[Python-Dev] Decimal data type issues (original) (raw)
Andrew P. Lentvorski, Jr. bsder at allcaps.org
Thu Apr 15 14:08:36 EDT 2004
- Previous message: [Python-Dev] Decimal data type issues
- Next message: [Python-Dev] Decimal data type issues
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]
On Thu, 15 Apr 2004, Michael Chermside wrote:
I strongly object. I have one particular use case for decimal, and I consider it to be a very important one.... to store and manipulate monetary amounts. Putting a limit as small as that used for floats would significantly harm my use case.
I presume that you are objecting to only having a default limit of 15 digits of precision? I can certainly understand your position.
However, why does having a soft default limit set at this level cause you an issue? If you need more, you can move it up. You have to override the default precision (currently 8) anyway.
If you want a default limit much higher than this, then there needs to be a way for people to determine when they leave the realm of "fast, but narrower range" and enter the realm of "slow, but much greater range".
What I am trying to establish is a "sensible default" limit which allows most people to use fast operations, does not limit those people who want greater range (at the expense of slower operation), does not impose undue constraints on an optimized version of the module (either software or later hardware), and provides useful feedback for those who suddenly find themselves moving from one group to another (either intentionally or unintentionally).
Part of creating Decimal is to provide a useful module for currency and other applications which require accurate Decimal arithmetic.
But part of creating Decimal is also to be newbie friendly. Letting people know when they pass outside of "sensible" regimes is friendly.
And here I was thinking that someone manipulating financial amounts might actually need speed on a silly operation known as "division". My mistake.get overflow errors instead of out-of-memory errors (and possibly some vague future advantage to implementers of a "complete numerical package" including things like sin() and cos()).
-a
- Previous message: [Python-Dev] Decimal data type issues
- Next message: [Python-Dev] Decimal data type issues
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]