Often times, management that doesn't have much experience with technology will take the path of "hack" before "track" in their development, which I don't agree with in the least. You can't build features on bad infrastructure. Anything you build on a faulty foundation will become almost completely worthless should you ever change the underlying layers. So what does this tie back to? Well, a while ago, I talked about delivering an actual product versus consulting and hacking in a solution.
When I settle in on a project, it needs to be done right. It's great when you get something that works, but in the long run, you need to finish it. There are manuals, documentation, standards that are a bare minimum necessity of handing something off to another person. So then, what exactly does a product "have"? In my mind, something you have as a deliverable needs to have at the very least a complete product. All facets of the item need to be complete, and the entire item must have a polished look and feel. The second piece of a finished product is installation information. If it is a single exe, a setup file, or something naturally intuitive, that's fine. If it's a list of copy and configure instructions, that's okay too. Even a readme.txt, install.txt, or instructions.doc would be of value. If you don't know how to install the product, how do you install it again if you ever need to? The final thing any product should have is usage documentation. How does it function, how does it work, what does it do? Most every product comes with some kind of manual detailing how to carry out certain tasks. These three things make for a complete deliverable product. I had to supply such things for my Senior Project, and I supplied them when I was at ASAP.
So then, if you are someone working, and management gives you the choice to do it completely or to do the minimum, what are you to do? Accept management's decision to do it wrong, or go for broke and try and make it as perfect as possible? I know the answer is probably somewhere in the middle, but I can't help but aim for one of the two outcomes. If it's the low bar I would target, I do think I would try to make sure it works on future portals, and if it was the high bar, I'd probably skimp a bit on the documentation. I hate leaving things unfinished. It isn't something I like doing.
I'm also moving this weekend. That should prove to be interesting.
In response to "Programming a Product":