Some pieces of software are intended to be feature-complete replacements for competitors. Others approach the same or similar tasks using very different methods. This doesn't change the relative merits of the software and soda analogies to books. Software is a much better analogy than soft drinks.

The rareness of feature-complete replacements for books, except in certain non-fiction areas, detracts from the software analogy to be sure. It still makes a lot more sense than 'coke vs. pepsi.'

The bottom line is that Google Search doesn't work very well - at least, not anymore. While it previously supported search expansions which could be taken advantage of by skilled searchers, it's since been focused on quick, lowest-common-denominator responses to the most common questions. As a result, searching for slightly abstract notions is virtually impossible, and some searches which should be straightforward also fail.

One example of a simple failure: "fireworks today" or "fireworks today san francisco" returned nothing after I chanced to see fireworks the other night. Using the date ("fireworks san francisco may 21 2014"), the only relevant result was a set of Coast Guard and DHS documents describing safety precautions for the event (Giants game). Of course, fireworks games are well publicized outside of interntal government safety documents.

A more abstract example: try to design a search for articles about names which are or have become insults, such as "Dick."

A mother who had been infected for a long period of time might have multiple strains due to accumulated mutation. About 25% of the time the fetus becomes infected, and in about half of these cases only one of the mother's strains will appear in the infant, which is usually assumed to mean that only a single virus of that strain was able to penetrate the in utero defenses (which are quite good, considering 0 viruses get through ~75% of the time).

To be clear, it's not that all the infections across individuals begin with the same strain, it's that each individual infection (frequently) does.

As dreamchaser said, HIV isn't actually all that easy to transmit, so in many cases it is only a single virus which begins the infection. In the other cases (heterogeneous initial infection) there was probably a unusually high viral dose.

Thought I would also mention how bizarre the schistosoma (genus) parasites actually are (more details can be found e.g. on wikipedia).

The eggs are deposited into water by infected humans, and infect certain snails. The snails later release a larval stage which has a tail, but no functional digestive system. These penetrate the skin of a human host, losing their tails in the process. The next larval stage is a few micrometers in size and has a sucker; it can now eat. They typically migrate to the lungs, where they will mature for about 1 week, living off the blood of the host.

As they mature into adult parasites, which are up to ~10 mM in size and visible to the naked eye, they migrate again, often to the liver or intestine. If possible, they will also find mates. The female parasite lives in a canal formed of by the male flatworm's body rolling up. The breeding pairs are monogamous and individual parasites may live for decades. During this time, the pair will reproduce continually, producing up to thousands of eggs per week (depending on species).

The eggs are deposited into the infected person's feces, where they find their way back into the water supply, but about half of the eggs become trapped in the body. These eggs mature normally, releasing many antigens which contribute to an active immune response which leads to most of the symptoms of the diseases.

The parasites are highly evolved to fit their human hosts, to the extent that they depend on human enzymes in order to complete certain essential metabolic pathways. They also have genes which are highly similar to human genes, which may help the parasite evade the immune system (such genes could have evolved or been acquired through horizontal gene transfer).

If it sounds unlikely, consider that a single bacterium, which only doubles every generation, can rapidly give rise to large colonies. For example, with a reasonable doubling time of 1 hour, 1 bacterium will become ~268 million in 48 hours. In contrast, viruses can create hundreds or thousands of copies with each generation. When actively replicating, they can spread very quickly.

The variety in diversity of viral populations in recent infections is probably caused by a variation in initial viral dose (b/c more particles make genetic diversity more likely), and also by effects like the two you mentioned. I'm a graduate student in biochemistry, and I've learned never to assume that those types of phenomena are mutually exclusive.

By the way, this popular summary describes a model experiment on the feasibility of very low dose infections and how genetic diversity of the viral population varies probabilistically with the initial dosage.

These authors aren't the only ones to have identified a link between schistosomiasis and HIV, see this 2011 paper in PLoS NTD for example.

It seems pretty obvious that any source of genital sores is going to increase the risk of HIV transmission, especially considering that this is the entire basis of increased infection rates for individuals engaging in anal sex.

Still, the link to HIV should only slightly increase our desire to deal with schistosomiasis, because schisto is already the second-most socio-economically devastating illness in the world (after malaria). It causes liver and kidney damage, diarrhea, and vascular disease. It is associated with bladder and colorectal cancers, increases metastasis of other cancer (including breast cancer), impairs mental and physical development in children, and leads to reduced cognitive function and work ability in adults. Further, only one drug is widely used to treat the disease (praziquantel, aka PZQ), and resistance has begun to be observed. 200 million people are believed to require treatment, and more than 600 million are at risk of infection.

Disclaimer: I develop automated drug screening methods against NTD, including schistosomiasis

When DNA is replicated in most life-forms, there are extensive proofreading mechanisms which identify replication errors, cut them out of the DNA, and then re-copy those sections.

In contrast, HIV does not extensively proofread new DNA during viral replication. That means that copy errors are quite likely, and over time lead to a heterogeneous population of mutant viruses in an infected person.

Like other viruses, HIV has evolved a balance between correct and incorrect DNA replication which keeps the virus deadly while still evading the immune system.