Python is slow because it does a lot of dynamic dispatch, which dwarfs the cost of actual operations such as addition. So it's the other way around - Python, of all things, could probably switch to decimal by default without a significant slowdown.
What would be much slower is all the native code that Python apps use for bulk math, such as numpy. And that is because we don't have decimal floating point implemented in hardware, not because of some fundamental limitation. If decimals were more popular in programming, I'm sure the CPUs would quickly start providing optimized instructions for them, much like BCD was handled when it was popular.
Eh, maybe. I don’t think many NumPy people would reach for a decimal dtype even if it were available and implemented in hardware. It just isn’t useful in the primary domains where NumPy is used.
That leaves vanilla Python. You’re right that compared to dynamic dispatch, a decimal op implemented in hardware is nothing, but the issue here (IMO) is death by a thousand cuts. Python is already slow as hell, and currently using decimal as a default would require a software implementation in many if not all places, which will be slow. Trade off does not seem worth it to me.
Sorry, my comment was about it being slow, not about it being a bad idea: It is a bad idea, but it will not really make your python even slower than it is today. The Intel supports most of ieee754-2008 which has most of what you would need.
Judging from other comments here, it is not clear how widely supported IEEE754-2008 is. If it isn’t supported everywhere, it would make a VERY bad default for a numeric type.
It also appears that the logic for implemented something like this standard is indeed slower than standard IEEE754. Even if it’s only a bit slower, seems bad to make it the default.
All this just to fix something which is confusing to a novice programmer… and this is leaving aside the additional complications a fixed width decimal has which a floating point type doesn’t have.
It's fully supported. However, some parts only via a library. Anyway, to drill down my argument. It doesn't matter how slow or fast it is: python's floats ints,... are boxed. Your performance is down the drain before you even start looking at the value at hand. If you want any kind of performance at all while crunching numbers you will not be doing it in Python (which is exactly what happens when you use things like numpy)
