Sorry for the delay between posts, but I only get to check /. infrequently these days.

Lappe and Baily are not technically wrong because they qualified their argument with 'could', not 'will'.

The canonical basis for their argument is something like the Flavr Savr tomato. What happened in that case was a biotech company with no traditional plant breeding experience employed a very early and expensive GM transformation process to a single variety of tomato (which, incidentally, happened to suck). Because of the development expenses and lack of understanding of the market, they only developed that one variety, hence a lack of genetic diversity of flavr savr tomatoes.

Taking a step back to explain, I'll start with the fact that all transgenes on the market in field crops today do not actually impact the quality of the product to the consumer at all. They don't make them taste better, look better, more nutritious, etc. This is actually largely due to regulatory reasons, as the FDA has much more stringent testing requirements if a product is actually supposed to impact the consumer as opposed to lighter restrictions if you can prove that there is no interaction intended or in practice (e.g., BT transgenic crops produce a protein that does not interact with animal digestive systems at all (and that organic farmers actually spray on their crops, incidentally)). All current field crop transgenes make the production process cheaper, more energy efficient, etc.

Anyway, in effect, transgenes basically add on a feature to an existing variety, leaving all other traits of that variety unmodified. If you added the round-up ready transgene to the red delicious apple variety, they would taste exactly like normal red delicious apples. As far as the consumer was concerned, there would be no discernable difference (unless you had a genetic or protein analysis lab at your disposal).

If you're a developer of apple varieties, and you wanted to offer your full lineup with transgenics, you would have to add the genes to every single variety of apple separately.

Development of varieties of a crop, and development and integration of transgenes into a crop, are basically two separate processes.

Due to current technological limitations, market pressures, regulatory requirements, and other business factors, transgenes are used for single-gene qualitative traits (e.g., herbicide/insect resistance) rather than quantitative ones (e.g., yield).

As if this post weren't long enough, let me use a car analogy:

Traditional breeding is like developing new cars. You use better manufacturing techniques, materials, and refined aesthetic focus to improve model lines or even come out with new line altogether.

Transgenics are like developing accessories that will fit a wide variety of vehicles, such as stereos, A/C systems, power steering, etc.

The Flavr Savr guys made the mistake of thinking that their transgene would singlehandedly ensure success despite the fact that they used a crap variety as the base of their product. It was comparable to OnStar thinking their system is so awesome that they could throw it into a Pinto and it'd sell like hotcakes even though it would be competing with similarly priced BMWs.

Lappe and Baily's argument is essentially based on the premise that all companies offering transgenes will ignore Calgene's mistake and repeat it blindly. Companies have done many stupid things, so this is not entirely unthinkable, but it is hardly likely and has not happened in practice.

Compared to transgenics, the advent of hybrids and commercial production farming have caused a decline in genetic diversity to a much larger extent. They have also caused orders of magnitude greater production, without which we would have been suffering global famine since the 60s, so it is something to weigh against the potential risks: narrowing the genetic diversity of production field crops is actually preventing global famine right now, but could also potentially contribute to widespread crop failure in the future.

Genetically modified corn isn't any different than non-GM corn except for the presence of the transgene, and contains the same minor genetic variations you'd find in any field of hybrid corn. The seed is propagated the same way as other corn, as cloning is simply too expensive for production use.

Other methods, like doubling haploids, can be used to create genetically identical populations, but those methods are not specific to transgenic material at all.

I personally think that the PVPA and related enforcement laws are messed up, and that Monsanto are a bunch of assholes, but you are conflating issues here.

A small amount of contamination in plants will not result in RR plants dominating the population as they did on Schmeiser's farm. Had he simply indiscriminately saved seed, the gene would have only been present at trace levels, or possibly been eliminated from the population without continued selection pressure favoring the gene.

In short, the only way his farm could come to be dominated by RR plants in that time frame was for him to deliberately select for RR plants.

Had he been honest he might have said 'yeah, I took those plants but I disagree with the laws that allow the patenting of plant varieties, so I don't recognize their legitimacy, especially because I was not responsible for the initial plants contaminating my farm'.

Instead he lied.

Is the law unjust? In my opinion, yes. Was Schmeiser guilty of deliberately propagating unlicensed seed? Yes. The two are not mutually exclusive.

None of those have anything to do with farmers' crops being contaminated with the terminator gene (hint: because it never happened).

It's the law (look up the PVPA of 1970) that prevents'you from re-planting, not the fact that it's GM.

Speaking as someone working in the industry, I would say that your argument is based on an impossible premise. There can be no 'one perfect set of corn' world-wide because there is too much environmental diversity. The 'perfect' variety is specific to a given environment, so you will see as many varieties as there are environmental conditions.

In fact, as the market matures, genetic diversity of GM crops continues to increase.

Essentially, in the early stages of commercialization of GM crops, the transformation and development process was far too expensive relative to the projected market size to apply to all but a few of the prize lines of corn, soya, etc. Now that competition has increased and the development processes have been refined and costs reduced, instead of 2 or 3 lines with a given GM trait, you get thousands, and that number is just getting larger.

Not. There are numerous non-GM techniques to induce sterility.

1) Depends on orbital altitude over the theatre of operations. Low (enough) latency links can be accomplished with sats if you're willing to spend money on a satellite constellation of sufficient density to allow for full coverage of the area with low altitude sats. Considerably lower latency than a Hawaiian quake player would experience on a west coast server is possible.

2) Yes, but that is neither an unsolvable nor unsolved problem.

I would also look at Scala. Shares Groovy's strength of transparent Java interoperability, but opts for a radical departure in language design and concepts (for the better, IMHO).