Drosophila naming conventions are not as simplified as nematode naming systems, nor as arcane as the numerous names given to the same mammalian protein or protein-like activity. They are simply based on the mutant phenotype of the early alleles recovered.

White mutants have white eyes. Wingless reflects the phenotype of the first homozygous viable allele, rather than the stronger phenotypes of null alleles. OTOH, it turns out to be substantially important across phyla. Notch reflects the dominant phenotype of the first allele, a notch in the wings. Aceate and scute (both reflecting a lack of bristles) are key bHLH proteins involved in cell type determination downstream of notch. Hedgehog may be initially named after the larval mutant phenotype (nulls have a lawn of 'deniticles' (hair-like cuticular protuberances normally found in limited areas of the ventral portion of each segment)) leading to the name, but the mammalian versions were given names associated with a video game.

Shavenbaby describes an embryonic lethal mutation that removes the hair like structures (tricomes) on embryonic larvae. In this context, the name makes sense, particularly since this describes a loss-of-function phenotype of sufficient severity to be lethal. All of this developed in the context lacking molecular data. Nematodeologists use a similar system, but with a simplified set of nameable phenotypes: Uncoordinated (unc), sickly short and maybe fat (dumpy, dpy), lethal (let), etc., each followed by a number indicating isolation order. (Is it easier to remember acheate, scute and notch or Let 13 26 and 30?)

Human gene names are either based on a single biochemical assay or a disease, neither of which are more informative than either worm or fly names. Since the fly community and worm community spotted so many of the key genes first, the MDs are now worried about unPC names and are trying to develop a seemingly meaningless nomenclature that won't irritate patients (trying to avoid the Far-Side-cartoon scenario of saying to a patient, "Mr. Farnsworth, I'm sorry, but you have cows."

BTW, if building a functioning body is compared to building a house, the mutants people who work on worms and flies generally concentrate on are the things involved in making solid foundations, building walls, floors, roofs and critical aspects of plumbing, heating and insulation. What human genetics finds are all the add ons that happen after those key events have happened. The things that people of who study flies and worms fined would generally be lethal in humans. Human diseases, or variants, occur in 'animals' in which the more fundamental things went very well.

Does evolution proceed by this, that, or the other thing?

Yes.

When you have a complex process such as evolution you cannot be limited to one procedure. One change to a gene may result in a small change to the organism, or to a large change to the organism. The deletion of a protein marker in cell membranes that allows the entry of HIV RNA, or a modification in the development of the hindlimbs or example.

Look into the matter of the newly discovered cnidarian worm, and ask youself, "what happened in later animals to eliminate the slime mold phase which this worm apparently experiences under the proper conditions?

BTW, if building a functioning body is compared to building a house, the mutants people who work on worms and flies generally concentrate on are the things involved in making solid foundations, building walls, floors, roofs and critical aspects of plumbing, heating and insulation. What human genetics finds are all the add ons that happen after those key events have happened.

I agree, which is why I like this study and the look at the developmental network. as the authors say, there are a number of mutations that could have given the same phenotype (in the lab), but evolution was constrained to using only those that fell at a certain "evolvable" point in the network. pretty striking.