The lack of compatibility is easy. Software designed to use the network, won't work on the model A. Ok, fair enough, but if software unintentionally uses the network, or has a flawed check for availablity of the network it can fail.
If network==online then check_for_update() and start(), else if network==offline then start(). This simple check will fail if you do not anticipate that the network might be undefined.
I know loads of software that can't be used without a mouse, even though I see no reason for it to require a mouse. I know software which won't work without a floppy drive just because it assumes there is always an A: drive and by default shows the file listing of that drive and thus crashes. I know of software which won't run on a PC without a graphics adapter (why require a graphics adapter in a server?) or keyboard attached.
Still, it all probably seemed like safe assumptions at the time, or it was just overlooked.
Software which requires both the mouse and the keyboard will start just fine, but require an extra investment (Like a usb-hub or more expensive mouse+keyboard set) to be porperly usable on a model A.
And it's not software alone. There are a lot of cases being designed, none of which cover the network slot (which you could cut away for the model B or something)
Nobody is considering the model A. And why would they/how can they?
There are specific orders for a pile of Model B's. It's is not yet known what the demand for A's is.
It's also not know how much of the model B's orders would have been a model A order if that were an option. And you won't know how much demand there is for the model A as long the model A is postponed because the model B is stil selling. Part of the success of the model B is the fact that there is no Model A.
It's like a bakery saying "We only sell raspberry pies because everybody apears to want a raspberry pie", ignoring the people who turned around because they wanted a strawberry pie and implying that people who ordered a raspberry pie, ordered it because they wanted raspberry pi (while some might have prefered strawberry).
And who would want a strawberry pi in the future when all plates have funny raspberries printed on them, all sauces are designed to taste great with raspberry, etc. You must love strawberry very much to still go for the unpopular choice.
If anyone want's a A, the only reason NOT to get a B and not use the extra's, is the extra $10.
Some extra reasons:
- Ordering a model B increases the waiting list for a model A. If you want a model A somewhere in the future, the fastest would be not to order a model B for the time being. You won't be any quicker inside the movie theater when you leave the line to get a coke and come back later when the line is shorter.
- There are 20+ cases for the model B and none for the model A. it will (probably) fit, but with a gaping hole where the network connector would be. you'll always have the feeling you're missing out.
- There's software which requires both mouse and keyboard or a network-connector. All software is garaunteed to work on a model B, and will most likely work on the model A. The safe choice is model B.
- Why buy a model A when everybody else has a model B?
Ok, there's only a little bit of risk that some software won't work (without an email to the programmer who made it, or some extra hardware), less choice in cases, a little less support from the community. Those are all minor issues and only increasing a little bit over time, but when do they add up to $10? which also isn't much.
Would you buy a device which is $10 cheaper and said to be 'compatible'? or spend that extra $10 and go for the real thing? When do we have to say to schools "you could save a few buck buying the model A if you really want, you're only making it just a tiny more difficult on yourself".
The longer the model A is ignored, the more this grows. (not taking pre-orders for the model A, while taking 300.000 preorders for the model B, IS ignoring the model A) And the difference between the model A being usefull and being useless is only $10.
What is the use of the model A when everybody has the model B? It's not an upgrade. Is a $10 savings really worth it? Or is it only so we can keep the slogan "$25, have a bite" at the top of our page?