How big are your users’ screens?

Building a web site? One important question, if you’re developing (as most do) to a fixed pixel size, is how big your users’ screens are. Not just the average user, either; you should keep at least 95% of your potential audience happy. Even 1% of visitors annoyed is a risk: are they the ones deciding if they’re going to buy or recommend your product? If you exclude Mac users, for example, you’re missing out on a vocal population of important bloggers and pundits.

It’s happened to me. In the late 90s I designed a multimedia project called Xstream to use an 800×600 screen, using Director. Of course, the one guy with a 640×480 screen was the guy who had to approve the thing.

So, how big are their screens? Well, it’s not even that simple. On Windows, most users browse at full screen. On Mac, few do; the multi-window philosophy has been around too long. As screens get larger and full-screen browsing becomes less practical, browser sizes will become less predictable.

OK, enough waffle. Pretty, interactive distribution graphs for browser width and height from Foldspy. EDIT: Sadly, this site has expired.

The bottom line? If you want to please 95% of the global audience, your core content should fit within 787×423 pixels without scrolling. Not scary enough? Only about 50% of the audience can see more than 600 pixels vertically or around 1024 pixels horizontally at once.

Lastly, bear in mind the iPhone’s variable sizing and lack of Flash support. It’s going to be big and you don’t want to be left out.

Let me just spew out some discussion here on what the hell I’m going to do about the new iPhone. I have to get one; I’ve been waiting for a year now. It’s looking likely that an unlocked iPhone will be available locally. They’ll be on prepaid from Optus and Vodafone here (if not for the subsidised US$199 promised) and even if they’re tricky to unlock they’ll definitely be available somehow. Apparently phone locking is illegal in Belgium, for instance.

Anyway, I really don’t want a $60+/month contract to get one. Many people don’t view this as a big deal — some don’t even have a home line any more — but that’s not an option for me.

ADSL is the only way I’m getting fast internet here, so I need a phone line. The line rental on Homeline Budget is only $20/month, so it’s not too painful. We make all our outgoing calls through VOIP with MyNetFone (one of many providers, and they’ve been OK) for 10c untimed national calls and 20c/min to mobiles. Telstra rates can’t come close.

My current mobile plan is so cheap I can’t use it all. I pay $25/45 days for Virgin Mobile’s Bean Counter (only available online) which offers 10c/minute calls to any Australian number, including mobiles, plus free voicemail retrieval. Nothing else is anywhere near as cheap, not even VOIP if I’m calling a mobile. Virgin resells the Optus network.

Of course, one reason I can’t use all my credit is that making calls from home isn’t really an option as the reception isn’t good enough to make calls and I can barely receive them. Could be partly down to the handset, of course, but we’re on the edge of all the maps and not even part of the planned Optus/Voda 3G network expansion. Yay.

So, I can stick with my kinda crappy reception at home but very cheap calls and add some data, right? I don’t really need data except for Google Maps while somewhere new and for a little mobile browsing. I don’t need data while I’m at home on Wi-Fi, which is at least 3-5 days a week, and I can always call people back on VOIP if mobile reception is especially iffy. Very sadly, Virgin don’t offer data for prepaid people. If they did, everything would be easy: I’d sign up for their $10/month for 300MB deal.

Move to Virgin post-paid? All the “capped” plans on any carrier are horrible to compare because of the confusing mix of call rates and “included value”. So, I can get “$50 worth of value for $20” but my call rate increases to 80c/minute? Working the maths, that means my call rate roughly triples. Unless, of course, I go over that $50, in which case it goes up by 800%. Yipes. Every cap plan I’ve seen has nasty, expensive call rates around 80-95c/minute. The only way to justify this is if you’re sure there’s enough included credit to cover you, which is just what they’d prefer.

So what else is out there? A minefield. Most of these plans are relatively expensive because they’re subsidising a handset. I can recognise that it’s worth paying an extra $10/month over two years to save $240 or so, but the maths are not in your favour. There are lots of carriers out there (Boost, Crazy John, Virgin Mobile, 3, Telstra, Optus, Vodafone, Soul) but only a few networks. Telstra, Optus and Vodafone run the standard (2G) mobile networks; everyone else resells. Telstra and 3 do share a 3G network, though Telstra’s now pushing Next G — just 3G on the 850MHz frequency, exclusive to Telstra. Vodafone and Optus are building a competing 3G network on 900MHz — not supported on this iPhone and not going to Mt Nebo.

Oy. This is getting long, but I still have a baby asleep in a carrier on my front, so I’m still typing.

I’m going to discount Vodafone because their reception is even worse than Optus out here. The phone often won’t even ring. Optus could be OK, which makes Virgin possibly OK too. Telstra on Next G should be good (according to their coverage checker) but they’re expensive. More expensive than I’m paying now, anyway.

For now, you really can’t go past this mega-spreadsheet of all the carriers and their cap plans. Just change the info at the top right to match your average usage patterns and it tells you your best plan. Lucky me: Bean Counter is my best option right now.

Ever have problems taking photos in public? Read this, then download the linked wallet document that explains your rights at the end.