Jeff Zeitlin wrote:
 > Simple question, straight up: Is it plausible that canids (or
 > xenopseudocanids) of some sort could be domesticated and bred up to a
 > size that would make them usable as cavalry mounts?

Maybe. How big and fast does a cavalry mount need to be?
Who are the riders and what are they carrying?

Look up Epicyon ("more than a dog"), a Miocene era predator native to
North America that cracked 100kg body mass.

Biggest mammalian land carnivore looks like a variety of sabre-tooth
tiger, 'Smilodon populator' that topped out at 440kg.

There are biomechanical constraints on limb loading, most notably the
humeri, that may limit further increases in size - you can't run any
faster without risking catastrophic forelimb fractures.

Spine structure isn't that big a deal as long as the load is distributed
to be borne over the shoulders/thoracic spine.

Logistics are important as well, as other posters have noted. A
herbivore without a fussy diet enables foraging rather than carrying
food. Contrast horses with cows, sheep and goats for points along the
'fussy diet spectrum'.

Large carnivores are going to need a supply train of meat. Lots of it.
Otherwise you are going to deplete the countryside of other animal life
with every march.

Alternate cavalry animals? I'd use elephants, larger antelopes, or
theropod dinosaurs as a starting point.

Selective breeding will take a few dozen animal generations to choose
required traits; with Traveller tech levels this could be shortened by
perhaps five fold with advanced genetic engineering (do the introduced
traits remain stable in the population without deleterious effects?).

Rob O'Connor