Monday, August 20, 2007

Numbers rocks: how I forgot about the review and ended up doing my budget

Apple has made a trial version of the iWork suite available as a free download. Pretty smart move: the suite is relatively small (it fits on a CD), so this is a great way to get people test drive the latest version of this emerging little office suite. (Let me get back to the "office suite" part later.) You can buy an activation code online to unlock the trial version, so basically, Apple is distributing iWork '08 as shareware. Cool.

I've put Keynote and Pages through their paces, and they're OK. But what I've been most interested in was Numbers. Why? Here's a list why:

  1. It's new. Duh.
  2. It's a spreadsheet app, and those are relatively rare. Word processors are a dime a dozen.
  3. I wanted to see if Numbers is competitive with Excel.
  4. I work with data a lot (Excel, FileMaker, and so on), and wanted to see if this new tool is of any use for me.

As anecdotal as it gets, but still, wow...

So I fired up Numbers, and started off by using one of its built-in templates. I noticed one that was called "Budget," [edit: originally I wrote "Home Budget," not sure how I'd got that wrong] and thought, what the hell, let me try that one.

I've been putting off drafting an annual personal budget for quite some time now. I was looking for the right tool for the job. Now, it's important to know that I'm a tool freak. (Also a Tool freak, but that's probably beside the point.) That is, I can obsess about the right tool for the job a lot more than about the job itself. It's almost a policy. Yep, I know this can be a flaw. But not always.

Anyway. So far, I've tried creating FileMaker Pro databases, using and extending Excel templates, and I've always given up after a certain point. Building a FileMaker database is almost like writing an application: you need to do a lot of work before you can start using it. About Excel, I just didn't know where to start. The built-in templates weren't much use, and as for rolling my own: the task seemed a little too intimidating. Before getting on with the already daunting task of drafting a budget, I would need to decide on how many worksheets to use, what kind of tables to design, and how to interconnect them, etc. I'm not bad with Excel, but whenever I embarked on this budget project, I must admit that I always ended up giving up.

So, last week, I fired up Numbers, and opened its Budget template. It was pretty straightforward, I just about immediately figured out how it worked. And to my utter surprise, it was almost exactly what I needed. I made some small adjustments, and started putting in my numbers. Then I made some more adjustments to the template, consulting the help file two or three times.

After about five hours, I was still frantically, furiously working on my budget. I was sweating, but what I was fighting was my numbers, not Numbers. I didn't even notice the app was there.

And that's just about the best thing you can say about an application. It gets out of the way, and lets you do your thing. Oh, and the template is very nice, too. Maybe that's where Excel lost me on this one, and Numbers won me.

These are my first main observations about Numbers

Numbers doesn't have one workbook with several infinitely large worksheets. It has pages with tables, which are the size you want them to be. This doesn't only make your numbers nicer to present, but also makes it easier to work with them: you can see all the tables at the same time, you don't need to switch between worksheets.

If you mouse near the border of a table, some controls pop up. You can insert, delete, or drag and drop columns and rows, you can sort and rename columns, and so on. These operations are extremely intuitive, though really mouse-heavy: there are no keyboard shortcuts for most of these. Working in Numbers feels a bit more like working in InDesign, and less like in Excel, where you can let go of the mouse for quite some time if you want. To me, this is a clear shortcoming, but a tolerable one.

Numbers is very good with defaults: it knows that most users will want their tables to have headers; and that when you sort, you'll usually want to sort the entire table. (This is a pet Excel peeve of mine: using auto-filters, you can never be sure if your entire table is being sorted. Some pesky little thing can prevent some columns from being sorted, and you'll end up with useless data.)

While sorting is dead simple, there are no auto-filters in Numbers. Filtering is dialog-based, and clearly more cumbersome than in the Microsoft suite. Also, the only way to tell if your data is filtered is by noticing missing row numbers. Excel has other visual clues, and they are important. Probably Apple's research shows that people don't really use filters that much. It's a pity, because I do.

Tables can have headers by default (there are several table templates you can choose from, but you can fully modify a table after creation), and they can also have titles (captions). These are great time savers as you add and arrange new tables to your work area (called a Canvas).

Numbers makes sure your spreadsheets are neatly organized and beautiful. Just like a great schoolteacher, it will instill a sense of work ethic in you, inviting you to keep your work clean and well-organized. (Don't use Numbers for committing tax fraud or plotting evil schemes. You will break down with guilt and give up.)

One annoying bit: as you move or resize a table that has another table on its right side, Numbers will always move that table too, keeping the distance constant between the two, even if that's not what you want. (Thanks to the reader who pointed out that this behavior can be turned off in Preferences.) And believe me: you will care about how your tables look. Numbers will make you.

Cross-references between tables and cells are quite like in Excel, except that they use column and table names, not numbers. Luckily, these names update dynamically.

There's a generous helping of functions, and for obvious reasons, they have the same syntax as in Excel. Not nearly all of Excel's functions are present, though. Worse, I've been relying heavily on Excel's extensive help system when constructing a function: as you type, it displays the syntax for you, and mousing over each part will show you additional details. It's very easy to get specific help for each function. Not so in Numbers: you're pretty much on your own, and help is awkward. Functions are probably also considered an advanced feature that relatively few users would be interested in. Hopefully, Apple will beef up this part of Numbers for the next version.

There is one very useful feature, though, that immediately made me a fan (that is, if one can get fanatic about spreadsheets). Select a few cells in Excel, and the app will display the sum of all the numbers in them. Numbers takes this concept a step further: not only does it display their sum, count, average, minimum and maximum, but also lets you drag these to your table, as a really quick and easy way to create a summary field. Well done, that one!

Bloatware vs. clutterware

So Lasso is here, and it's sexy indeed. But does it take on Excel? Well, yes, and no. Excel has macros (though you'll have to kiss them good-bye soon, as they will be absent from the next Mac version.) It also has tons and tons and tons of advanced features that Apple did not include in Numbers.

Now, there will certainly be people who dismiss all these tons of Excel features as "bloatware," but I will certainly not go down that road. I'm with Joel Spolsky here: he believes that the size (and the complexity and the feature count) of applications increases as do our needs. He also gives us his spin the 20/80 rule, i.e. that while it may be that 80% of users use only 20% of the features, it's not the same 20% for everyone.

I do believe that software can be too complicated and intimidating (and Microsoft Office is certainly like that). However, that doesn't have much to do with the number of features, but rather with their presentation. I would rather call it clutterware than bloatware. Features are necessary, but throwing them all at the user in a big scary mess is wrong.

For a version 1.0 release like Numbers, Apple did have to narrow its focus on the most commonly-used features. However, here's hoping that the scope of Numbers will grow in time. And knowing Apple, I'm fairly confident that Numbers will never become clutterware. Bloatware maybe -- but, as Joel says, that's actually a good thing.

Is iWork an office suite then? It would probably be an inaccurate moniker, and one that Apple seems to want to avoid (never calling it an office suite, going with "productivity suite" instead). This has to be at least partly due to a marketing effort that carefully tries to avoid the appearance of competing directly with Microsoft. But marketing materials, as well as iWork templates, also clearly indicate a focus on the home, small business, and educational markets, Apple's traditional strongholds. Besides, large corporations would need collaboration features clearly missing from iWork.

I wonder whether Apple will, over time, address the corporate market more aggressively. We can say that, with iWork '08, it's on the doorstep, but not yet knocking.


Thursday, August 09, 2007

iMacs ditch iPod look for iPhone colors

When the iMac G5 was released, its slogan was "From creators of iPod: iMac." I wasn't sure if the irony in that line was intentional, but it was pretty rich: how much has the world turned that the iMac brand, once a cultural icon, needed to be propped up by the popularity of a mere MP3 player?

But now, the iMac looks like the iPhone. It's silver and black, and its screen is made of glass (with its glossiness touted as a feature, which makes me shudder). Its screen has a black border. Even its iSight webcam looks a bit like the iPhone's single button.

Is this a new design trend for Apple's consumer products? Until now, Apple's professional hardware was very visibly made of aluminum, while most of its consumer products were black or white. The only exceptions were the colorful, metallic iPods (nanos and shuffles).

The iPhone is a brand new product category, and it comes as no surprise that its color scheme defies that of Apple's other products. But now the iMac is following suit. This can mean one of two things in my opinion, and only time will tell which one is the case:

1. Apple is shifting its design policies yet again, moving away from white/black plastics for consumer Macs. Such shifts have happened in the past, just think about the various color schemes for iMacs and iBooks (from bondi blue to various selections of multiple colors to psychedelic patterns to an elegant white) in the late nineties, or the similar changes in the professional line-up from the same bondi blue to various shades of graphite, and from black to titanium to aluminum in the case of pro laptops.

The color scheme has never been as straightforward and consistent as it was recently (until the arrival of the new iMac), we have to wonder whether Apple is giving up some of that consistency. Future revisions of the MacBook line will be a certain indication.

But luckily, I think every major color scheme transition so far has pointed to a classier direction (if we disregard the few last revisions of the G3 iMac).

2. The iMac is no longer a consumer Mac. Well, one can argue that it's now a pro-quality machine, or simply one that's sitting on the fence dividing the pro and consumer categories. Its software bundle doesn't contain any specifically consumer or pro apps, but then neither does any current Mac's: Tech Specs pages indicate that Apple seems to have quietly stopped bundling third-party apps with Macs.

In any case, the iMac seems to be a mighty machine. Some further random observations:

  • I wonder what the keyboard feels like. I have some doubts. Apple hasn't always exactly been a champion of input device ergonomy.
  • The mouse now looks out of place. Will there be one made of aluminum too?
  • Apple's digs at Windows PCs, specifically Dells, are back. The comparison photo is particularly mean.


Monday, August 06, 2007

Random anti-bad-idea post:

I learned about this a while ago via Daring Fireball. John Gruber was endorsing the idea.

Basically, Mike Davidson complains that answering even short e-mail questions may take very long, and also that as he gets too many e-mails, he tends to prioritize the easiest ones to answer, not the most important ones.

Don't we all. But his solution makes me shudder. He now always responds in five sentences. He wanted to do a word or char count, but that's difficult, so he arrived at counting sentences. He then posts a short Q&A at the end of all of his messages ("Q: Why is this e-mail five sentences or less? A:") linking to a little explanatory website he operates, where he sums up his policy.

What's wrong with this? Everything. First, writing short messages is not always easier than writing longer ones. You'll soon find yourself fighting the tool. Second, why be so damn restrictive and dogmatic? Couldn't you just strive to spend less time on answering e-mails? How about a timer? Or nothing at all, just an effort to keep it short? Maybe you'll average on five sentences, maybe you won't, but why this "one size fits all" approach?

And don't even get me started on the sociopathic explanatory link at the end. If your correspondants are insulted by your brevity, a pre-recorded explanation will only rub salt into their wounds. Guess what, you've been handled by a policy. You're not so important.

My suggestion: strive to spend less time on answering e-mails. If some of them end up too brief as a result, add this sentence at the end: "Sorry to be brief, I'm really busy."

No weblink or clever URL, though.