whoops, accidentally replied to the last sender and not the whole list...
On 7/5/06, Kevin Damm [EMAIL PROTECTED] wrote:
One approach I've found some success with (but haven't tested a great
deal) is to load one page that checks a session variable or cookie
value and if it isn't found loads
One of the restrictions on XMLHttpRequests is that they can only work
on the same domain as the script they originated from. This is a
javascript security feature to avoid malicious scripts found on
another domain. Unfortunately, this means that you can't even share
script information between
Sometimes a web server is set up with a root directory different than
the system's root directory... it could be that /abspathto/ is not the
real path to MochiKit. Try treating the root directory as the
document root for your web server and see if that works.
- Kevin
On 8/29/06, [EMAIL
From what I understand, it's discouraged to use functions like isIe,
isGecko, etc. to fork your condition on how you calculate or
manipulate objects.
There are a few good reasons for this, especially that the box model
in future versions (or different current versions) of the same browser
might
On 10/31/06, Zachery Bir [EMAIL PROTECTED] wrote:
On Oct 31, 2006, at 3:03 PM, abargnesi wrote:
Hello,
I am using the DOM Manipulation js library that is part of Mochikit
and
I like it a lot. And it seems quite fast. However I would like it to
handle one case that it doesn't.
It looks like more than just IE has a problem with the onchange event.
According to quirksmode
http://www.quirksmode.org/js/events_compinfo.html#forms
the onchange event is buggy for Safari and Opera as well, not being
implemented for some form elements. It may be that the event doesn't
I am more concerned with when some input fields are left, so that I can update
some calculations.
I don't have a Windows box nearby to test IE on, but onblur propogates
to the parent form in Firefox. This way you can connect one signal
and catch when the user leaves a form element (assuming
Or,
return (compare(americanDate($('exsupplier_date').value),
americanDate($('arrival_date').value) != 1);
On 2/26/07, Matthew Kwiecien [EMAIL PROTECTED] wrote:
More simple how? So you don't need the split and piecing back together? If
so, then how about:
if
It's probably an issue with IE7 aggressively caching the request. Try
adding a unique string (e.g. datetime and a random number) to the end
of your request as part of the query string so that IE views it as a
unique page.
So, an example,
http://mysite.com/ajax-process.php?var=value
would become
Nothing really pops out at me from your example code. Do you get any
Javascript warnings or errors when that input field gets focus? Or,
better yet, do you have a publicly-accessible demo page that we can
poke around at?
- Kevin
On 11/5/07, feesh [EMAIL PROTECTED] wrote:
Hi all,
The most likely reason you are getting this message is because your
page is served from an https:// site and some content (images, etc.)
is being served from an http:// site.
This message appears (even without javascript usage) when content from
insecure http: is on the same page as secure
Maybe use MochiKit.Base.merge or MochiKit.Base.update:
function CHECKBOX() {
return INPUT.apply(this, update(arguments, {'type': checkbox}));
}
- Kevin
On Dec 17, 2007 1:30 PM, machineghost [EMAIL PROTECTED] wrote:
I recently found myself doing this a lot:
var a =
Oops, that should probably be arguments[0]
On Dec 17, 2007 2:00 PM, Kevin Damm [EMAIL PROTECTED] wrote:
Maybe use MochiKit.Base.merge or MochiKit.Base.update:
function CHECKBOX() {
return INPUT.apply(this, update(arguments, {'type': checkbox}));
}
- Kevin
On Dec 17, 2007 1:30 PM
it in everywhere just because
it might be possible :)
-bob
On 12/17/07, Kevin Damm [EMAIL PROTECTED] wrote:
Oops, that should probably be arguments[0]
On Dec 17, 2007 2:00 PM, Kevin Damm [EMAIL PROTECTED] wrote:
Maybe use MochiKit.Base.merge or MochiKit.Base.update:
function CHECKBOX
But, then you are not url-encoding the quotes in the query string.
From the documentation for encodeURIComponent:
http://developer.mozilla.org/en/docs/Core_JavaScript_1.5_Reference:Global_Functions:encodeURIComponent
it looks like the conversion from Latin-1 umlaute to UTF-8 byte
sequence is
15 matches
Mail list logo
