Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]

Groups > comp.lang.javascript > #18163 > unrolled thread

IE8 attachEvent problem

Back to article view | Back to comp.lang.javascript

Contents

  IE8 attachEvent problem "Mel Smith" <med_cutout_syntel@aol.com> - 2013-01-20 10:42 -0700
    Re: IE8 attachEvent problem Martin Honnen <mahotrash@yahoo.de> - 2013-01-20 18:55 +0100
      Re: IE8 attachEvent problem "Mel Smith" <med_cutout_syntel@aol.com> - 2013-01-20 10:59 -0700
        Re: IE8 attachEvent problem "Mel Smith" <med_cutout_syntel@aol.com> - 2013-01-20 11:24 -0700
          Re: IE8 attachEvent problem Martin Honnen <mahotrash@yahoo.de> - 2013-01-21 12:06 +0100
            Re: IE8 attachEvent problem "Mel Smith" <med_cutout_syntel@aol.com> - 2013-01-21 09:52 -0700
    Re: IE8 attachEvent problem "Evertjan." <exxjxw.hannivoort@inter.nl.net> - 2013-01-20 20:11 +0100
      Re: IE8 attachEvent problem "Mel Smith" <med_cutout_syntel@aol.com> - 2013-01-20 13:21 -0700
        Re: IE8 attachEvent problem Bjoern Hoehrmann <bjoern@hoehrmann.de> - 2013-01-20 21:44 +0100
          Re: IE8 attachEvent problem "Evertjan." <exxjxw.hannivoort@inter.nl.net> - 2013-01-21 00:07 +0100
          Re: IE8 attachEvent problem "Mel Smith" <med_cutout_syntel@aol.com> - 2013-01-20 16:19 -0700
            Re: IE8 attachEvent problem "Mel Smith" <med_cutout_syntel@aol.com> - 2013-01-20 17:18 -0700
      Re: IE8 attachEvent problem Martin Honnen <mahotrash@yahoo.de> - 2013-01-21 18:28 +0100
        Re: IE8 attachEvent problem "Evertjan." <exxjxw.hannivoort@inter.nl.net> - 2013-01-21 19:07 +0100
          Re: IE8 attachEvent problem Martin Honnen <mahotrash@yahoo.de> - 2013-01-21 19:33 +0100
          Re: IE8 attachEvent problem "Mel Smith" <med_cutout_syntel@aol.com> - 2013-01-21 13:05 -0700
            Re: IE8 attachEvent problem "Evertjan." <exxjxw.hannivoort@inter.nl.net> - 2013-01-22 09:00 +0100
    Re: IE8 attachEvent problem Arty Effem <ignoring@mail.com> - 2013-01-23 13:49 -0800

#18163 — IE8 attachEvent problem

Hi:

Environment: As a client using an IE8 browser on a Win XP Pro (sp2) machine:

    Within  the javascript code in my page there is a 'function remember()'. 
It is shown as --- the problem function --- below. The function 'remembers' 
where in a long table my user was viewing, so that when he 'returns' to the 
page, I position his viewport to the 'remembered' set of lines. It has 
worked fine for a long time but under IE7. Recently, I have started using 
IE8 as a testbed (in addition to Chrome and FF), and set IE8 to show me JS 
errors when viewing pages.

    When this remember function is called , IE8 shows a 'Type Mismatch' 
error at character 4 of
the offending line (indicated by asterisks ("*****") below.

    When viewed with Chrome and FF, there is no problem because these 
browsers use addEventListener as the event handler (instead of attachEvent).

    I'm not a newbie, just a stupie,  and would like to continue using this 
'remember' function without the error.

    Hope that someone can help me fix  this error on IE8

Thank you.

--
Mel Smith
Mesa, Arizona

------- the problem function ----------

function remember(e) {
var targ,tid;
if (!e)
  {
  var e=window.event;
  }
if (e.target)
  {
  targ=e.target;
  }
else if (e.srcElement)
  {
  targ=e.srcElement;
  }
if (targ.nodeType==3) // defeat Safari bug
  {
  targ = targ.parentNode;
  }
tid=targ.id;

if (window.addEventListener) {
   window.addEventListener("load",document.getElementById(tid).scrollIntoView(),false) 
;
}
else if (window.attachEvent) {

   // ********** next line shows a 'Type Mismatch' error at char(4) 
********
   window.attachEvent("onload",document.getElementById(tid).scrollIntoView()) 
;
}
else {
   alert("c. No adding or attaching") ;
   window.onload = "document.getElementById(tid).scrollIntoView();"
}

return ;
}






-- 
Mel Smith

[toc] | [next] | [standalone]

#18164

Mel Smith wrote:

> function remember(e) {
> var targ,tid;
> if (!e)
>    {
>    var e=window.event;
>    }
> if (e.target)
>    {
>    targ=e.target;
>    }
> else if (e.srcElement)
>    {
>    targ=e.srcElement;
>    }
> if (targ.nodeType==3) // defeat Safari bug
>    {
>    targ = targ.parentNode;
>    }
> tid=targ.id;
>
> if (window.addEventListener) {
>     window.addEventListener("load",document.getElementById(tid).scrollIntoView(),false)
> ;
> }
> else if (window.attachEvent) {
>
>     // ********** next line shows a 'Type Mismatch' error at char(4)
> ********
>     window.attachEvent("onload",document.getElementById(tid).scrollIntoView())
> ;
> }
> else {
>     alert("c. No adding or attaching") ;
>     window.onload = "document.getElementById(tid).scrollIntoView();"
> }
>
> return ;
> }

With both event APIs (MSHTML/IE as well as the W3C DOM) you need to pass 
in a function object that you want to be called, not call a function. So 
you need
   window.addEventListener("load"
     function() {
       document.getElementById(tid).scrollIntoView();
     },
     false
   );
respectively
   window.attachEvent("onload", function() { 
document.getElementById(tid).scrollIntoView(); });

On the other hand if your remember function is called onload then it 
should suffice to simply call the scrollIntoView, as your current code 
does anyway, only accidentally.

So depending on when/where you execute the remember function, you might 
not need to use addEventListener respectively attachEvent, you can 
simply put your
    document.getElementById(tid).scrollIntoView();
in the remember function.

[toc] | [prev] | [next] | [standalone]

#18165

Martin said:

>
> With both event APIs (MSHTML/IE as well as the W3C DOM) you need to pass 
> in a function object that you want to be called, not call a function. So 
> you need
>   window.addEventListener("load"
>     function() {
>       document.getElementById(tid).scrollIntoView();
>     },
>     false
>   );
> respectively
>   window.attachEvent("onload", function() { 
> document.getElementById(tid).scrollIntoView(); });
>
> On the other hand if your remember function is called onload then it 
> should suffice to simply call the scrollIntoView, as your current code 
> does anyway, only accidentally.
>
> So depending on when/where you execute the remember function, you might 
> not need to use addEventListener respectively attachEvent, you can simply 
> put your
>    document.getElementById(tid).scrollIntoView();
> in the remember function.
>

    Hi Martin:

    I'll print and study your response.

Thank you !

-Mel Smith

[toc] | [prev] | [next] | [standalone]

#18166

Hi Martin:

    Below is how the 'remember' function is currently called:

    There are approx 150 buttons on the page. When *any* button is clicked, 
I 'submit' the form and a new page is sent to the browser. When the user 
'backs up', then I expect the previous page to be displayed in the viewport 
at the appropriate set of lines which were 'remembered'.

    It *does this* but still shows the error on IE8.

Thanks,

-Mel

----- remember() is *only* 'called' here --------
<form id="someid" name="frmmenu" action="/cgi-bin/xxx/myscript.exe"
 method="post" onclick="remember(event);">
------------------------------------------------

[toc] | [prev] | [next] | [standalone]

#18187

Mel Smith wrote:

>      It *does this* but still shows the error on IE8.


I am not sure about older IE versions, according to 
http://msdn.microsoft.com/en-us/library/ie/ms536730%28v=vs.85%29.aspx 
scrollIntoView takes a boolean argument.
Does the error vanish when you explicitly call
    document.getElementById(tid).scrollIntoView(true);
? Or which statement currently gives the error?

[toc] | [prev] | [next] | [standalone]

#18189

Martin said:

"Martin Honnen" <mahotrash@yahoo.de> wrote in message 
news:50fd2118$0$6563$9b4e6d93@newsspool4.arcor-online.net...
> Mel Smith wrote:
>
>>      It *does this* but still shows the error on IE8.
>
>
> I am not sure about older IE versions, according to 
> http://msdn.microsoft.com/en-us/library/ie/ms536730%28v=vs.85%29.aspx 
> scrollIntoView takes a boolean argument.
> Does the error vanish when you explicitly call
>    document.getElementById(tid).scrollIntoView(true);
> ? Or which statement currently gives the error?
>

Hi Martin:

    In a 'later/lower' post, I corrected statement in question and used 
'document.body. etcetera' .  This stopped the error.

    With that correction, I now have the following results:

    IE8 - Works corectly
    FF -   Works correctly
    Opera - Works correctly
    Chrome - Page displays at *top*, not where it was 'remembered'
    Safari - Page displays at *top*, not where it was 'remembered'

    Since my users are older retired north americans, and since in this 
demographic (IE and Chrome) are predominant,  I'd like to determine why 
Chrome is not scrolling to the remembered position (which is a multi-line 
cell in a long 3-column table)

    btw,  I *did* place a 'true' as an argument in the scrollIntoView() 
function.  But, this is the default anyway and made no difference.

    Also, I placed an Alert() just ahead of the 'scroll' code to ensure that 
the ID was correct. It _was_

    So, if I change the 'onload' event to point at my new ID, why is it 
ignored by Chrome and Safari when I 'return' to the page ??

Thanks for the comment.

-Mel Smith

[toc] | [prev] | [next] | [standalone]

#18167

Mel Smith wrote on 20 jan 2013 in comp.lang.javascript:

>    window.onload = "document.getElementById(tid).scrollIntoView();"

If you change this to:

body.onload = 'document.getElementById(tid).scrollIntoView();'

... shouldn't it work on most browsers, 
"eventually" not needing any other "attaching"?

-- 
Evertjan.
The Netherlands.
(Please change the x'es to dots in my emailaddress)

[toc] | [prev] | [next] | [standalone]

#18168

Evertjan said:
>>    window.onload = "document.getElementById(tid).scrollIntoView();"
>
> If you change this to:
>
> body.onload = 'document.getElementById(tid).scrollIntoView();'
>
> ... shouldn't it work on most browsers,
> "eventually" not needing any other "attaching"?
>

Hi Evertjan:

    I've changed my remember() function to what is show below, and 
re-tested:

Result:

    On IE8, I get the error 'body is undefined'

    BUT, it works on IE8 and I return to the 'remembered' set of lines in 
the table even with the 'indefined' error ???

    On Chrome, FF, ands Safari, I'm not sure what is going on ...

-Mel

--------------------------
function remember(e) {
var targ,tid;
if (!e)
  {
  var e=window.event;
  }
if (e.target)
  {
  targ=e.target;
  }
else if (e.srcElement)
  {
  targ=e.srcElement;
  }
if (targ.nodeType==3) // defeat Safari bug
  {
  targ = targ.parentNode;
  }
tid=targ.id;

body.onload="document.getElementById(tid).scrollIntoView();"
return;
}

[toc] | [prev] | [next] | [standalone]

#18169

* Mel Smith wrote in comp.lang.javascript:
>    On IE8, I get the error 'body is undefined'

>body.onload="document.getElementById(tid).scrollIntoView();"

You are probably looking for `document.body`.
-- 
Björn Höhrmann · mailto:bjoern@hoehrmann.de · http://bjoern.hoehrmann.de
Am Badedeich 7 · Telefon: +49(0)160/4415681 · http://www.bjoernsworld.de
25899 Dagebüll · PGP Pub. KeyID: 0xA4357E78 · http://www.websitedev.de/ 

[toc] | [prev] | [next] | [standalone]

#18174

Bjoern Hoehrmann wrote on 20 jan 2013 in comp.lang.javascript:

> * Mel Smith wrote in comp.lang.javascript:
>>    On IE8, I get the error 'body is undefined'
> 
>>body.onload="document.getElementById(tid).scrollIntoView();"
> 
> You are probably looking for `document.body`.

Right. 

However the above still cannot be working 
in a function that is executed after the page is ready,
as the unload executes as the last part of the loading process.


I suppose the OP wants the page to store somewhere the position
of the page shown, and retrieved in a next invocation of that same page 
later. The document.body.onload='' or the <body onload='..'> should work on 
that next page if executed during the loading. 

-- 
Evertjan.
The Netherlands.
(Please change the x'es to dots in my emailaddress)

[toc] | [prev] | [next] | [standalone]

#18175

Bjoern said:

>* Mel Smith wrote in comp.lang.javascript:
>>    On IE8, I get the error 'body is undefined'
>
>>body.onload="document.getElementById(tid).scrollIntoView();"
>
> You are probably looking for `document.body`.


Hi Bjoern:

    After taking a little nap, I awoke with a light-bulb bubble over my head 
which was, of course, which you noted above.

-Mel

[toc] | [prev] | [next] | [standalone]

#18178

Bjoern, Martin and Evertjan:

>>>body.onload="document.getElementById(tid).scrollIntoView();"
>>
>> You are probably looking for `document.body`.

Hi:

The statement: 
document.body.onload="document.getElementById(tid).scrollIntoView();"

    ... works perfectly for IE8 and Safari

    For Chrome, the original page displays at the top and not where the 
viewer left it.

    I'll do more testing.

Thanks to y'all !

-Mel Smith

[toc] | [prev] | [next] | [standalone]

#18190

Evertjan. wrote:
> Mel Smith wrote on 20 jan 2013 in comp.lang.javascript:
>
>>     window.onload = "document.getElementById(tid).scrollIntoView();"
>
> If you change this to:
>
> body.onload = 'document.getElementById(tid).scrollIntoView();'
>
> ... shouldn't it work on most browsers,
> "eventually" not needing any other "attaching"?

I am confused, which browsers support assigning a string with Javascript 
code as an event handler to the DOM property reflecting an attribute?
Wouldn't you need to assign a function?

[toc] | [prev] | [next] | [standalone]

#18191

Martin Honnen wrote on 21 jan 2013 in comp.lang.javascript:

> Evertjan. wrote:
>> Mel Smith wrote on 20 jan 2013 in comp.lang.javascript:
>>
>>>     window.onload = "document.getElementById(tid).scrollIntoView();"
>>
>> If you change this to:
>>
>> body.onload = 'document.getElementById(tid).scrollIntoView();'
>>
>> ... shouldn't it work on most browsers,
>> "eventually" not needing any other "attaching"?
> 
> I am confused, which browsers support assigning a string with Javascript 
> code as an event handler to the DOM property reflecting an attribute?
> Wouldn't you need to assign a function?

Your testing is as good as my testing, I suppose.

Methinks both should do fine, you test it:

================
document.body.onload = 'alert("I`m loaded");'
================

and

================
function sureIam() {
  alert("I`m loaded");
}; 

document.body.onload = sureIam;
================

Mind: no () in the latter.

-- 
Evertjan.
The Netherlands.
(Please change the x'es to dots in my emailaddress)

[toc] | [prev] | [next] | [standalone]

#18192

Evertjan. wrote:
> Martin Honnen wrote on 21 jan 2013 in comp.lang.javascript:
>
>> Evertjan. wrote:


>>> If you change this to:
>>>
>>> body.onload = 'document.getElementById(tid).scrollIntoView();'
>>>
>>> ... shouldn't it work on most browsers,
>>> "eventually" not needing any other "attaching"?
>>
>> I am confused, which browsers support assigning a string with Javascript
>> code as an event handler to the DOM property reflecting an attribute?
>> Wouldn't you need to assign a function?
>
> Your testing is as good as my testing, I suppose.
>
> Methinks both should do fine, you test it:
>
> ================
> document.body.onload = 'alert("I`m loaded");'

I don't see why assigning a string would do, it doesn't work for me 
(http://jsfiddle.net/kF8se/). Tested with current versions of FF, IE, 
Chrome, Opera on Windows 8.

Assigning a function is different (http://jsfiddle.net/t6JVX/1/), works 
fine with above named browsers.

What you could do is use setAttribute (http://jsfiddle.net/TEE8c/1/) if 
you really want to work with a string but I wouldn't recommend that due 
to the way setAttribute works with older IE versions.

[toc] | [prev] | [next] | [standalone]

#18193

Evertjan said:

> Methinks both should do fine, you test it:
>
> ================
> document.body.onload = 'alert("I`m loaded");'
> ================
>
> and
>
> ================
> function sureIam() {
>  alert("I`m loaded");
> };
>
> document.body.onload = sureIam;
> ================
>
> Mind: no () in the latter.


Hi Evertjan:

    How do I change my 'remember' code to make it an 'onload function call' 
and pass it the target and ID -- instead of a string.  I'm getting more and 
more confused here :(.

Thank you

-Mel 

[toc] | [prev] | [next] | [standalone]

#18196

Mel Smith wrote on 21 jan 2013 in comp.lang.javascript:

> Evertjan said:
>> document.body.onload = sureIam;
>> ================
>>
>> Mind: no () in the latter.

>     How do I change my 'remember' code to make it an 'onload function
>     call' 
> and pass it the target and ID -- instead of a string.  I'm getting
> more and more confused here :(.

Sorry, I only tried to address the onload issue.

-- 
Evertjan.
The Netherlands.
(Please change the x'es to dots in my emailaddress)

[toc] | [prev] | [next] | [standalone]

#18226

On Sunday, 20 January 2013 17:42:14 UTC, Mel Smith  wrote:
> Hi:
> 
> 
> 
> Environment: As a client using an IE8 browser on a Win XP Pro (sp2) machine:
> 
> 
> 
>     Within  the javascript code in my page there is a 'function remember()'. 
> 
> It is shown as --- the problem function --- below. The function 'remembers' 
> 
> where in a long table my user was viewing, so that when he 'returns' to the 
> 
> page, I position his viewport to the 'remembered' set of lines. It has 
> 
> worked fine for a long time but under IE7. 

That function has never worked, it just looked as though it did.

attachEvent/addeventLister expect a function reference and that's not what you're passing them. You are executing scrollIntoView immediately, without waiting for load.

window.addEventListener("load", function(){ document.getElementById(tid).scrollIntoView(); } ,false);


etc. It would be better to define the function separately and just pass its name ( without () ) in the three places in the code.

[toc] | [prev] | [standalone]

Back to top | Article view | comp.lang.javascript

csiph-web