History | Log In     View a printable version of the current page.  
Issue Details (XML | Word | Printable)

Key: WET-171
Type: Improvement Improvement
Status: Closed Closed
Resolution: Fixed
Priority: Critical Critical
Assignee: Unassigned
Reporter: Raghu Venkataramana
Votes: 0
Watchers: 0
Operations

If you were logged in you would be able to see more operations.
WET

Create the Browser DOM Wrapper directly from C#

Created: 16/Aug/06 05:31 AM   Updated: 18/Apr/07 11:11 PM
Component/s: None
Affects Version/s: 0.9.8_beta1
Fix Version/s: 1.0.0

Original Estimate: Unknown Remaining Estimate: Unknown Time Spent: Unknown
Issue Links:
Relationship
 
This issue Relates to:
WET-263 Getting the exception saying that RPC... Major Closed


 Description  « Hide
Right now, this step is done by communicating with the ruby layer - C# sends an 'os process command' to WET core (ruby). WET in turn writes back a DOM XML to give a representation of the elements. This techinique not only opens up opportunity for a lot of bugs, but also makes the entire refresh process very slow.

 All   Comments   Work Log   Change History      Sort Order:
Raghu Venkataramana - 21/Aug/06 01:03 PM
Actually, there are many more issues in doing this. We may end up having to rewrite the entire WET core layer in C# to accomplish this. Reason: The current way of accessing the DOM by the WET Core layer uses a lot of its internal DOM construction technique. And WET's DOM Construction methodology is pretty exhaustive and we should try to reuse it rather than reinvent.

Raghu Venkataramana - 02/Sep/06 03:02 PM
Getting bitten by this once again. It appears that we are better off fixing this rather than being hit by a much bigger problem later on. It probably implies some additional effort but it may well be worth it.

Raghu Venkataramana - 03/Sep/06 12:33 PM
As of now, I did find another workaround for the problem that I mentioned yesterday. Yes, I agree it is very important that we put a permanant fix for this (that is, resolve this issue). However given the fact that there is a workaround, this can probably wait for a bit more.

Raghu Venkataramana - 02/Oct/06 04:49 AM
I keep going back and forth on whether or not to implement this. If we do implement this, we are likely to cause big time regressions. Since most of the methods for the dom traversal in ruby use the built in eval method, it may not be appropriate to reimplement this. Instead, try to stick to the current implementation and keep improvising it when we run into new issues.

Raghu Venkataramana - 23/Jan/07 03:02 AM
The jury is finally out. This HAS to be done and HAS to be done now. In fact for the past month.5+ I have been working a lot towards this. At this stage I have completed implementation for most of the objects. What remains is the unit tests for each of these.

Raghu Venkataramana - 20/Feb/07 09:11 PM
The RPC problem seems to be getting hotter. It is important that we fix this before 1.0

Raghu Venkataramana - 23/Mar/07 03:32 AM
We are closer to getting this completed. Shruthi J is working hard towards this. Most of the tests are complete now. What remains is the tests for checking if definition is created correctly

Raghu Venkataramana - 02/Apr/07 10:07 PM
I was wrong in my previous comment (about being close to getting this completed). We are now working on creating a browser dom wrapper for the simulated browser and this seems to be time consuming;

Raghu Venkataramana - 18/Apr/07 11:11 PM
The implementation of this has been completed. We should now start opening separate issues for specific problems.