Thanks guys for your replies

I can't help but think that almost all of the large professional telescopes are truss designs and so it must merely be a matter of construction quality to achieve the required ridigity?
I think I would like to make a true
Serrurier Truss Newtonian which will be mounted on the G11 via the central box construction. I found an excellent example here:
http://www.aao.gov.au/local/www/sl/sl-tels.html#ni-tel There is a great photo of the truss design here:
http://www.aao.gov.au/local/www/sl/p.../30cm-new1.jpg
Alex/Clive, do you know if such an approach would be stable enough for imaging? Anything in particular to watch out for?
My local Bunnings sells 1" alu tubes for $10 pr. metre, both square and round. I don't have the equipment and skill to try welding alu, but I figure I can just bolt/screw it together and cut some thick plywood rings for the top end and make a mirror cell.
For the optics I have long been thinking about R.F. Royce mirrors, I checked yesterday and a 10" conical sells for US$850 which I think is not bad. One of the many advantages of these mirrors are that they simply mount with a bolt through the middle which further simplifies the design of the mirror cell.
I do dream of a 12" or larger though, but the 12" Royce is US$1700 which is a bit steep I think, compared to the 10". I would like to hear if anybody knows if it's worth it, or if I would be better off with an average run-of-the-mill 12" or larger? I do know that these optics certainly yield great results for planetary imaging, but I'm mainly thinking of long exposure deep sky imaging here. Perhaps the blurring due to seeing effects during long exposures may negate the advantages of the Royce optics to a degree?
By the way, I've attached some photos of my current mirror, as you can see it hasn't been resurfaced in 15 years

I must have lost some 30-40% of the reflectivity already when combined with the secondary...