W. H. Heydt,
The original Pi was laughably obsolete when you could first buy it.
I have heard this opinion expressed many times and I'm really wondering what it's supposed to mean.
For a year or two prior to the launch of the Pi I was keeping an eye out for small, cheap and capable ARM boards to replace big, expensive industrial PC boxes. After all, phones are cheap, ARMs are cheap, why not cheap embedded boards using an ARM?
A little bit before the Pi launch one of the cheapest and best options you could get was a IGEP board from ISEE.
https://www.isee.biz/products/igep-proc ... pv2-dm3730
. Very nice but it set you back nearly 200 euro. Still a lot cheaper than those industrial PC's we were using.
Then comes the Pi. What? A board that can do all we want for 35 dollars? Amazing!
So, yes, maybe the processor was a bit old. So what? It was the cheapest kid on the block that could all the work we needed for nearly a tenth of the cost.
The Pi was revolutionary in that respect. Certainly not "laughably obsolete". So revolutionary in fact that it has inspired many followers since.
"Laughably obsolete" is not the same thing as "useless." There are tasks for which a 300MHz Pentium would work just fine, and that's the speed that is usually quoted as the 2835 functional equivalent.
The *price* of the Pi is a big part of what makes it revolutionary. The Beaglebone that was around before the Pi launched was $90. The "upgraded" BBB was $45 (and the current version is $55...and we'll get to see how long *that* price lasts). The Cubieboard-1 hit the market at $50 (Cortex-A8 at 1GHz). The first board I've seen that matches the Pi on price and beats it on raw specs is the Odroid-C1 (quad-core A5) and that comes after the abortive attempt to copy the Pi with the Odroid-W.
The RPF has reduced the price of SBCs for the educational and hobbyist market. The RPF has also shown that sold, well-wrking OSes are a *required* part of selling an SBC into these markets. The Pi2B has now made performance at this price point an issue. It is going to be extremely interesting to see how the other board makers respond, and to see if they *can* respond.
Are there features I'd like to see on a future board? Certainly. If there is a "must have" feature that the Pi lacks, I'll buy a board that has it, even though I'd prefer to have a Pi with that feature. If the RPF then comes out with a board that has the feature I'm looking for, I will--in all likelihood--switch to the Pi that has it. At this price point, the boards are consumables.
Some people have talked of wanting a Pi2A. I have mixed feelings about that. I think one will happen, but it'll take a while for two reasons--neither of which are exactly the one that has been given--the first being that the engineering work has to be done to develop it. The second is that, with the surge of sales from the Pi2B (150K in two days, and that limited by the initial stock) it's going to take a while for the factory to meet demand...for the Pi2B, for the B+, for the A+, etc. If the RPF announced a Pi2A tomorrow, you wouldn't be able to buy one for weeks or months simply because the factory capacity isn't there to make them. It's be the original Model A ordering problem all over again...and writ larger.
Then there is this... The major use of the A+ is probably robotics and other battery powered projects. A Pi2A is going to require more current, making it *less* desirable in these applications. There will be some where the increased compute power and memory is essential, and those applications will want to use a P2A. Applications that are more power sensitive and can live with a less capable CPU will stick with the A+. Indeed, I think this makes a better argument for continuing to produce the A+ than a corresponding one to continue making the B+, though the B+ will certainly be useful as a development tool for applications where the final device uses an A+.
And finally... After the RPF engineers catch their breath and catch up on sleep, I'd bet that the RPF will start looking around to see what could potentially be used to make a Pi3B. I don't expect to see one for 2 to 5 years, but the search starts as soon as the relevant folks come up for air. Since this isn't a "wish list thread", I won't say what new or revised features I'd like to see, but I'd be inclined to bet on it having "more of the same"...faster cores and more memory.