The Dark Clouds of Choice

I enjoy Go (the programming language/tooling). It rapidly facilitates efficient, concise solutions. Whether it’s processing some data, building a simple connector web service, hosting a system monitoring tool (where I might layer it on some C libraries), it just makes it a joy to bang out a solution. The no-dependency binary outputs are a bonus.

It’s a great tool to have in pocket, and is fun to leverage.

I recently pondered why this was. Why I don’t have the same feeling of delight building solutions in any of the other platforms that I regularly use. C# or Java, for instance, are both absolutely spectacular languages with extraordinary ecosystems (I’d talk about C or C++ here, using both in the regular mix, but while there are usually very practical reasons to drop to them they don’t really fit in this conversation, despite the fact that from a build-a-fast-native-executable perspective they’re the closest).

Goroutines? Syntactic sugar over thread pools. Channels? That’s a concurrent queue. Syntactical sugar is nice and lubricates usage, but once you’ve done it enough times it just becomes irrelevant. A static native build? A bonus, but not a critical requirement.

There is nothing I build in Go that I can’t build at a similar pace in the other languages. And those languages have rich IDEs and toolsets, while Go remains remarkably spartan.

The reason I enjoy it, I think, is that Go is young enough that it isn’t overloaded with the paradox of choice: You throw together some Go using the basic idioms and widely acknowledged best practices, compile it, and there’s your solution. Move on. A lot of strongly held opinions are appearing (about dependencies and versioning, etc — things Go should have gotten right in v1), and an evolutionary battle is happening between a lot of supporting libraries, but ultimately you can mostly ignore it. Happily build and enjoy.

Doing the same project in Java or C#, on the other hand, is an endless series of diverging multi-path forks. Choices. For the most trivial of needs you have countless options of implementation approaches and patterns, and supporting libraries and dependencies. With each new language iteration the options multiply as more language elements from newer languages are grafted on.

Choices are empowering when you’re choosing between wrong and right options, where you can objectively evaluate and make informed, confident decisions. Unfortunately our choices are often more a matter of taste, with countless ways to achieve the same goal, with primarily subjective differences (I’ll anger 50% of C# programmers by stating that LINQ is one of the worst things to happen to the language, and is an inelegant hack that is overwhelming used to build terrible, inefficient, opaque code).

We’re perpetually dogged with the sense that you could have gone a different way. Done it a different way. I actually enjoy C++ (I admit it…Stockholm syndrome?), but with each new standard there are more bolted-on ways to achieve existing solutions in slightly different ways.

I of course still build many solutions on those other platforms, and am incredibly thankful they exist, but I never have the completely confident sense that it is optimal in all ways, or that someone couldn’t look at it and ask “Couldn’t you have…” and I could firmly retort. I continue an internal debate about forks not taken.

The Best of the Best of the Best…Maybe?

I’ve talked about the consulting thing on here a lot, and the trouble involved with the pursuit. While I’ve been working with a fantastic long-running client, and have primarily been focused on speculative technology builds, I considered keeping skills diverse and fresh by mixing things up and working occasionally through a freelancing group that purports to have the best of the best. Doing this would theoretically remove the bad parts of chasing engagements, pre-sales, badgering for payments, etcThe parts that are a giant pain when you hang your own shingle.

If it was just challenging engagements with vetted clients, cool. Just the fun and rewarding parts, please and thank-you.

Freelancing groups almost always end up being a race to the bottom, generally becoming dens of mediocrity, so the notion of a very selective group made it more interesting. I like a challenge, and if someone wants to build a collection of Boss Levels for me to battle, the barriers yielding a group that customers would pay a premium for, maybe it’d be interesting.

So I took a look to find that one of their requirements — not mandatory, but strong recommended — is that you post on your blog how much you want to work with them. This is before you know anything about the credibility of their process, rates, the quality of peers, etc. And this isn’t something you can easily find: they demand that you don’t talk about their process or virtually anything involved with working with them, so in the absence of any information about them (beyond some very negative threads I later found on HN, primarily posts by throwaway accounts), you need to tell the world how eager you are to join them.

This is a profound asymmetry of motives.

Who would do that? It is an adverse selection criteria, and it instantly deflated my impression of their selectivity and had me clicking the back button: The sort of desperation where someone would pander like that — while certainly possible among fantastic talents in bad situations — is not a good criteria. To try to harvest cheap link and namespace credibility like that itself makes the service look lame, like a cheap 1990s long distance carrier.

I still want those challenging smaller efforts, however — variety keeps me fresh and in love with this field, and some extra cash and contacts are always nice — so instead I’m going to start offering micro-consulting via yafla, pitching it on here as well: Pay some set amount (e.g. $500), describe your problem and supply supporting data, and I’ll spend from four to eight hours on your problem, offering up appropriate deliverables (utility, shell of a solution, decision input, directions, analysis, etc). I’ll get that online quickly.

Going for the smaller chunk efforts, of the sort that someone with a corporate credit card can easily pay for, should prove much more tenable than seeking significant engagements where the sales process just drags on forever.

It also is a cross-motivation desire to encourage me to spend more time on this blog, and if I’m pitching bite-sized consulting on posts and actually seeing uptake, it’ll keep me authoring content that people want to see.

While I have always enjoyed the cathartic aspect of blogging, it has never been profitable for me. Every engagement has always come via referral and word of mouth, and even when I’ve sought traditional employment I’ve always been amazed that no one ever even does a simple Google search, much less read anything on here. I never expected any return, and adore that anyone reads these things at all, but it does make it so at times it’s tough to justify time spent making content.