Changes between Version 3 and Version 4 of WG1Ballot5Results


Ignore:
Timestamp:
04/09/12 07:46:13 (5 years ago)
Author:
alexshinn
Comment:

updating missing Shinn vote for #275

Legend:

Unmodified
Added
Removed
Modified
  • WG1Ballot5Results

    v3 v4  
    8989    * [wiki:WG1BallotLucier Lucier]: yes 
    9090    * [wiki:WG1BallotMedernach Medernach]: yes 
     91    * [wiki:WG1BallotShinn Shinn]: no 
    9192    * [wiki:WG1BallotSnellPym SnellPym]: yes, no 
    9293  * '''Results:''' '''yes''', no 
    93   * '''Ratios:''' 6:1 
     94  * '''Ratios:''' 6:2 
    9495  * '''Rationales:''' 
    9596 
     
    102103 `Medernach`:: 
    103104  The sign bit of `NaNs` is meaningless in the IEEE Standard 754 floating-point formats, but in order to avoid using it as an identifier we may consider both as synonymous. However I really would prefer using `NaN` instead of `+nan.0` or `-nan.0` (as it is not a number but an indication of failure, to help diagnostic, it is neither signed in general, nor exact nor inexact, isn't it ?) 
     105 `Shinn`:: 
     106  I don't understand the motivation behind this. NaN can neither be positive nor negative - the "+" is just a hint that the value is numeric (which is no longer true with the symbol extensions that now allow an initial "+"). Thus "-" serves no purpose and does not simplify the syntax - if we allow multiple forms, do we also allow "nan.0", "+nan.", "+nan", "+nan.1", etc.? One is easier to remember than two. 
    104107 
    105108=== #278 Shrink division routines to just truncate and floor === 
     
    118121Vote `shrink` to prune to `truncate-*` (R5RS) and `floor-*` (R5RS `modulo`), moving 
    119122the extra operators to the large language; `shrink/core` to do the same as `shrink` 
    120 but move the remaining operators to the core language; or `1` to keep all 18 
     123but move the remaining operators to the core language; or `keep` to keep all 18 
    121124division operators in the small language. 
    122125