Health continues to improve, and the weekend on Loch Fyne is currently “on”. Rachel and Ed will pick us up on their way there from London. We have arranged to communicate by text tomorrow, as they battle their way northwards through the holiday traffic – very 21^st century.

Yesterday I re-potted my chilli plants. Today we count pills and pack.

I think I am going to re-cast-off the entire upper edge of my sister's shawl. At first glance, I thought it was garter stitch, but I've had a look at the pattern and (a) it's moss stitch, much less fun; and (b) there is a lot of adroit short-rowing at the end, or rather long-rowing as stitches once left behind are gradually incorporated – only the last two rows involve the entire stitch-count. So I mustn't unravel mindlessly.

Nevertheless, it should prove relatively peaceful kitchen-table knitting. The Bridal Shawl is distinctly anti-social.

I had a look at Liz Lovick's book. Sure enough, she has a section called “designing with frames” but they aren't the sort of frames I was talking about yesterday. On careful re-reading, I think Sharon Miller's magnificent framed shawl could be reconstructed from “Heirloom Knitting”. I wonder why she never published it as a separate pattern? Feeling that no one would want it, since it was already in the book? I think she's wrong, if so.

What Lovick does say, is that island knitters almost invariably use k3togtbl for the double-decreases which are so plaguing me. I had a go yesterday – the current motifs have got a lot of them, and of course the surrounding trellis is composed of nothing else. I think perhaps k3togtbl does work a bit better in the trellis – it's easier and more secure than the centred decreases I've been doing, and easier to retrieve in the unhappy cases where such a decrease has to be unpicked.

In the motif, it doesn't work as well. It all depends on how the stitches to be decreased were formed in the preceding round. I've settled down with slip 1, k2tog, psso for those. Both alternatives produce angled results. I was astonished when I first learned from “Heirloom Knitting” that the angle of a decrease doesn't matter in fine lace knitting. Lovick agrees.

Miller's instructions for the framed shawl (such as they are), are worked centre-outwards. Lovick seems to like to do it that way, too. Not me, if only to avoid that long cast-on along one edge of the centre. I was feeling a bit frustrated about the resulting problem of turning the border pattern on its head, if I insist on knitting edging-inwards. Myrna Stahman says she includes lessons on how to do it, in her lace tutorials.

But then it occurred to me, I know how to do it. I've done it.

I've used Bridget Rorem's lace alphabet several times, and at least once, I had to reverse the letters, and did so successfully. In 2009 (I discover, by googling my own blog) I knit a First Holy Communion veil for James' and Cathy's daughters, incorporating the initials of both girls. It was knit from the top down so the letters had to be reversed. Here are finished pictures of it:

And here is an account of my struggles with the initials. I remember that day well, when I thought I was knitting the mirror-image of the effect I wanted.

Many border patterns wouldn't suffer much – including the one I'm knitting at the moment – if they were simply knit in the other direction. Not by turning the chart upside down, which might present problems, but by fitting the pattern as given (centre-out) into the greater abundance of initial stitches available when knitting edging-in. If you centre the pattern with care, it shouldn't be difficult.

I was rather struck, reading those old blog entries, with how much earlier I used to get up, five years ago.

The current (or about-to-be-current) Economist has a 14-page pullout by James, I am told, although he has unfortunately been pushed off the front cover by the Ukraine.