Nikon 35mm f/2 AF-D

Concept: 2 out of 5
Execution: 2 out of 5
Yeah, but: No, I didn't inadvertently use one photo twice.

The Long Version: I've never really liked the Nikkor 35/2 AF-D lens, but I bought it anyway. Simply for the sake of versatility I wanted something wider than my 50/1.8D for my F-mount cameras. As always with SLR lenses, price ramps up very quickly as the focal length and apertures decrease, and the 35/2D was as far as I could bring myself to go. Nikon wide lenses simply aren't as good as my Zeiss or computer-assisted Panasonics, so I knew I would need to lower my standards a bit.

This is the point where I should be saying that I've changed my mind and now have glowing opinions of the lens, but that just isn't how it's worked out. It's the best 35mm "FX format" lens – a term nearly a decade away when it was designed – that costs less than a thousand dollars, but as it stands there's a very big gap between the f/2D and the new f/1.4G 35mm lenses. It would be nice to have an option in between the functional and the phenomenally expensive.

All Nikon AF-D lenses need a camera with a screw drive to autofocus, making them duds on the cheapest bodies but they can be screaming fast on the higher-end cameras. The 35/2's focusing is very quick and has an excellent minimum distance of 25cm, or about a handspan from the camera body to the subject. There is some focusing noise, but if anyone thinks that it's objectionable then they'll probably be knocked over by the ensuing thunderclap from the shutter. Optically the results are decent, with strong centre sharpness that peters out toward the corners, but the big sin for me is its strong barrel distortion.

The salvation of the 35/2D is that digital redemption is possible with software lens profiles. Lightroom/ACR has this built in, and other programs are also capable of fixing some optical flaws. There's a lot to be said for lenses that don't need to be fixed in photoshop, but as the Panasonic in-camera processing demonstrates, it's an increasingly viable way to solve problems.

The 35/2D is a midrange film-era lens, and it shows. With film vignetting isn't as strong; sharpness and distortion is less significant on film than on those unforgivingly flat and flawless digital chips. As a lens of only minor personal importance, the performance of the 35/2D is adequate but leaves me uninspired. As a result this lens normally lives on my F100, where it's comfortable and not often disturbed by needing to be moved to my D700.

I've never really liked the Nikon 35/2 AF-D lens, but I'm glad that I bought it anyway. I almost exclusively use the 85/1.8, 85/2.8, or 105/2.8VR on my Nikons, but without the 35/2D, there are times when I would have had to leave the big SLRs at home. So even though the 35/2 is in third place – at best – on the list of wide-angle lenses that I reach for, it's the enabler lens that completes the Nikon set. And when the cameras are that good, sometimes it's okay for the optics to just be good enough.

last updated 27 nov 2011

Rain Alarm (Extended) for iPhone

Concept: 4 out of 5
Execution: 3 out of 5
Yeah, but: Chicken Little eat your heart out.

The Long Version: I enjoy planning. I like to know what to expect, and I like to have some idea what to do if that's not what happens. One manifestation of this is that I really want to know if it's likely to rain.

Rain Alarm is a useful little application for iOS. It taps into publicly-available weather radar and watches for precipitation around the phone's location; both the radius and the excitability of the program can be selected from its settings menu. It then ties into the Notification Center in iOS5 for messages that read "Precipitation about 5.2 km away (strength 30 of 100, area 4 of 100)". The application also shows the weather radar, and can animate the maps, making it easy to see exactly what's coming and how far away it is.

So it's goodbye to that charming forecast "50% chance of scattered showers" that sounds significant but is short on specifics. By using Rain Alarm for the past month I've been able to make better choices about when to bike to and from work, and when I'll need an umbrella versus getting by with a hat. By paying attention to the program as well as the actual weather outside my window, I've learned how to interpret the programs's strength ratings and choose the gear that I need to cope with it. Yes, I'm still caught unprepared sometimes, but that's really just part of the fun of being obsessive.

The reason why I bought an iPhone – aside from all of the accessories to dress it up in – is because I wanted a portable computer. Clever applications like Rain Alarm, that can use the phone's on-board sensors to pull useful information out of a data stream, are exactly what these things are good for. It's a fairly simple task, but it's done well.

last updated 24 nov 2011

RoboSport

Concept: 5 out of 5
Execution: 3 out of 5
Yeah, but: Mario, schmario.

The Long Version: I don't usually review things that I haven't used for almost twenty years, but one of the lingering injustices of electronic gaming is that the 1991 Mac/PC/Amiga classic RoboSport has never been reissued for a modern platform.

I started to play Robosport on a friend's Mac SE, long before I had a computer of my own. It and Spaceward Ho! were our favourite two-payer games, because they're both turn-based and would calculate the outcomes of everyone's actions without needing to be played on two computers or a fast network. The premise of RoboSport was a near-future televised sporting event that had teams of combat robots competing in special arenas. Not too shabby an idea, especially considering that this predated the radio-controlled cars of BattleBots or reality TV.

A team could be as many as eight robots, and each one could be armed with a rifle, machine gun, heavy machine gun, or missile launcher. The rifle robot had the best armour and long-range accuracy, the heavy machine gun had the most close-in firepower but the least armour, and the missileer had much better accuracy than the grenades that anyone could throw. While the game arenas had a top-down view, it wasn't necessarily omniscient – robots had distinct fields of view, cones of fire, and can't see past walls or obstructions. There were different maps, some laid out as ruins, others as suburbs, and some as computer circuitry. I mostly played capture the flag, hostage rescue, and last robot standing, but other modes included treasure hunt, where the robots need to hunt coins as well as each other, and "baseball" where they have to run bases by reaching certain waypoints.

So far it's just a basic-but-quirky squad-infantry game, but the fun comes from needing to pre-program each robot's moves. Stand here, look there; wait fifteen seconds and then throw a grenade before rushing into the next room; call shots on specific targets or just wait for targets to appear: it became a complicated dance with everyone reacting to what happened in the last round. Sometimes it worked out the way I'd expect, but usually the results were amusing and unintended. Opponents rushing past each other in doorways to take up defensive positions in the room that the other just left, a complicated outlay of firepower aimed at nothing while a lone rifleman plinks away from an unexpected direction, obliterating your own team with some badly thrown grenades – there's nothing to do but watch the results and choose your actions to program for the next round.

Computer games don't need great graphics, just good ideas. The computer that once ran Robosport had less power than a pre-WebOS Palmpilot, and my current phone has a higher-resolution screen. If some modern advances could be brought to the same game concept – better AI, network multiplayer, colour, not crashing all the time – I would buy it in a heartbeat. What's more, if it ran on a platform that I don't own, I would buy one just to play Robosport with. The irony of having it on an Android tablet would capture some of the inherent sense of humour of the game, so that would be just about perfect. Fingers crossed.


postscript: as often happens, in writing this review I've found something else. Looking for background on Robosport led me to the contemporary game Frozen Synapse, and playing it has been enough to keep me from Twitter for almost two days. There may be a review of it in the future, but for now I can say that no matter how good it is, it's still not Robosport.

last updated 19 nov 2011