# Ruby-dev summary 28495-28605

**URL:** <https://www.ruby-forum.com/t/ruby-dev-summary-28495-28605/59341>\
**Category:** Ruby\
**Created:** [May 3, 2006, 6:57pm UTC](https://www.ruby-forum.com/t/ruby-dev-summary-28495-28605/59341 "2006-05-03T18:57:44Z")\
**Posts on this page:** 1\
**Page:** 1

<div class="post-metadata">

**Author:** ![Takaaki\_T](https://www.ruby-forum.com/letter_avatar_proxy/v4/letter/t/dc4da7/32.png) [@Takaaki\_T](https://www.ruby-forum.com/u/Takaaki_T)\
**Post date:** [May 3, 2006, 6:57pm UTC](https://www.ruby-forum.com/t/ruby-dev-summary-28495-28605/59341/1 "2006-05-03T18:57:44Z")

</div>

Here is a summary of the articles posted to ruby-dev.

[ruby-dev:28502] rb\_gc\_call\_finalizer\_at\_exit()

Nagai, who is a maintainer of tcltklib, asked how to know that  
functions are  
called by rb\_gc\_call\_finalizer\_at\_exit() in the C layer in order to  
fix a  
problem of tcltklib. However, the current ruby doesn’t provide such a  
way.

[ruby-dev:28456] Kconv.toxxx with -m0

Naruse requested for comment on the proposal to change the behavior of  
the  
methods of Kconv such as “tojis”. Currently, those methods perform  
mime decoding  
in default, since nkf runs with the option ‘-m’(enable mime decoding)  
in default.  
If this change is accepted, the nkf module of ruby calls nkf with  
‘-m0’(disable  
mime decoding) and those methods will not perform mime decoding in  
default.

[ruby-dev:28509] improving the speed of rational

Funaba proposed that ruby should adopt the rational library[1]  
developed by Hara, since  
it is faster than the current rational library. This proposal has been  
accepted, and it  
will be included in ruby-1.9.  
[1] [http://raa.ruby-lang.org/project/rational/](http://raa.ruby-lang.org/project/rational/)

[ruby-dev:28589] Float#div and Float#divmod

Hara proposed that -3.4.div(1.1) should return -4 instead of -3 in  
ruby-1.8.4 as is in the  
case of ruby-1.9. In addition, he proposed that -3.4.divmod(1.1)  
should return [-4, 1.0]  
instead of [-4.0, 1.0], since it should be the pair of the results of  
div and %. This  
proposal has been accepted.