Nick Malik has posted another thoughtful piece on the whole Mort topic that's been going around these past few days. While I appreciate where Nick is coming from, I can't agree at all with his perception. Ayende also voiced his opinion on the topic here.
Trying to control and bound a developer and intentionally doing everything possible to keep him blissfully ignorant about the tools he's working on has not proven very successful so far, and frankly, I think it's not only a bad idea (as in, it doesn't work at all, abstractions are leaky and all that), but I also find the idea incredibly depressing and dreadful.
Here's just a few things that are incredibly wrong with this whole idea, imho:
- It is a self-defeating attitude: It means you stop focusing on improving the situation and purposefully keeping it in a bad, unsustainable position.
- It is really elitist, far more than any of the times the whole ALT.NET idea has been brought forward. It essentially means you just split the developer profession in two mutually exclusive, segregating communities (casts?): Those that know and do, and those who don't know and do.
- There's no way to cross the bridge: Once you've started as one of these Morts guys, you stay there. The whole proposition is based on the idea that Mort has no need to learn anything new on the job (and possibly didn't even need to know much to start with). Indeed Mort should not try to learn anything or become smarter lest he tries to work around the walls of his prison (thus causing damage).
- This, in effect means that once a Mort, always a Mort. Mort is not likely to look for a new work place, since he'd only be able to work as Mort anywhere else, anyway. No need to keep working hard trying to attract and keep those pesky developer types, whooohooo!.
Incidentally, this is why this whole idea is the wet dream of most project managers and IT shop owner's I've met: it promises cheap, untrained, replaceable labor, and low cost to ensure high personnel retention. Goodness gracious great balls of crap!
Now, is it good to try to make it easier for the developer to focus on solving the business needs? Absolutely. Is it good to try to raise the level of abstractions we work for so that we can concentrate most of the time in the parts of applications that actually matter? Again, absolutely agree. But making the leap from here that the obvious end result we should aspire to is to this? I mean, if that's the best we can do, we might as well just abandon all hope and give up.
An interesting aspect of Nick's proposition is that it is indeed a win-win situation for a company like Microsoft, while I'd say pretty much all IT shops (partners) in the MS ecosystem would come out loosing. See, in Nick's ideal world, Mort would just drag and drop till he drops (sorry, couldn't resist :-)), while someone smart would create the tools and services he use. The next proposition in line would be that those tools would be done by MS (of course) and the services well, by none other than MCS [1]. Partners around could put in those Morts instead of supplying development talent, but, well, what do you know! Since Mort doesn't really have any special skill set (particularly none that's very software specific), well, the business themselves might decide to supply it! [2]
And now we've just gotten round all the way back to the old idea of CASE tools and 4GL languages.
[1] I have no grudges against MCS; indeed I know many people working there. It's just the logical conclusion :-)
[2] This would also likely mean that those lucky enough to actually work on the Developer side would probably be able to get paid a lot more for their work, being a scarce resource and all...
