Open Bug 289323 Opened 19 years ago Updated 2 years ago

Back button loads page in wrong frame

Categories

(Core :: Layout: Images, Video, and HTML Frames, defect)

Product:
Core
Component:
Layout: Images, Video, and HTML Frames
Platform:
x86
Windows 2000
Type:
defect

Tracking

()

People

(Reporter: jchittu, Unassigned)

References

Details

(Keywords: helpwanted) 

User-Agent:       Mozilla/5.0 (Windows; U; Windows NT 5.0; en-US; rv:1.7.6) Gecko/20050317 Firefox/1.0.2
Build Identifier: Mozilla/5.0 (Windows; U; Windows NT 5.0; en-US; rv:1.7.6) Gecko/20050319

This is a unique reproducible case of an earlier bug 28142. Oracle configurator
is one of the Oracle application product and this bug can be reproduced in this
product. 

Oracle configurator is DHTML product heavily uses Javascript and Frames.  We
have 3 top level frames header, content frameset and a invisible frameset
consisting of 2 frames which we call src and proxy. The content frameset will
contain contain max of 2 frames and we call tree and a display frame. The tree
frame display a DHTML tree control and the right side display frame will display
the details of the selected node in the tree. The src frames holds all the
javascript to render dhtml controls on the tree and display frame and also
handles events raised from the tree and display frame. All the events are
handled in src frame and we use proxy frame to post to the server. The server
response comes to proxy frame in the form javascript commands to create and
update controls state.

When running the product in Mozilla 1.7.x, Firefox1.0+ or Netscape7.x we run in
to this bug. Whenever browser back button is pressed the contents from the
display frame is rendered wrongly in src frame.

Reproducible: Always

Steps to Reproduce:
This bug can only be reproduced by running the Oracle configurator DHTML UI. I
could not reproduce this bug in a simpler test case. We are happy to show the
bug through a netmeeting session and we are ready to assist you in all the way
we can to make both Mozilla and our product experience better for our customer.

To reproduce this bug we need to launch configurator DHTML ui and click on tree
nodes to do couple of screen navigation. When i press browser back button the
display frame contents (html page) gets rendered on the src frame. It is
supposed to render on the display frame and not in the src frame window. This
causes our product to loose all application state and any further interaction
leads to javascript error. Since the src and proxy frames are hidden the user
does not know what is happening and it resembles that the client application has
died.

I could not reproduce this bug in a simple test case but i can reproduce this
bug from our product everytime at will.


Actual Results:  
Contents from the display frames gets rendered in the src frame.  

Expected Results:  
This is wrong, contents must get rendered correctly in their own frames when
pressing browser back button
*** Bug 289332 has been marked as a duplicate of this bug. ***
When you say Firefox 1.0+, do you mean 1.0.x, or the current betas that are on
the way to becoming 1.1?

If the former, could you possibly test a current trunk build (or at least
Mozilla 1.8beta1)?

Past that, I'm not sure what we have in the way of being able to do a netmeeting
session.... do we have something like that set up at the Foundation?
This is an automated message, with ID "auto-resolve01".

This bug has had no comments for a long time. Statistically, we have found that
bug reports that have not been confirmed by a second user after three months are
highly unlikely to be the source of a fix to the code.

While your input is very important to us, our resources are limited and so we
are asking for your help in focussing our efforts. If you can still reproduce
this problem in the latest version of the product (see below for how to obtain a
copy) or, for feature requests, if it's not present in the latest version and
you still believe we should implement it, please visit the URL of this bug
(given at the top of this mail) and add a comment to that effect, giving more
reproduction information if you have it.

If it is not a problem any longer, you need take no action. If this bug is not
changed in any way in the next two weeks, it will be automatically resolved.
Thank you for your help in this matter.

The latest beta releases can be obtained from:
Firefox:     http://www.mozilla.org/projects/firefox/
Thunderbird: http://www.mozilla.org/products/thunderbird/releases/1.5beta1.html
Seamonkey:   http://www.mozilla.org/projects/seamonkey/

Yes this bug is still reproducible in the latest version of Firefox browser. 
As we said earlier we are happy to assist with everything from Oracle side in 
fixing this bug to make both the browser and our product better.
Gerv, this bug is waiting on the Mozilla foundation folks answering the question
in comment 2, basically.  I can't tell what's going on here and whether it's a
valid bug or not given the information I have thus far.

jchittu@gmail.com, is there anything you can do to let me try this given that I
do not have NetMeeting available?  Is there a way I could access this DHTML app?
The MoFo does not have any facilities for enabling NetMeeting beyond the ability
to run NetMeeting (it comes with every version of Windows). Boris: can you find
a MoFo Windows machine and get Myk/Dave to open the required ports in the firewall?

Alternatively, I may be able to get some space on a demo installation of my
company's "MeetingServer" product, which uses Java to achieve much the same effect.

Other alternatives include VNC and Windows Remote Desktop.

Let me know if I can be of more help.

Gerv
Hmm...  I'd need to vnc into the MoFo machine from over here (Chicago).  If I
can VNC directly into an Oracle machine where I can test that would be much
simpler, probably.
We can use oracle product to do a web conference session. Please to
http://imeeting.oracle.com. Let's fix a time so that i can start a web
conference and i can let you join web conference to see the bug. When do you
want to do. We can do it today also. I'm on east coast please let me know when
do you want to see it. You can update this bug or send a mail directly to me to
see this bug.

Thanks
CK
> Please to http://imeeting.oracle.com.

That seems to require Windows to run, right?

I've been thinking about this bug, and I think the simplest way for it to appear
would be if the page removes or adds frames dynamically; our session history
implementation has known issues with those (especially removals).  Is that being
done here?
1. To use Oracle Web conference you have to have IE browser. I guess even IE on
MAC will work
2. About your question on dynamically adding and removing frames - No we don't
do that. Our framesets are defined in jsp/html pages.
Oracle web conference can support Netscape 4.75. You can use that to see the
bug. Let me know when you folks want to see this.
So my options for browsers are whatever runs on Linux or Solaris...

What I'd really want is to not so much see the bug as see the source of the
pages involved; just seeing a demo of the page isn't going to give me any more
information than is present in comment 0...
I'm pretty sure that NS 4.75 runs on Solaris.
Yes, but when I go to that page you cited in comment 8 using Netscape 4, I get
it complaining that it's unable to download Oracle Web Conferencing Console...

As for when, I think the earliest time I could do that's during business hours
eastern time is Thursday morning sometime (so in 6 days).  Would that work for
you?  Something like 11am Eastern, say?
Next thursday works for me. Let's touch base on Wednesday to confirm the time.
You can also use this URL for oracle web conference
https://conference.oracle.com/imtapp/app/prelogin.uix

I spent several hours today trying to find a way to run the conferencing stuff
on a Linux system, but no matter what I do the site just hands me a .exe file. 
So I really don't see how I can use that stuff, offhand....

Given that, is there any way to get me more information via e-mail or something?
 What would the web-conferencing software allow us to do that we can't do
(possibly a little slower) over e-mail?
To be more precise, the software involved doesn't like my Windows system because
that has no sound output method, and doesn't like my Linux system (which does
have sound output) because it's not Windows....
A web conference session will show you the problem and morover you can take
control and run script debugger to see yourself what is going on. At this point
i suggest you to find a windows machine to see and diagonize the problem. I
spoke to our support people here and they told most of the web conference works
better on windows than other OS. If you can tell me any other web conference
tool than can work beyond wirewall, i will be happy to use that also.
> and run script debugger to see yourself what is going on

Ah, I see.  Unfortunately, there are no Windows machines that I could install
software on other than the one I already mentioned....

bryner, do you have a Windows machine on hand that you could use to look at this
session history issue?
Here is the architecture our configurator product. An xml message is posted to
our product servlet and the servlet returns a top level frameset 
<FRAMESET rows="190,*" ..>
  <FRAME name="header" id="header" SRC="/OA_HTML/czHeader.jsp"...>
  <FRAME name="template" id="template" SRC="/OA_HTML/czFraTemplate.jsp"...>
</FRAMESET>

czFraTemplate.jsp inturn contains 

<FRAMESET rows="*,20"...>
  <FRAME name="czContent" id="czContent" src="/OA_HTML/czCntnt.jsp"...>
  <FRAMESET cols="50%,*"...>
    <FRAME name="czSrc" id="czSrc" src="/OA_HTML/czSource.jsp"....>
    <FRAME name="czProxy" id="czProxy" src="servlet url"...>
  </FRAMESET>
</FRAMESET>

The frame czSrc contains all the necessary JS sources to push content to the
czContent frame. Normally the czContent frame will consist of 2 inner frame
czTree and czDisplay. The czTree display a tree like strucutre and when user
selects a node the selection action is posted to the servlet through the proxy
frame. Servlet returns JS commands that will invoke JS functions on the source
frame to push html content dynamically in to the czTree and czDisplay frames. A
mockup of our application is hosted at 

http://users.rcn.com/ckhima/ieperf/frameset.htm

Click on the header Page 1 Page 2 links, this will simulate an event post on the
proxy frame which will inturn render contents onthe content frame.

Try clicking on browser back button the contents are loaded in their respective
frames correctly as per the browser history. When we run the same thing in our
confiugrator product when we press back button the html content from the
czDisplay gets rendered on the czSrc frame. Due to this all the JS object state
that were there on the czSrc frame is lost resulting in errors

I tried my best to explain the situation in words. Let me know if you need more info
So that mockup shows the general app setup but doesn't show the bug for you, right?
Yes i tried to show the general setup of our product functionality. The bug is
not reproducible you can run and try clicking browser back button you see the
content gets rendered correctly in their respective frames. The bug is
definitely reproducible everytime when we run our product.
Did you guys find a windows machine so that you can see and take a look at the bug?
Confirming bug just in case to save it from Gerv.

It doesn't look like I'll be able to debug this in a reasonable way; could
someone at the foundation take a look?
Status: UNCONFIRMED → NEW
Ever confirmed: true
We are ready to show this bug in a oracle web conference session whenever
someone is ready to look in to this bug
Flags: blocking1.9a1?
Keywords: helpwanted
Flags: blocking1.9a1? → blocking1.9-
Reporter: Is this still a problem ?
Product: Core → Core Graveyard
Component: Layout: HTML Frames → Layout: Images
Product: Core Graveyard → Core
Severity: normal → S3
You need to log in before you can comment on or make changes to this bug.