sucks up a lot of the energy I'd have for being social with people I'd actually like to hang out with.
At the moment, I think I'd be comfortable living off about one-quarter of my gross salary (not counting some recent stock plan stuff that has done quite well.) Not very ERE

I'm probably about a third of the way toward accumulating enough assets to comfortably retire.
I and my wife would probably like to buy a house fairly soon, though we're not in a great hurry, thankfully. While I don't like how overpriced houses are in the US (in large part due to the subsidies), there are a number of reasons I'd like to own rather than renting too.
Rather than grind away at working for another 10-20 years, I'm starting to consider other possibilities. There are certain limits on how much I drop my cost of living without significant lifestyle revision, but I have been looking at ways of making money, not the least of which is trading stocks and options, something I've gradually been accumulating experience at.

[dragoncar]

Maybe you could work for a smaller outfit? If the problem is boredom, a smaller place will give you responsibility over a greater array of problems. If the problem is hours, could you try contract work, or do they expect fill time work?

[jacob]

It seems to my moderately informed self that a lot of the software craftsmanship was lost when hardware started developing faster than software. It's amazing what people could stuff into the bootsector in the 1980/90s, whereas now something like a compiled version of print hello world takes up 8kb.
(I hate how my fully adequate setup is slowly becoming obsolete as new features are introduced. It would be marvelous if development efforts were aimed at making existing codes faster rather than adding new features.)
Disagreeing with the values of the entire industry is hard to escape (and it's part of the reason I no longer work in physics); are there any alternatives?

[dragoncar]

This slide from an Apple Keynote says it all. Try to guess what year this is from:

Talk about throwing away the old to make way for... the old.

[B]

I recently left the software "engineering" (love the quotes) industry, with only a couple years under my belt. The reason is simple, and ICouldBeTheWalrus and jacob both put it very well. It NOT engineering. Shiney bullshit calls the shots, not efficient, effective designs.
I have no debt, modest expenses, and a fair amount of savings. If you can say the same, I think it's time to try something different. A smaller company, freelance work, anything is better than laboring for something you don't believe in.

[jacob]

It would seem to me that something like numerics, controls, finance, where speed rather than features is of the essence, may solve the problem.

[George the original one]

> I really wish there was more time for gradual

> improvement of products, rather than learning

> the same hard lessons over and over as

> organizations throw out the old for the new

> and shiny
That is exactly why I didn't major in computer science when I began college in 1980. Even back then, programming was all too often a matter of reinventing the wheel.
> And ultimately, it seems that no job can let

> me explore enough of my wide interests to be

> worth spending 40+ hours a week at. Also I

> just find it plain tiring to work for other

> people and feign some sort of loyalty to the

> corporate entity.
And those are the exact reasons we all tend to want to leave employment.
You're in good company now!

[elai]

It's called software engineering so that when you want to move around the world it will fit into usually easier 'engineering' work visa categories. More accurately software engineering can feel like software construction, or software design or even software engineering at times. Software has gotten to a point where they're like cars, a lot of similar parts doing the same thing, and often you got to mold your own metal shell, even though it has the same car shape.

[dot_com_vet]

Maybe you should look into DevOps. It's about doing everything right, sounds like it would be up your alley.

[KevinW]

Well, since the 1980s CPU speeds and demand for software have been growing exponentially, and the supply of Master-level labor (as in Apprentice, Journeyman, etc.) has not kept pace. When CPU cycles are always getting more plentiful and labor is always getting more scarce, the Pareto-optimal solution is to substitute cycles for labor.
At first this was done with things like Lisp which operates at a higher level of abstraction and automates things like garbage collection so a Master can express in 10 lines what used to take 100. Over time abstraction shifted from a productivity tool for Masters, to a support system for Apprentices fumbling through Journeyman-level work. Sadly a lot of code is "written" by copy-pasting code from websites without a real understanding of how or why it works. The result is often inelegant, inefficient, buggy, and unmaintainable. Due to the general trend of job title inflation, this is often labeled engineering, even though it doesn't fit my definition of the term.
That being said, there are groups that still do real software engineering (no quotes) so I wouldn't write off the entire field.
I agree that most software is flashy, unnecessary fluff, but is there any segment of the consumer economy where that isn't the case?
@jacob One solution to system requirements creep is to settle on a stable software environment. Some people live in Emacs. FVWM and GNUstep run fast on old hardware, but they make your computer look like a Unix workstation circa 1990.

[ICouldBeTheWalrus]

Thanks for all the ideas.
@dragoncar: Yeah, it's terribly amusing how people keep rediscovering the idea of a document's changes being automatically saved.

[blah]

@Jacob, I agree with you that efficiency is the beautiful side of software development that has been lost in the commercial sector because of hardware speed / specs. It is a sad side effect. I am glad for the improvements in hardware but only because of scientific research applications which can be utilized to calculate results faster than on slower boxen as you also mentioned above. Your supposition that working in such a field is a antidote to the problem is partially correct, if you find the right lab or group to write for, but many will want you to use a high level language and produce results quickly rather than efficiently as well. So, alas, despite the greater need in scientific research for efficient coding, it doesn't always work out that way in practice. People are in competition for publishing first and unfortunately, discovery rather than improvement in method or technique usually is what gets published. Never mind that improvement in technique and efficiency is what ultimately drives research and discovery forward for everyone. Blame the scientific journals for that one. </end rant>