[webkit-dev] Stumped by Date() results in Safari

[Kevin McCullough]

Hi and sorry for my delayed reply.  Yes I've done some re-work in the  
date code recently, which should treat dates outside of 2038  
correctly, but which are not available in any Safari version that has  
shipped yet.

I wouldn't be too sure that it's not Safari's fault just because  
someone should have found it by now, it may have been encountered, or  
it may be that it's not that big of a deal for most users, since  
dates after 2038 are usually not interesting.  However Chris is right  
that that year is the edge of the "compatible range" which is a fun  
quirk in most date implementations, that may have existed before my  
changes.

Likewise I don't know of a repository of Safari quirks, but hopefully  
you can get what you need from this mailing list or our irc channel  
at #webkit on irc.freenode.net.

Hope this helps.

ð Kevin McCullough
kmccullough at apple.com



On Feb 13, 2007, at 1:47 PM, Chris Brichford wrote:

> I have also encountered problems with the date code in WebKit.  I was
> able to fix the bugs I care about ( windows port crashed ) by  
> upgrading
> to the latest code from WebKit.org.  Since it sounds as though you are
> an AJAX developer trying to get your app working in Safari, that might
> not help you too much.  The date object's implementation seems to have
> been totally redone recently.  I would not be surprised if the old  
> code
> is broken for dates outside of the "compatible range" ( 1970/1/1 -
> sometime in 2038 ) in one way or another.
>
> If you want to see the C++ code in question look at:
> http://trac.webkit.org/projects/webkit/browser/trunk/JavaScriptCore/ 
> kjs/
> date_object.cpp
> http://trac.webkit.org/projects/webkit/browser/trunk/JavaScriptCore/ 
> kjs/
> DateMath.cpp
>
> User kmccullo seems to be the person who has submitted most of the
> changes in this area recently.
>
> Hope this helps,
> Chris
>
> -----Original Message-----
> From: webkit-dev-bounces at lists.webkit.org
> [mailto:webkit-dev-bounces at lists.webkit.org] On Behalf Of Jason White
> Sent: Tuesday, February 13, 2007 1:21 PM
> To: Geoffrey Garen
> Cc: webkit-dev at lists.webkit.org
> Subject: Re: [webkit-dev] Stumped by Date() results in Safari
>
> Hmm. Yes, that's it. The machines having the problem all run 10.3.9  
> and
> Safari reports being 1.3.2. I suppose it could be a problem that was
> fixed, but since I think people would have noticed this I've got a
> feeling that it's probably something external to Safari.. I'm just
> really puzzled by why it would only affect that browser on the very  
> same
> machines.
>
> I'm in a situation where I have to support Safari for various reasons,
> however, I'm not really an expert on it.. so if someone spots a big
> newbie mistake please accept my apologies! As I said, I tried to find
> the answer in documentation but got nowhere. Thanks for sharing the
> bandwidth and time..
>
> Jason
>
> On 2/13/07, Geoffrey Garen <ggaren at apple.com> wrote:
>> Jason,
>>
>> I tried this in Safari on 10.4.8:
>>
>> javascript:alert(new Date('01/19/2039').getTime())
>>
>> and got:
>>
>> 2179036800000
>>
>> which is greater than 214748364700.
>>
>> Am I testing the right thing?
>>
>> Geoff
>>
>> On Feb 13, 2007, at 12:57 PM, Jason White wrote:
>>
>>> If I'm posting to a very wrong place, please let me know. I'm having
>
>>> a problem with a Javascript date in Safari that is preventing me
>>> from supporting it alongside Firefox, IE, and Navigator.
>>>
>>> It appears that Safari has a problem when I attempt to use new
>>> Date() and pass in a string formatted date greater than 01/19/2038.
>>> Using  x = new Date('01/19/2038').getTime(), it appears that the
>>> milliseconds since 01/01/1970 never goes over 214748364700 no matter
>
>>> what you set the date to..but it can be less. IE and Firefox will go
>
>>> much higher than that. Converting that number to binary, it doesn't
>>> look like anything is getting overflowed (it's not all 1s)
>>>
>>> Can anyone help me figure out what is wrong here? I can't imagine
>>> it's Safari's fault, or someone would have noticed by now. I don't
>>> think it's a system configuration issue because the other browsers
>>> don't have the same problem.
>>>
>>> Also, if anyone knows of a good reference book on this browser
>>> please let me know! For some reason I'm having a hard time finding
>>> much information on developing for it's peculiarities.
>>>
>>> Thanks,
>>>
>>> Jason White
>>> Developer
>>> New England Research Institute
>>> _______________________________________________
>>> webkit-dev mailing list
>>> webkit-dev at lists.webkit.org
>>> http://lists.webkit.org/mailman/listinfo/webkit-dev
>>
>>
> _______________________________________________
> webkit-dev mailing list
> webkit-dev at lists.webkit.org
> http://lists.webkit.org/mailman/listinfo/webkit-dev
> _______________________________________________
> webkit-dev mailing list
> webkit-dev at lists.webkit.org
> http://lists.webkit.org/mailman/listinfo/webkit-dev

-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.macosforge.org/pipermail/webkit-dev/attachments/20070213/c3899343/attachment.html