Dear Peter, Thank you for review. I submitted -06 (-05 contains some nits) as a revised version.
On 2015-06-26 08:15, Peter Yee wrote:
Some of the parameters are described with their bit lengths, others are not. Lengths are listed haphazardly. Be consistent in either including them or not. Timer lengths are listed as n-bit integers, but this may be confusing given that the document says "Short floating point format is used to describe the wide range of timer values". So, are the timer values floating point or integer? Is the combined implication supposed to be that the numbers are formatted as floating point (for greater range), but only take on integer values? Given that the values in this document are all given in reference to division by the value of TUNIT, it's possible that values might not be integers. If that happens and only integers are desired, is rounding or truncation preferred? Also note that Appendix A indicates that "short unsigned floating point" was dropped. It's not completely clear if that means that "short [signed?] floating point" was adopted in its stead or whether actual short integers were preferred.
Short floating point is dropped but some texts have not removed (my mistake). Thank you for pointing out. All values should be in integer. Thank you again for your efforts on review (and sorry for lot of nits...) Yusuke _______________________________________________ Gen-art mailing list Gen-art@ietf.org https://www.ietf.org/mailman/listinfo/gen-art
