Chad Perrin: SOB

14 September 2006

Joel Spolsky has not jumped the shark.

Filed under: Cognition,Geek — apotheon @ 04:55

Joel Spolsky has increasingly often been described as jumping the shark lately. For instance, see Has Joel Spolsky Jumped the Shark? by Damien Katz and Has Joel Spolsky Jumped the Shark? by Jeff Atwood. In a comment at DHH‘s weblog post Fear, Uncertain, and Doubt, reader Phil even posted a picture of Joel jumping the shark. The use Perl; weblog of someone coincidentally using the name “perrin” (even though I don’t even know this person) said “Are we done with Joel Spolsky now?”, to which another use Perl; user operating under the name Alias said “So, like, now can we say ‘Joel has jumped the shark’ ???” Darrell Norton asks “Is Joel Spolsky jumping the shark?” I have seen at least a dozen other references to this potential shark-jumping condition of Joel’s, and almost all of it was prompted at least indirectly by his Joel on Software post Language Wars.

The first time I saw such a reference, I was immediately tempted to agree that he has jumped the shark, or at least to believe he was never quite good enough to be said to have jumped the shark. It’s true — he has said some excellent things over the years, and made some intelligent observations, but at the same time he has carried some obvious biases and (mis)conceptions that nearly offset the good material. I don’t think the opinions expressed in Language Wars and subsequent references to it at Joel on Software are newly formulated. Instead, I think Joel has just finally said something that made proclivities that have existed all along so clear that nobody can reasonably ignore them.

I have also recently, though I don’t recall exactly where, run across the incredible statement by someone in online discussion that Paul Graham may have jumped the shark. My first reaction was, of course, disbelief and denial. The suggestion was accompanied by a brief statement that made me stop to think about it, however. This accusation (or speculation) was based on the idea that Mr. Graham may have begun to stray from his open source, hard-coding, ninja guru Lisp hacker roots in his zeal for making a fast buck with his startup investment business Y Combinator. The extreme vaporware state of Graham’s next-generation programming language Arc seems to support the idea that he may have jumped the shark somewhere. With dismay, remembering the excellent and insightful essays of his that I’ve read, I began to question.

After long consideration, I came to realize that (assuming for a moment Joel’s writing has begun to fall off in quality of thought, rather than merely showing the worn threads that were always there) there are qualities both situations share that do not match up with the shark-jumping phenomenon. It is my belief that neither has jumped the shark, not because their most recognizable pursuits are necessarily still improving or at least holding steady quality, but because the term “jumping the shark” cannot apply to either one of these people. Humans are not so simply constructed and engaged.

People of prominence in some niche such as Paul Graham and Joel Spolsky, such as ESR and RMS, and even people such as Elizabeth Taylor, Donald Trump, and members of Metallica, never simply jump the shark. Only certain pursuits of theirs may have done so.

People are collections of interrelated processes, when measured by what they do, like story arcs with beginnings, middles, and ends. They are an ecosystem of intertwined passions and vocations. When one arc ends, dies, is completed, it is best to let it go and continue to pursue other, more promising goals for the future. These personal arcs arise naturally within us and within our lives, and we ought to go with the flow, picking up what newly engages us and leave in our wakes what has reached completion and closure. Just as a given entrepreneurial pursuit should be allowed to end when its time has come, rather than entering an economically parasitic stage of its existence where it must destroy others to sustain its own pathetic life well beyond its natural lifespan as most corporations do, so too should our pursuits be allowed to fade into memory as new pursuits take over.

Productive intellectual pursuits such as Joel Spolsky’s writings about software development ultimately reach a point, if they are pursued effectively and long enough, where the individual engaged in such an activity has reached the culmination of his or her contributive efforts. Recognizing when this has occurred and knowing how to let go of it while transitioning to new pursuits is the key to knowing when best to “quit while you’re ahead”. Attempting to maintain relevancy without the same spark and inspiration that originally existed, despite all obstacles, will hollow out the soul in your work and ultimately result in ignominious decline. I have not followed Joel on Software closely enough to make a guess at when he may have reached that point, but I think it may be pretty clear at this point that he has in fact done so. That being the case, he should have moved on to something new.

Because he has not moved on, it may be that his writing about software development has begun to mutate into some kind of pathetic shadow of its former glory, begging for the scraps of Internet attention primarily to pass that mindshare on to the software and services provided by his company Fog Creek Software. It has become advertising, a hip, Web 2.0 marketing vehicle, dressed up in the costumery of a programming weblog. Like Metallica, it looks like Joel has sold out — but this time, to Google PageRank and his own addiction to filthy lucre. Maybe my speculations in this paragraph are all wrong. Maybe he’s always been that way, or worse yet a mere attention whore. Maybe he’s just as relevant and insightful as we always thought he was, barring this one (major) aberration. Maybe it’s something between these two possibilities. Then again, maybe everything I’ve said is true.

If that’s the case, Joel Spolsky has nonetheless definitely not jumped the shark, and neither has anyone else. Joel on Software, on the other hand, very well may have leapt eight times the length of a football field over a collection of ninety-seven specimens of the impressive biological killing machine of the sea we know by the name Great White.

Joel, I think you need a new schtick.

Paul, let us know: Have you already found a new direction? Is the perception I saw articulated, that you’ve jumped the shark, merely a result of us not yet realizing you’ve moved on? Have you clung to the ultrahacker image so long that what you’re doing with it, whatever you may call it, has jumped the shark at last?

Say it ain’t so.


  1. I think Joel is going the route of any person who works in tech but happens to write about it as well. You must become a bit divorced from reality. Let’s face it, not everyone wants to be slinging code their whole life. However, when you get out of the trenches and dump a lot of your day-to-day hands on work (or take on additional responsibility), not only do you stop keeping up with the current state of affairs, but you gain new perspectives as well.

    I have worked with people who were still stuck in a “Pascal/C/COBOL” mindset because that is what was good stuff when they stopped doing day-to-day coding and started managing, for example. Was it a bad thing? Not really, they certainly had technical aptitude, more than enough to understand what was what. But they also approached problems from certain angles that just did not make sense in a more modern context, either.

    It really is the “when ou have a hammer, every problem looks like a nail” syndrome in action, and some people have outdated or skewed hammers.

    You are also jumping on Joel, and I am just not seeing why. The “Language Wars” post was dead on correct except for maybe the bit about Wasabi at the end. Then again, I can understand even that; 20 years ago, my father wrote his own version of COBOL to run on Wangs, just so he could then write his accounting system in that language. Wasabi probably is not ever meant for or destined for glory and fame, it is a custom crafted language designed to meet Fog Creek’s unique needs.

    But Joel is 100% right in his assesment of the languages out there. Frankly, the “Big 4” (Java, PHP, VB.Net and various VB variants, C#) are all equally lousy. Personally, I feel that PHP is the most miserable of the four (lack of tools, lousy documentation, inconsistent libraries, just to begin with), but that is just my personal opinion. I would much rather be using any number of languages instead. But the reality is, using any other language, while having great tactical advantages, and even some strategic advantages, is a career killing blunder. When the customer comes back to me and tries making a change and wonders why the software is in some arcane language, I just activated the career warning lights. For private projects, sure, use less popular languages. Indeed, this is one of the major reasons why I’m finally starting to really like the .Net Framework; it makes it possible for goofball languages to have as much access to libraries, APIs, etc. as mainstream languages. F# is a great example. In one fell swoop, there is suddenly a functional programming languages that is 1) on equal footing with VB, C#, C++, Java, etc. in terms of ability to access the OS, Internet, file system, etc. and 2) able to be used in more than a character stream mode. That is fantastic news.

    In any event… I do not think that Joel has “jumped the shark” (and I agree that people have arcs, not defining moments like “jumping the shark” implies), but I do beleive that having his nose out of the daily coding feedbag is starting to color his judgement on these kinds of things.


    Comment by Justin James — 14 September 2006 @ 06:41

  2. […] Original post by apotheon and software by Elliott Back […]

    Pingback by » Blog Archive » Joel Spolsky has not jumped the shark. — 14 September 2006 @ 08:41

  3. Good insights, apotheon. People are far more complex than “Happy Days”.

    I think Joel still has a lot of insight. That’s what made this particular thread so jarring for me. It shows that even the smartest person can make mistakes, whether that’s Joel or the rest of us.

    Comment by SterlingCamden — 14 September 2006 @ 11:56

  4. You are also jumping on Joel, and I am just not seeing why. The “Language Wars” post was dead on correct except for maybe the bit about Wasabi at the end.

    Joel was dead wrong about almost everything. I don’t know that I agree that a proper DSL can’t be used for something like FogBugz, as one detractor suggests, but I know from what he says that Wasabi (it seems) must be quite a bit more than that as a superset of VBScript with additional features added to it. From the description, his shop tried to create some kind of Lisp/Ruby frankenstein language from a descendant of GWBASIC. Let’s just assume, for a moment, that Wasabi is actually a proper DSL though.

    If that’s the case, Wasabi may be the only right decision he made in that essay. The rest involves undervaluing some excellent languages, sabotaging his argument by mentioning Wasabi in the same breath as dismissing the usefulness of anything not one of his cherry-picked Big Four Languages, and using Pythonista anti-Ruby dogma and almost useless benchmarking to bolster his arguments without bothering to properly investigate the realities of the subject matter.

    The simple fact that he declares Perl a non-starter for web development pretty clearly marks him as an idiot on this subject. If everybody who uses Perl as a web development language dies a messy financial death, I guess Amazon must not be using Perl/Mason, and Slashdot must actually be using C#. Hell, maybe PerlMonks is actually written in Java with J2EE.

    I have a theory, but it’s a tentative one without any evidence: I think that someone on his development team did indeed develop a language called Wasabi. That language is a proper DSL. It’s excellent for the task to which they put it. It is not a functional dialect of VBScript with closures, et cetera, as he claims. Having turned into Mr. Manager with exactly the disconnect from real-world coding that you describe, he’s throwing buzzwords at the wall to see what sticks.

    But the reality is, using any other language, while having great tactical advantages, and even some strategic advantages, is a career killing blunder.

    That’s not the argument Joel Spolsky was making. It’s not even close. It’s also only true in certain circumstances — in other circumstances, it’s 180° in the wrong direction. I’m not even talking about “private projects”, either.

    No, he wasn’t “dead on correct” about everything but Wasabi. Wasabi might be the only part that he got right — if he was wrong about what Wasabi actually is.

    Comment by apotheon — 14 September 2006 @ 12:40

RSS feed for comments on this post.

Sorry, the comment form is closed at this time.

All original content Copyright Chad Perrin: Distributed under the terms of the Open Works License