[ 
https://issues.apache.org/jira/browse/SLING-4258?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14280221#comment-14280221
 ] 

Bertrand Delacretaz commented on SLING-4258:
--------------------------------------------

Thanks! 

Do you know if we can reliably expect timezone names like 
""America/Los_Angeles" and Europe/Moscow to be available on all "reasonable" 
JVMs? 

The TimeZone javadocs says you get GMT if a name is not present, which makes 
the test pointless in such a case IIUC.

Using TimeZone.getAvailableIDs(...) with a range of offsets to select a few 
timezones looks safer to me, WDYT?

I hate to make things more complicated, but such subtle issues in tests can be 
painful, so safer sounds better to me!





> Please document better how dates are handled by the Post servlet
> ----------------------------------------------------------------
>
>                 Key: SLING-4258
>                 URL: https://issues.apache.org/jira/browse/SLING-4258
>             Project: Sling
>          Issue Type: Improvement
>          Components: Documentation
>            Reporter: santiago garcía pimentel
>
> Im currently doing some things with dates in Sling that involve timezones and 
> I find that the documentation regarding it is not particularly clear.
> according to 
> https://sling.apache.org/documentation/bundles/manipulating-content-the-slingpostservlet-servlets-post.html#date-properties
> several formats are defined. 
> I found that the only format that saves a provided timezone is the ISO8601 
> format, rest of them relies in a Date object, which does not have timezones. 
> Could this be clearly stated?
> Also, the ISO8601 parser is problematic. It relies on the Jackrabbit parser 
> which uses format "±YYYY-MM-DDThh:mm:ss.SSSTZD", but according to 
> http://www.w3.org/TR/NOTE-datetime the ISO format does not have milliseconds 
> on it ("SSS"). So it is very hard to find a way to keep the timezone 
> information (I had to dig through the code to figure it out)
> Could we please replace ISO8601 with the actual format 
> "±YYYY-MM-DDThh:mm:ss.SSSTZD" so it is clearer?



--
This message was sent by Atlassian JIRA
(v6.3.4#6332)

Reply via email to