Thank you Alan. The numbers I want to display, come from a file, to the server, then the server sends those numbers to GWT. So I don't have control generating the numbers as rational numbers. The solution that worked for me is to read the numbers on the server as double, then send them to GWT. GWT displays the numbers correctly and performing calculation works well too. I think it's the conversion from float to double that causes this problem.
- Tony On Wed, Feb 25, 2009 at 9:50 AM, Alan Hadsell <ahads...@gmail.com> wrote: > > > > On Feb 24, 1:52 pm, "tony.p.." <tony.t....@gmail.com> wrote: > > > This seems like a bug in JS, more than just the way it handles float. > > Because I'm not doing any calculations on the GWT (JS) side, I'm just > > printing what comes from the server. So if 15.2 if printed as > > 15.2000000000, that would be ok, but to be printed as > > 15.199999809265136 is really weired. Am I right or am I missing > > something? > > You're missing something. In JavaScript there is no such number as > 15.2. JavaScript uses IEEE-754 standard floating-point notation, > which can only represent numbers of the form (i times 2 ** j) where ** > represents exponentiation, and i and j are integers. See > http://babbage.cs.qc.edu/courses/cs341/IEEE-754references.html for > details. The effect is similar to trying to represent 1/3 with an > exact decimal representation; you can come very close but you can't > get it exactly. > > If you want to represent exact decimal numbers in JavaScript you have > two choices: Use strings, and forego doing arithmetic on them, or use > rational numbers, where all numbers are internally in the form (i/j) > with i and j integers. > > One very useful subset of rational notation is scaled decimal, where j > is a constant power of 10 depending on the maximum precision you want > to retain. So if you're maintaining 4 digits after the decimal place, > j is 10000, and 15.2 is represented as 152000/10000. The denominator > (also called the scale factor) can be implied in this case (i.e. you > don't need to store it with each number; think of it as a static final > value). > > This notation makes it easy to add and subtract; just add or subtract > the numerators. Multiplication is easy if the numbers are small -- > multiply the numerators and then divide by the scale factor. Division > is relatively easy; multiply one numerator by the scale factor and > then divide by the other numerator. In both multiplication and > division you need to think about rounding, and also about overflow if > any of the intermediate results might exceed 15 digits. > - Show quoted text - > > > > --~--~---------~--~----~------------~-------~--~----~ You received this message because you are subscribed to the Google Groups "Google Web Toolkit" group. To post to this group, send email to Google-Web-Toolkit@googlegroups.com To unsubscribe from this group, send email to google-web-toolkit+unsubscr...@googlegroups.com For more options, visit this group at http://groups.google.com/group/Google-Web-Toolkit?hl=en -~----------~----~----~----~------~----~------~--~---
