Here’s an excellent example:
Decimal floats are symptomatic of poor design.
You’ve got to be pretty fucked up to consider a hack, however elegant, designed to work around the storage and performance limitations of binary computers, the “right design” while a system which, with certain penalties, produces actual, accurate mathematics, a “broken design.”
(Of course, if you don’t understand the rounding problems in IEEE floating point mathematics well enough that you think adding more bits will make it accurate enough for financial work, it’s possible it’s not Coder Stockholm Syndrome and just a bad case of Too Stupid To Code With People’s Money.)
It’s pretty dissapointing to read some of the comments in the blog post that kicked off the discussion, though.
Certain special interest groups subverted the standardization process (again) and pressed through changes to introduce in the C programming language extensions to support decimal floating point computations. 99.99% of all the people will never use this stuff and still we have to live with it.
Those special interest groups are, presumably, “people who would like accurate numbers when they do mathematics, and don’t mind taking a speed and storage hit to get them.” Since they tend to be people like me who make sure your bank accounts add up correctly, I think it’s a reflection of slightly more than 1% of people, if only indirectly.
I refuse to add support for this to glibc because these extensions are not (yet) in the official language standard. And maybe even after that we’ll have it separately.
Boy, that will improve the quality of glibc.
We’ve come a long way since then. Memory use is not that much of a premium anymore
Until we start virtualising systems, a topic that has become near and dear to me of late.
