When posting *anything* on *either side* of a crevo thread.
There, I've set my web, let others walk into it :-)
Cheers!
PH, you think locomotion is evolutionarily more important than foraging? I would think that the use of silk to create webs that can catch prey would be a more important evolutionary event; the use of silk as a means of locomotion being a next step to better situate the means of ambush, not the other way around. This can actually be supported by the fact that some species of spider create ground traps with silk rather than create webs that are arial.
What's more "important" is not necessarily the function that evolves first. What's "easiest" is often the first thing to evolve (i.e. involves the least amount of modification of existing structures or biochemistry).
Evolution usually proceeds by small stepwise modifications, and the first few small steps towards the development of modern silk and silk usage may not have involved the catching or holding of prey, because that would involve stronger silk and more complicated usage of it than one would expect during the earliest stages of silk production.
However, it wouldn't take much modification of existing insect physiology for an early insect to have had a viscous substance (a proto-silk, as it were) seep out of its feet in order to improve traction as it walked (rather in the way that snails and slugs currently leave slime trails). If that was the first step (no pun intended), from there evolution would have driven the stepwise improvement of the sticky substance itself, its applications, the behaviors which drove how it was used, etc., until much later it was suitable for creating "suspension bridges", traps, wrappers, etc.
Oh, and speaking of caterpillars, while moth caterpillars often spin protective cocoons to shield themselves while they pupate, butterfly caterpillars don't, they just lay a very sparse (almost invisible) layer of silk as they walk across leaves and twigs, in order to improve traction and lower the odds of falling off, and also make tiny "pads" of silk to anchor their "tails" when moulting their skins (including one to anchor themselves as they moult their last larval skin and pupate).
So in butterflies, at least, silk production is still used solely for "attachment" purposes, if that helps shed some light on this issue.
Even if catching prey is more important than locomotion, that doesn't answer the question of which came first. Evolution doesn't give creatures what they "need most". It just works with whatever happened to be there. Let's suppose there was some sort of scent gland in the feet of the ancestral spider; it might have been a small modification to turn that into a sticky foot adhesive, but it would take many modifications to turn that into the abdominal wonder gadget spiders have today.
It's possible that the fossil record might someday tell us what came first. A better route might be to look at the genes, and see where the foot-spinneret fits into the cladogram of arachnid abdomen-spinnerets. If the foot-spinneret came first, it will be genetically equidistant from the abdomen-spinnerets of all spiders, because all abdomen-spinnerets will be modifications of some close version of its gene(s). If the foot-spinneret came second, then its gene(s) will be closely related to the abdomen-spinneret gene(s) of the tarantulas.