Thursday, June 30, 2005

Big-endians vs Little-endians

The Lilliputian people was broken apart between the Big-endians who believed that eggs should be opened at their big end and the Little-endians who believed eggs should be opened at their little end.

In a 1980 paper (On Holy Wars and a Plea for Peace), Daniel Cohen used this part of Switft's Gulliver's Travels to express his opinion about the byte ordering question.

What does this have to do in a blog supposedly about i18n and l10n ? Well, this story is related in Yves Savourel's book XML Internationalization and Localization.

Wednesday, June 29, 2005

Belgacom == concurrence ???

[ If you're not Belgian, you're probably not interested in this post (in French)]

Lu dans la Libre ce matin, à propos de Belgacom TV:

"[...] L'arrivée de Belgacom va faire souffler un nouveau vent de concurrence."

Donc, Belgacom rimerait avec concurrence !? Ce n'est plus la 4è dimension mais au moins la 37è...

Tuesday, June 28, 2005

Why is there no .NET support (yet) in appTranslator ?

I sometimes receive e-mails from people telling me that Win32 apps are so 20th century ! Why doesn't appTranslator support .NET ?

Well, first of all, appTranslator doesn't support .NET yet.

Who are my clients ?

When I started to work on appTranslator, I was wondering who my clients would be. And I figured that software that needs to be translated is more likely shrinkwrap software rather than consultingware or internal software. I believe .NET is not yet the platform of choice for this kind of software (although I admit this is changing very rapidly). Reasons of course include the overhead of the .NET framework redistribution/deployment. But it's no longer the main one.
Starting-up companies don't usually care much about translation. They have work enough getting their product ready and filling their local market. After some time, their product is getting better, they start thinking of hunting further from home (understand: abroad) and realize they need to translate their software. Which means that even if .NET is becoming more and more widely adopted for shrinkwrap software, the people looking for a localization solution have a product whose development started probably a few years ago, hence is not likely .NET based.

What am I good at ?

There is another reason. This one is probably even more important. When you start-up, you basically need to do everything in your company, from sales to support, from development to marketing, and I wish that would be it. And of course, you can't be good at everything. I, for example, am a developer. I'm very used to managing projects and designing software products. But even though I believe I have good presentation skills, I'm not a sales guy. And marketing is not what I'm used to do.

Since I know there are areas of my occupation where I'm probably not very good at, I'd better be as good as possible in areas that are my strong points. As far as my development skills are concerned, I'm way better at Win32 than at .NET. It's just a question of experience.
So it's undoubtful that I would do a better job at solving a problem for Win32 apps than solving the same problem for .NET apps.

That's why I focused on Win32 in version 1.0 of appTranslator.

OK, but will appTranslator ever support .NET ?

OF COURSE !!! And better sooner than later.

OK, don't make me say what I didn't say! .NET support is not coming next month. Neither in August. But it will come. Yep. Promised.

Sunday, June 26, 2005

Please stop the Wimbledon championship !

Or turn it into a bowling tournament. Or invent tennis without a ball.

Because there's life in there !

(Source: AFP)

Thursday, June 23, 2005

Effective toolbars are short

A couple of weeks ago, I was demonstrating appTranslator to a potential client. That client was using a competitor's product. The kind of product which has a load of features but horrible usability. For example, it has several toolbars totalizing close to 50 buttons. Yes, Ma'am: FIFTY buttons !

Long toolbars are counter-productive

At some moment, I showed him a button and asked what was the function. The guy didn't know. Then I asked him if there was a button to do function X. He didn't know neither. Actually, the button I showed and the function I asked are among the most important UI functions in a localization tool. And the guy had no idea how to use the toolbar to trigger these functions.

The point is that toolbars with too many buttons are worse than useless: They are counter-productive. Don't think that adding buttons to a toolbar is harmless. You may think that users who don't know what a button does can simply ignore it while power users have more shortcuts.
WRONG! Adding buttons to a toolbar contributes to hide the important ones into the mass. Result: The buttons that should be obvious for every one are no longer obvious to anyone.

Two kinds of buttons

Keep your toolbars short. Don't add too many buttons. Because you'll quickly hit the limit of the icon paradigm: It's almost impossible to make a self speaking 20x20 pixels drawing. And the more you add such drawings, the more they will look like one another, making it even more difficult to distinguish which one is which.

Actually, the only buttons that users are able to recognize are the ones they are used to recognize because they see them in most programs they use: Open, Save, Cut, Undo,...

Besides these buttons, you can add a few ones for functions specific to your app. But remember that users won't recognize them. Some will fly their mouse over these buttons to see if there is an explanation (tooltip). They may even remember one or two of them but not much more. That's the best you can expect.

If you really want to please your power users, add support for customizable toolbars. But don't expect many people to use that.

The appTranslator toolbar contains 10 buttons, 8 of which are the most standard buttons you can think of. And the remaining 2 are usual appTranslator commands (Build and Dialog Layout Mode). And I'm pretty sure almost nobody uses the Build button.
There are also two comboboxes whose contents should hopefully be fairly obvious. Because it's easier to make self explanatory text than icons. After all, the default layout for the Internet Explorer toolbar displays 'Search' besides the magnifier. Which means that people don't associate the magnifier with the search command.
Comboboxes fit very nicely in toolbars: They allow users to make choices in text lists, which is easier that choosing a button among 3 or 4 which all look like each other.

By the way, FogBugz doesn't have toolbars at all. Only menus. And frankly, I'm happy with these menus. Why ? Because even if my bug tracking system is opened most of the time, even if I use it several times per day, I never spend hours in it. So there's no chance I would remember what its icons mean.

Tuesday, June 21, 2005

I've got a new friend !

Do you remember I mentioned Michael Kaplan's blog some time ago.

Today, I learned I'm his good friend! :-)

Wait! appTranslator is even on his radar! [ Well, OK, he and I know that I rarely miss an opportunity to ask for a link. And, no, I won't tell you that story !]

Michael even plans to blog about my product ! Is it Christmas already ? Or should I wait until I read his post before being happy ? Nah... No risk ! You know why ? Because MichKa and I are good friends ! ;-D

BTW, the amount of exclamation marks and smileys in this post indicate that I'm happy ;-)

Thursday, June 16, 2005

How to make usability a selling point ?

A few days ago, I asked that question on the Business of Software forum hosted by FogCreek.

(I admit my post sounded somewhat baby crying. I guess I wasn't in very mood).

The discussion that followed was really interesting. People made various very good points about software (and usability) in general. I'm too tired now to make a summary and pack everything into one (or more!) articles, so I let you read the discussion ;-)

Tuesday, June 14, 2005

Do you want spam in your own language ?

I just ask because info at received its first spam today. I thought that if spammers know about appTranslator, they could as well decide to use it! :-)

Friday, June 10, 2005

How do you decide if a product is worth evaluating ?

This is very important. Managing to get people to your site is not easy. But once they get there, you don't want them to leave immediately. You want to make sure you keep as many people as possible. Because if they go away, so do your sales. :-(

You managed to attract people to your home page. You'll have to drive them to your download page (if you have a demo version. If you're selling shrinkwrap software, you have a downloadable demo version, right ? No kidding! A couple of days ago, I found what looked a great little tool. And guess what! No demo! Just a money back guarantee. Come on!)

Getting people from your home page to your demo

How do you get people to download your demo version ? Obviously, you have to convince them that your product is very probably what they are looking for. Which brings us back to the question in this post's title: How do people decide if a product is what they are looking for ?

  • I read the overall description.
  • Then I probably browse through a features list if there is one.
  • I may give a look at screenshots, just to see how professional it looks.
  • And of course, I look at the price ! No need to spend time evaluating something that I can't afford.
That's why the pages in my site's menu are ordered that way.

And that's why it's been some time now that I spend every morning telling myself that I must absolutely rewrite my appTranslator overview page because it's way too wordy and not appealing enough.

How do people think ?

And I of course do a big mistake: I figure people think the way I think.

Earlier today, I gave an in-depth look at my site's log. Big surprise ! The average visitor goes exactly the other way around :
  1. Home page. (OK, that's where they enter the site)
  2. Screenshots.
  3. Price list.
  4. Download.
  5. Features.
  6. Overview.

Wait, even the short page with the differences between editions gets more hits than the overview !

People don't read

Steve Krug's excellent book (Don't Make Me Think) explains that users don't read, they actually scan (or was it Joel Spolsky ?).

You know what ? He's damn right ! Not only people don't read, but they perfectly know where to click to avoid reading... Pages 2 to 6 above are ordered ascendingly by number of words !


  • What you think is probably not what people think.
  • Visitors don't read.
  • Don't ask Jeeves, ask Urchin !
  • Do screenshots.

Oh! BTW, there's something else I don't quite get: Why do so many people rush to the download page but don't actually download the program ? What do they expect to find there that they don't find ? Freebies ? I don't know...

Thursday, June 02, 2005

Microsoft opens the Office docs file format

Well, not the current format(s). But they decided to create a new XML-based (what else!) format.

No scoop here, this probably the fastest spreading news in the history of the blogosphere.

I didn't read much about it but I would be surprised if nobody at MS thought that it's nice to have what could eventually become the killer argument for making people upgrade their copy of Office.

[Through Ian Landsman]