Many people don't understand just how scalable software is.  Computers are really good at performing simple tasks over and over.  They don't get bored.  They don't quit.

So let's look at some hypothetical software and talk about how difficult it is to do different things.

Let's say you're building a site that lets users publish blogs. To publish one blog (such as this one) you need a table to store the posts, and another table to store comments.  I've got more bells and whistles, but those two tables store just about everything.  So I'll assume that this hypothetical software only has those two tables.

This means that the software supports exactly one blog.  You can add as many posts as you want but there's no way to differentiate between multiple blogs.  In order to support more blogs, you need to add another table that stores information about each blog which links to the individual posts.  What I'm saying is that there is a fundamental difference between supporting one blog and supporting two.

Those of you that have experience programming know that adding the extra table takes about 10 minutes.  It's hardly any work at all, but there is work involved.  The thing is. once you have a table that stores information about the individual blogs, you can add as many entries to that table as you want.  There is no additional development work necessary to go from supporting two blogs to supporting three (or four or 5000).

I mention this because sometimes the way software is distributed makes it seem like it's difficult to duplicate certain things.  Maybe you only get to add so many user accounts or you can only perform an action a certain number of times before you have to pay more.  This is almost always an artificial limitation imposed by the software company as a way to get you to pay more.  Don't get me wrong, these companies have every right to use whatever fee structure makes them the most money, but we as consumers need to be more aware of these practices so that we know when to call BS.

I first thought of this because the Palm Pre application launching screen only allows for three different pages (or whatever they're called).  You can add links to apps and contacts anywhere you want on these three pages, but you can't make a fourth.  For a device that is generally very well designed, it seems crazy that they would allow more than one page without allowing as many as I want.

Notes:  Because at least one of you works at Facebook, I should say that scalability is an issue when you have millions of visitors every day.  Really high traffic websites can't just scale infinitely, but this isn't an issue for 99.9% of software out there. 

More Notes:  Sometimes making the jump from one to two doesn't actually require any work (as was the case with a project I recently had to write specs for).  Most of the time there is a little work.

I just got a Palm Pre.  It's my first smart phone and I couldn't be happier with it.  I'm going to try to avoid writing about why I like the phone itself and instead focus on why the phone is a big step forward for technology in general.

First, a disclaimer:  It's impossible to talk about the Pre without comparing it to the iPhone.  When I favor the Pre in this post, I'm not necessarily saying that you as an individual would like a Pre better than an iPhone. I'm not trying to criticise people that are happy iPhone owners.

With that said, something bothers me about how most companies are operating these days.  Everyone seems to try so hard to "synergize" their various revenue channels that they forget about making the best product they can.

For example, the iPhone isn't just a great phone.  It's a way for Apple to sell more songs on itunes, make money from the app store, incentivize users to sign up for the ridiculously overpriced MobileMe service, etc.  For the most part, the iPhone seems great and I certainly don't fault Apple for trying to make money, but there are a lot of improvements that haven't happened because it wouldn't fit with Apple's strategic positioning.  Apple generally takes a very closed approach with all of their products for this same reason.

Verizon is famous for disabling features on their phones.  Some of their phones don't sync automatically with Macs because they have a deal with Microsoft.  Some of their phones have the GPS or bluetooth disabled for reasons that I can't even understand.

And then there's Palm.  Palm is a failing company.  They needed to swing for the fences with this product.  It almost seems like a team of fancy-pants executives were sitting in a room trying to figure out how to leverage Palm's corporate strengths (or some other weird business jargon) and some young upstart suggested something crazy: "Let's just try to make the best phone we can!"

I'm sure this suggestion got some laughs originally but when no one had any better ideas, they gave it a shot.  It sounds stupid, but this is the first product I've purchased in a long time where it really seems like they just tried to make what consumers wanted most.  Here are some ways in which the Pre seems to have ignored all the dumb business stuff and instead focused on the user:

• You can plug the palm into a computer and it looks just like a USB hard drive.  You copy files to the pre and they're available on the phone.  This includes moving any MP3 file into the "Ringtones" folder.  They don't mess around with DRMs or trying to partner with a desktop media app.
• You can sync with most accounts that people actually use: Gmail, Google Apps, Microsoft Exchange, Facebook, iTunes, Amazon (for downloading music), photobucket, AIM...
• You can open the phone up.  This may sound unimportant but it means that third parties can make hardware upgrades for the phone which isn't an option on the iPhone from what I understand.
• GPS, bluetooth, wifi, etc. all seem to be fully enabled with no restrictions.  Sprint normally charges users to use their GPS program but the Pre comes with a pretty solid one for free.
• There is a normal 1/8" headphone jack.  A lot of phones have some proprietary jack so that they can sell extra hardware.
• The pre uses an open format for the charging cable (micro-usb).  I've read that you might need a proprietary cable to actually sync data with a computer, but being able to charge is enough for me.  Everything is syncing over the air anyway.

On Tuesday I talked about how I don't like modal dialogs.  Here's one way to avoid using them when designing an application.

I learned this one from Gmail, but there are other applications out there doing the same type of thing. Imagine if someone accidentally clicks the "delete" button in your application.  The results could be disastrous.  That's why most websites pop up a window saying "are you sure?"

The problem is, most of the time the use wasn't clicking the delete button by accident.  Most of the time the user actually wants to delete something and it's annoying having to deal with the confirm dialog.

With most important decisions, gmail just goes ahead and performs the action.  Then they pop up an unobtrusive message at the top of the page allowing the user to undo the action.  This makes sure that accidental clicks don't result in catastrophe but the user isn't inconvenienced in situations where they're doing everything correctly.

Sounds simple enough, but it's far from standard practice.  Sometimes it's not necessary to go through the effort of undoing things, but other times it can be a huge time saver.  If you remove enough minor inconveniences, you'll end up with great user interface.

I'm trying to incorporate not only this specific strategy at work, but any design ideas that follow the same line of thought.  For example, we will be redesigning one of our sign-up processes soon and I've already figured out one way to make it much easier.  Rather than showing the user every single option they can customize during signup, I think we'll just show the basics and then the advanced users can go in later and make changes that the majority of our users don't want to deal with.

Once again, this is an obvious idea which makes it all the more important to implement.