thirdbIT

Joel Spolsky recently posted the transcript a speech he gave to the Yale CS department on November 28th. It reads a bit like an extended recruiting pitch, which, of course, it is. Because why would he give a talk at Yale CS and not attempt to pick up some smart people to work for him while he was there?

Anyway, one section of the talk is all about how it sucks to write in-house software. Joel says it sucks because, first, you don’t get to work with cool technologies, because no one will let you:

you’re not going to allowed to build things with Ruby on Rails no matter how cool Ruby is and no matter how spiffy the Ajax is going to be. You’re going into Visual Studio, you’re going to click on the wizard, you’re going to drag the little Grid control onto the page, you’re going to hook it up to the database, and presto, you’re done.

Second, it sucks because you don’t get to work on a project past the point where it minimally solves the business problem at hand and because of that, you’re stuck building and fixing ugly, broken things for not enough money and no respect.

The key point about in-house development is that once it’s “good enough,” you stop. When you’re working on products, you can keep refining and polishing and refactoring and improving, and if you work for Facebook, you can spend a whole month optimizing the Ajax name-choosing gizmo so that it’s really fast and really cool, and all that effort is worthwhile because it makes your product better than the competition. So, the number two reason product work is better than in-house work is that you get to make beautiful things.

Third, it sucks because no one thinks in-house developers are very important:

Once at a Viacom Christmas party I was introduced to the executive in charge of interactive strategy or something. A very lofty position. He said something vague and inept about how interactivity was very important. It was the future. It convinced me that he had no flipping idea whatsoever what it was that was happening and what the internet meant or what I did as a programmer, and he was a little bit scared of it all, but who cares, because he’s making 2 million dollars a year and I’m just a typist or “HTML operator” or whatever it is that I did, how hard can it be, his teenage daughter can do that.

Therefore, says Joel, serious programmers should work at software product companies, like, say Fog Creek Software.

I’m sure a lot of in-house software development jobs are just like Joel says. But there are plenty of really rewarding in-house positions available. For example, what he describes is not at all what I experienced while working at Millennium Pharmaceuticals (2001 -2003, with some consulting work on and off after that). Informatics was pretty crucial to the company’s vision of how it would revolutionize drug research and development. Developers were respected, and we tried to make our software, if not very beautiful, at least not completely butt-ugly. There was room for plenty of innovation in what technology we used. It was a great place to work, and I worked with some amazing people there. I felt like the software I was making was actually helping people do their work, and it was work that they cared passionately about, and I understood how they could be passionate about it, even if I wasn’t myself passionate about it. Most of the code I wrote went to helping the molecular pathology department manage their experiment workflow. I thought the domain (biotech, and as far as what molpath was doing, basic disease research) was interesting. I was really happy there.

Conversely, I’ve found that I’m not really at my happiest working on consumer products. I don’t get excited about consumers. I think that their consumption is pretty much the least interesting aspect of people, and I think that when people are acting as consumers, they tend to be at their most demanding, irrational, and childish. I also don’t think that people like that Viacom executive who had so little respect for Joel are found only in big companies with big in-house software teams.

I like to help people get their work done. I like the philosophy of ‘good enough’. I don’t really care that much about ‘really fast and really cool’ unless someone convinces me there’s a very good reason to care about it, and the reason is better than “well, people will think it’s cooler and buy more of it.” I’m just not a very good capitalist that way.

In any case, plenty of software product companies make butt-ugly software that’s only just good enough, using awful techonology, so I really don’t think the distinction is a valid one. And I think that we’ll see that an increasing proportion of in-house software written in ruby on rails. Maybe it won’t have cool AJAX livesearch widgets in it. So what?

It’s not that I think appearances aren’t important. It’s not that I don’t think that there’s a place for consumer software products. And I know that lots of in-house software development sucks. But I think Joel is over-generalizing in order to make a stronger recruiting pitch. And I think that he wrongly assumes that what makes him happy is going to be what makes everyone else happy. That’s true at the level of basic human needs, of course, but not at the level of what floats your boat at work. Or rather, not in the way that Joel talks about it. Your satisfaction at work probably has a lot more to do with the length of your commute, whether you have clear goals and a sense of control, and if you have friends in your workplace than about whether you get to work with cool technology or make things faster and better just for the hell of it. Respect, sure, that’s important to have. But plenty of software product companies respect sales and marketing over development, and plenty of non-software-product corporations respect their developers.

Joel’s not really wrong here, of course. He just over-generalizes. He thinks his own experience must be everyone’s experience. He is sure that what he wants must be what everyone else wants.

Actually, it’s probably not that at all. He just really, really wants to pick up the cream of the Yale CS crop. He needs good developers, and gosh darn it, he’s gonna make his pitch to get them.

I broke down and did some AJAX crap on our big rails app. Cool, I thought. I’ve got my drag-and-drop lists with acts_as_list and my in-place-edits. Don’t we look all sexy and modern? Except in_place_edit_for doesn’t bother validating changes to the model. Someone wrote a custom implementation which does do validation but I haven’t tried it out. I just yanked out my in-place-edits for the moment. I’m irritated. Gotchas everywhere. Oh, and I’m having routing issues today too.

On the plus side, I did solve some troubles I was having with Attachment_Fu’s thumbnailing. Though it’s choking now on destroy (thumbnail-delete troubles, I think I read about those somewhere). Full post on Attachment_Fu to come when we’ve ironed out all our weirdness with it.

Found on Slashdot, a good article from Dr. Dobbs’ Journal comparing available AJAX frameworks.

Oh, man. AJAX.

So I’m working through the canonical rails text, Agile Development in Rails, and I’m breezing along, thinking “this rocks!” The last web app that I wrote, back in 2002-ish, was a Java/JSP thing, with some web services thrown in, toward the end. I had some exposure to Ant, JUnit, and the other open-source tools people were starting to use then, but not a lot. STRUTS was still bleeding-edge.

Anyway, so that web app I wrote at Millennium, for the Mol Path department, was an enormous undertaking, at least for one not-very-experienced developer. Legacy code, legacy database, ties to 5 or 6 other systems, disgruntled and demanding users, the works. I worked my ass off on that app, and most of the functionality of most of the code I wrote is now provided, from the start, in Rails. I’ve been breezing through the rails tutorial, excited as all get-out to think how much more quickly one can develop web apps with rails. It’s all awesomeness.

And then I get to AJAX. And I cringe. Because ye gods, I hate javascript with a fiery passion. I hate the DOM. And yet, if I want to write code, and I do, desperately (I’ve done doc-only gigs, and it’s not my favorite thing), then I’ve got to know the DOM. I have to fiddle with the CSS, make the shopping cart update on the fly, zoom the links, and all the rest.

Why is javascript still around? Why doesn’t someone, you know, standardize it? Why can’t it work the same way in all browsers? Why does IE have a “quirks mode”? I have long been convinced that javascript was invented to drive programmers insane. It looks sort of like a programming language, but actually it’s voodoo. I know with Rails I get to stay pretty far away from the actual javascript , thank goodness for small mercies, but still, not as far away as I’d like to be. And again, why is javascript still around? Still around, and used more now than EVER?? Grrr.

Still and all, Rails is damn cool.