Tomatoes that didn't quite work

I started them too early. Everyone always starts them too early. I know this — I've read it, I've said it to other people — and I still sowed in the first week of February because February is long and grey and seed catalogues arrive in January and it's hard not to. By April I had leggy, pale-stemmed plants that had exhausted their small pots and were living on a windowsill in light too thin for them. I transplanted them twice before they went outside and they never quite recovered their confidence.

The variety was Gardener's Delight, which is reliable and forgiving and the right choice. The problem was me, not the tomatoes. I missed side shoots in June when I was busy with other things, and two of the three plants turned into something more like small trees — branching, heavy, held up by an increasingly elaborate network of string and cane. They produced fruit. Quite a lot of fruit, actually. But the trusses were smaller than they should have been and the fruit took longer to ripen because the plant's energy was spread across too many stems.

Blight arrived in mid-August, which is about when it always arrives in a wet summer. The lower leaves went first — the characteristic brown patches spreading fast in the warm damp nights. I removed them as I found them and kept the plants going until September, but the harvest was reduced and the ones that did ripen had to be picked before they were fully ready and left on a windowsill to finish. Windowsill-ripened tomatoes are better than no tomatoes. They are not as good as vine-ripened tomatoes.

What I'll do differently: sow in the third week of March. One pot per plant from the start, moved up once. Fewer plants, more attentively managed. A weekly side-shoot check in the calendar. And some kind of cover — even a simple frame with fleece — to slow the blight down. It's not about getting it right; it's about learning what wrong looks like in enough detail that next year's wrong is slightly less wrong.