I cannot now find the first book I ever made, but I recall that in the publication data I said something like ‘printed and published, with labour and great pain, using Microsoft Publisher’.
Part of the pain was certainly Publisher, which I recall as a quirky beast that took some time to master, but far and away the greater part was the actual printing, which was done on an inkjet printer (an Epson, I think) which only printed one side at a time.
This meant that you had to set it to print one side of all the pages in a booklet (I didn’t know they were called ‘signatures’, then) then feed these back in to print the other side, ensuring that they were in the right order, the right way up, and the right way round, so that considerable concentration was required to avoid the full range of possible errors; while from time to time the printer would introduce errors of its own, drawing through two sheets at once, so that the whole booklet was ruined (believe me, trying to salvage a sheet that is blank on one side by printing the pages that should be on the other as a one-off was not worth the convolutions it involved).
The only benefit of this procedure, if it could be called a benefit, was that, like banging your head on a brick wall, it was nice when you stopped. It made completing a single book feel like a real triumph, but it was labour intensive in the extreme – you were always doing something printer-related (setting up the page sequence on the computer, sending the print instruction, keeping an eye on the sheets emerging, arranging them to feed back in, setting up the next page sequence and so on). It was clear that when it came to volume production, it was a non-starter.
Duplex printing was the only way to go, but there were still decisions to be made after that. The proof copies were printed individually and by the time I had done five of them it was clear that it was only a little less laborious than the old way – sure, the laser printer was much faster and duplex printing made everything much simpler, but it actually made the whole business more hectic and, crucially, still tied you to the computer and the printer, taking out the printed signature and setting up the instruction for the next one.
I did attempt to issue a single instruction to print an entire book as a sequence of booklets – which ought to have worked, by separating each instruction by commas (e.g. 1–32, 33–64, 65–96 etc) – but in the event it treated the whole thing as one giant booklet, with the outside sheet having the last page and the first on the reverse, with only the very centre of the whole book treated as facing pages in sequence.
The solution I hit on was to print ten of each signature (collated) at a time, which allowed the printer to chunter away and left me free for other things. When each set of ten was finished, I could set up the next, then sort the printed one into ten separate piles on a big table, and so on, till after nine separate runs (I was operating on 8 sheet/32 page signatures by this time) I would have ten complete books ready for the next stage, which was folding.
This proved a most practical solution, since it allowed steady progress at every stage – while one set of ten books was printing, I could be collating the next for folding, then putting them under weights to press them (and as the piles accumulated, I could use each one to press down on the next). Once I started folding, I could put the stacks of signatures back in the boxes that the paper came in, in two A5 columns, then when the box was full, stick a weight on top to compress it; and all the while the printer churned away, and everything mounted up in easily-counted units of ten.
It’s probably worth mentioning at this point that you are wise to anticipate all this sorting when you are at the printing stage. You can of course check page numbers to ensure that your sequences are in order (and you will find that it is good to make a written note of them that you can refer to, both for issuing print instructions and collating signatures – e.g. sig. 1: 1-32, sig 2: 33-64, etc. But it is also useful to identify each signature individually, and indeed that is why they have that name – the printer would put a mark, usually a letter, to ‘sign’ each one, so that the sequence could be easily established.
For the proofs, I printed a set of bars down the spine of each signature, in a different position for each, so that when a signature was assembled correctly, they would appear as a set of steps, ensuring that I had a visual check that everything was in order before I started sewing (since when you sew, it is the spine that is towards you). Though this method appealed to me, it was a bit of a fiddle to set up on InDesign (effectively you have to treat the spine of each signature – i.e. the centre line of the reverse side of the outermost page – as split into two margins, one on the first page of the signature, one on the last).
When I started full production, I opted instead to incorporate the signature number at the top end of the spine, which was much easier, and worked just as well – since the numbers were always at the same end, you could spot any signature that was upside down, and of course the sequence ensure that they were in order. As well as numbering the signatures, I also printed six regularly-spaced short bold horizontal lines across the spine of each signature as a guide to sawing, which you do at a later stage, to make the holes you are going to sew through: with the book assembled and held in a vice, the marks on each signature should line up to provide six guidelines – simply saw across these and you should have a suitable set of holes in each signature, all aligned, which saves a great deal of labour and fiddling about with individual measurement, marking and pricking holes with an awl, and so on.