Prev: Java Front end Developer | NYC | Long term (Only GC/Citizens)
Next: Google Closure: The Baseless Parts
From: Richard Maher on 18 Nov 2009 04:33
Hi,

Can anyone confirm that IE 8 prevents a page that was loaded via httpS from
injecting a script (createElement, appendChild etc) by throwing an
"Operation aborted" exception when trying to set the src attribute to a
HTTP*no*S URL?

I know it's probably a very sensible thing to do and this is only happening
at the moment as we haven't teed-up a certificate for our test server, but
if someone could just confirm the behaviour (or other explanation then
that'd be great.

BTW. Chrome, FF, IE<8 and IE 8 in compatibility mode all do not have an
issue.

Cheers Richard Maher

PS. My colleague found the ONERROR event handler today which is a very
useful, yet seldom talked about, weapon in the injection/callback arsenal.



From: David Mark on 18 Nov 2009 04:33
On Nov 18, 4:33 am, "Richard Maher" <maher...(a)hotspamnotmail.com>
wrote:
> Hi,
>
> Can anyone confirm that IE 8 prevents a page that was loaded via httpS from
> injecting a script (createElement, appendChild etc) by throwing an
> "Operation aborted" exception when trying to set the src attribute to a
> HTTP*no*S URL?

I cannot, but I haven't tried. I wouldn't put anything past MS.

>
> I know it's probably a very sensible thing to do and this is only happening
> at the moment as we haven't teed-up a certificate for our test server, but
> if someone could just confirm the behaviour (or other explanation then
> that'd be great.

I don't know how sensible it is.

>
> BTW. Chrome, FF, IE<8 and IE 8 in compatibility mode all do not have an
> issue.

[...]

>
> PS. My colleague found the ONERROR event handler today which is a very
> useful, yet seldom talked about, weapon in the injection/callback arsenal..

It certainly is _not_.
From: Richard Maher on 18 Nov 2009 17:54
Hi David,

>> PS. My colleague found the ONERROR event handler today which is a very
>> useful, yet seldom talked about, weapon in the injection/callback
arsenal.
> It certainly is _not_.

Seeing as it's coming up to pantomime season in the UK, let me just say "Oh
yes it is!" :-) Unless of course you like polling, don't care whether a
script loaded or not, or are cunningly clever enough to have found another
way (outside of IE) to tell wehn a script failed to load.

Cheers Richard Maher

"David Mark" <dmark.cinsoft(a)gmail.com> wrote in message
news:d3ff725f-0f5c-46f4-9cf3-8527fed07dc4(a)g27g2000yqn.googlegroups.com...
On Nov 18, 4:33 am, "Richard Maher" <maher...(a)hotspamnotmail.com>
wrote:
> Hi,
>
> Can anyone confirm that IE 8 prevents a page that was loaded via httpS
from
> injecting a script (createElement, appendChild etc) by throwing an
> "Operation aborted" exception when trying to set the src attribute to a
> HTTP*no*S URL?

I cannot, but I haven't tried. I wouldn't put anything past MS.

>
> I know it's probably a very sensible thing to do and this is only
happening
> at the moment as we haven't teed-up a certificate for our test server, but
> if someone could just confirm the behaviour (or other explanation then
> that'd be great.

I don't know how sensible it is.

>
> BTW. Chrome, FF, IE<8 and IE 8 in compatibility mode all do not have an
> issue.

[...]

>
> PS. My colleague found the ONERROR event handler today which is a very
> useful, yet seldom talked about, weapon in the injection/callback arsenal.

It certainly is _not_.


From: David Mark on 18 Nov 2009 18:20
On Nov 18, 5:54 pm, "Richard Maher" <maher...(a)hotspamnotmail.com>
wrote:
> Hi David,
>
>
>
> >> PS. My colleague found the ONERROR event handler today which is a very
> >> useful, yet seldom talked about, weapon in the injection/callback
> arsenal.
> > It certainly is _not_.
>
> Seeing as it's coming up to pantomime season in the UK, let me just say "Oh
> yes it is!" :-) Unless of course you like polling, don't care whether a
> script loaded or not, or are cunningly clever enough to have found another
> way (outside of IE) to tell wehn a script failed to load.
>

It is very easy to determine if a script has loaded (in any browser)
and it certainly doesn't involve the onerror property. Use callbacks.
From: Richard Maher on 18 Nov 2009 18:35
Hi David,

> It is very easy to determine if a script has loaded (in any browser)
> and it certainly doesn't involve the onerror property. Use callbacks.

Umm, the "callback" doesn't get called because there was some error loading
the script (or SiteMinder SSO etc substitutes a logon screen) how do you
discover that has happened?

Cheers Richard Maher

"David Mark" <dmark.cinsoft(a)gmail.com> wrote in message
news:6502af84-5c14-49b4-b434-75b4fbef6a77(a)j35g2000vbl.googlegroups.com...
On Nov 18, 5:54 pm, "Richard Maher" <maher...(a)hotspamnotmail.com>
wrote:
> Hi David,
>
>
>
> >> PS. My colleague found the ONERROR event handler today which is a very
> >> useful, yet seldom talked about, weapon in the injection/callback
> arsenal.
> > It certainly is _not_.
>
> Seeing as it's coming up to pantomime season in the UK, let me just say
"Oh
> yes it is!" :-) Unless of course you like polling, don't care whether a
> script loaded or not, or are cunningly clever enough to have found another
> way (outside of IE) to tell wehn a script failed to load.
>

It is very easy to determine if a script has loaded (in any browser)
and it certainly doesn't involve the onerror property. Use callbacks.


 |  Next  |  Last
Pages: 1 2
Prev: Java Front end Developer | NYC | Long term (Only GC/Citizens)
Next: Google Closure: The Baseless Parts