About 4 months ago, a leap second is inserted. And it reveals that there
are many implementation diversity for leap seconds.
For example,
JavaScriptCore accepts the leap second in Date#toString format[1], but
rejects it in ISO 8601 format.
On the other hand, SpiderMonkey rejects it in Date#toString format, but
accepts in ISO 8601.
V8 rejects the leap second in ISO format and when it comes in Date#toString
format, heuristic second check(0-59) decides it isn't second and fail to
parse it.
Chakra rejects both, and Carakan shows the same behavior to SpiderMonkey
I think it is better to describe it as NOTE or standardize this behavior.
[1] Thu Dec 31 23:59:60 GMT 1998
[2] 2012-07-01T08:59:60Z
Hello,
About 4 months ago, a leap second is inserted. And it reveals that there
are many implementation diversity for leap seconds.
For example,
JavaScriptCore accepts the leap second in Date#toString format[1], but
rejects it in ISO 8601 format.
On the other hand, SpiderMonkey rejects it in Date#toString format, but
accepts in ISO 8601.
V8 rejects the leap second in ISO format and when it comes in Date#toString
format, heuristic second check(0-59) decides it isn't second and fail to
parse it.
Chakra rejects both, and Carakan shows the same behavior to SpiderMonkey
( quated from this blog post
http://uupaa.hatenablog.com/entry/2012/07/02/113201 )
I think it is better to describe it as NOTE or standardize this behavior.
[1] Thu Dec 31 23:59:60 GMT 1998
[2] 2012-07-01T08:59:60Z
--
Regards,
Yusuke Suzuki
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.mozilla.org/pipermail/es-discuss/attachments/20121125/556298fc/attachment.html>
Hello,
About 4 months ago, a leap second is inserted. And it reveals that there are many implementation diversity for leap seconds.
For example, JavaScriptCore accepts the leap second in Date#toString format[1], but rejects it in ISO 8601 format. On the other hand, SpiderMonkey rejects it in Date#toString format, but accepts in ISO 8601. V8 rejects the leap second in ISO format and when it comes in Date#toString format, heuristic second check(0-59) decides it isn't second and fail to parse it. Chakra rejects both, and Carakan shows the same behavior to SpiderMonkey
( quated from this blog post uupaa.hatenablog.com/entry/2012/07/02/113201 )
I think it is better to describe it as NOTE or standardize this behavior.
[1] Thu Dec 31 23:59:60 GMT 1998 [2] 2012-07-01T08:59:60Z