Bug 37426 - new-run-webkit-tests should use the ServerProcess abstraction in chromium.py
Summary: new-run-webkit-tests should use the ServerProcess abstraction in chromium.py
Status: RESOLVED WONTFIX
Alias: None
Product: WebKit
Classification: Unclassified
Component: Tools / Tests (show other bugs)
Version: 528+ (Nightly build)
Hardware: PC OS X 10.5
: P2 Normal
Assignee: Nobody
URL:
Keywords:
Depends on:
Blocks:
 
Reported: 2010-04-11 18:45 PDT by Dirk Pranke
Modified: 2011-05-31 18:54 PDT (History)
3 users (show)

See Also:


Attachments

Note You need to log in before you can comment on or make changes to this bug.
Description Dirk Pranke 2010-04-11 18:45:27 PDT
The code to launch and read from test_shell in chromium.py predates the ServerProcess abstraction that WebKit Mac uses. It should be rewritten to use it to minimize code duplication.
Comment 1 Dirk Pranke 2010-11-11 20:00:54 PST
Tony, I'm punting this over to you, because it becomes true when we remove the test_shell code and only use DRT.
Comment 2 Tony Chang 2010-11-12 10:05:14 PST
Ok, after we switch, I'll see about removing the test_shell related code.
Comment 3 Tony Chang 2011-02-04 16:34:59 PST
Hmm, as mentioned in bug 34826 comment 20, it is non-trivial to make use of ServerProcess on Windows python.  I'm not sure if this is something we want to pursue.
Comment 4 Dirk Pranke 2011-03-18 18:46:53 PDT
punting back to me ;)
Comment 5 Dirk Pranke 2011-04-06 19:44:43 PDT
I don't think it should block bug 34984 and I plan to clear the blocking flag on Mon 4/11. If you disagree, now would be a good time to speak up :).
Comment 6 Dirk Pranke 2011-04-12 18:05:31 PDT
clearing blocking flag.
Comment 7 Dirk Pranke 2011-04-15 17:11:49 PDT
See bug 37426 for a way to make serverprocess work w/o blocking on win32.
Comment 8 Dirk Pranke 2011-05-31 18:54:00 PDT
A more accurate description of this would probably be that we should just use the WebKitDriver implementation if we ever want to switch the Chromium DRT to webkit-style output.

Whether or not the WebKitDriver implementation works on windows is a separate issue that should probably be tracked under a separate bug.

I'm marking this as WONTFIX for now and clearing ownership; if anyone disagrees, feel free to reopen it.