FLB

[Gouse Basha]

Hi Michael,

I am  trying to scan our apps self register page ,user has to enters input for username and password etc... and user clicks submit button after entering mandotary

field.So i am invoking FLB before submit button ,after flb completes scan the page,page becomes blank and when my script clicks submit button,page just renders it wont got to next page .

Please check snap shot attached with this mail.I have enabled debug mode but i dont find any logs w.r.t to flb.

     please help so that i can integrate this great tool with our frame work ,to eliminate manual efforts to find layout regarding bugs.

thanks in advance.......

--
Thanks & Regards,

Basha

9538029173

[Michael Tamm]

Hi,

In FLB 1.5 when you enable debug mode, all FLB does is to set the log level for the FLB classes to DEBUG. (This will probably change in 1.6 were I want to improve diagnostic information). So if you did not configure any logging framework, the fallback is the JDK logging facility which will log to stdout by default. So you could just send me the the console output.

On the other hand side FLB is not designed in a way that you can reuse the analyzed page (besides clicking on a link). For example FLB clears all JavaScript timers so that any JavaScript animation stops and does not interfere with the layout bug detectors. In your screenshot I see a AJAX spinner. I'm not sure if you ever receive a callback after FLB has run over the page. So my advice is: Always navigate in a fresh window to the page you want to analyze and when you have analyzed that page with FLB, close the browser window.

Kind regards, Michael

2013/4/1 Gouse Basha <mohamm...@gmail.com>

[Saimadhu G]

Hi Tamm,

We (Gouse Basha and me) are running regression with FLB. While regression, we will come across multiple pages because of different kinds of inputs submitted. So, before submitting input in every page, we want to use FLB on the current page to gather any layout issues, then submit the page to get to the next page, do FLB, enter data and submit for next page, and so on. This way, we thought of collecting flb data on every page we come across using regression test case execution.

 

Now, when you say after analyzing with flb, close the page window. Then my doubt is how do we go to the next page for analyzing if the input is not submitted in the current page?

 

Is it possible, while regression save all pages using html source/ snapshots using selenium and then do flb on each page? We need your help in getting this done quickly.

 

Waiting for your reply.

 

Thanks

Madhusudan G

+91 9591244001

 

 

--
You received this message because you are subscribed to the Google Groups "Fighting Layout Bugs" group.
To unsubscribe from this group and stop receiving emails from it, send an email to fighting-layout-...@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.
 
 

[Michael Tamm]

Hi,

copying the DOM of the current page into a new window for analyzing it with FLB sound like a good strategy to me. Although I have not done it before. What you can do additionally is to take screenshot of both the source window and the target window and compare those two screenshots just to make sure that the same layout is rendered.

Kind regards,

Michael

P.S. If you implement this strategy, maybe you can contribute that code back to FLB?

2013/4/13 Saimadhu G <madhusud...@gmail.com>

[Saimadhu G]

Tamm,

In Selenium webdriver I am not seeing any function to get the DOM of the current page. I googled and found there are lots of people asking similar/ same question but no answer.

I discussed with our developers regarding taking the page source/html source (other than DOM) but there is no gurantee that the source window and target window match the same layout.

 

Now i am thinking of 1 possible workaround for the current issue:

Current issue:

As par of regression testing,

1. do flb on the current page

2. //Now we are not able to enter input and go to the next levels.

 

Workaround:

As par of regression testing,

1. store the current page url

2. open a new tab with the current page url

3. do flb on the current page in the new tab (synchronous call to flb from test framework) and close new tab

4. After flb returns, proceed entering the input and submit in the actual tab (not on new tab)

5. repeat steps 1-4 for all regression test cases.

 

the problem with this workaround is the regression execution time will be more because flb is taking too much time to scan the page and regression waits till flb returns.

 

In case, if we are able to get DOM of the current page to rebuild the current page in later point of time, then regression time will be saved much. Alternately, we can have a DOM of the current page in a new window and call FLB asynchronously from regression f/w.

 

Please let me know your views on the workaround and any possibilities of getting DOM of the current page using any tool (not limiting to selenium).

 

Thanks

Madhusudan G