Bug 108362 - Add a concept of dashboard parameters that invalidate others
Summary: Add a concept of dashboard parameters that invalidate others
Status: RESOLVED FIXED
Alias: None
Product: WebKit
Classification: Unclassified
Component: Tools / Tests (show other bugs)
Version: 528+ (Nightly build)
Hardware: Unspecified Unspecified
: P2 Normal
Assignee: Julie Parent
URL:
Keywords:
Depends on:
Blocks:
 
Reported: 2013-01-30 11:45 PST by Julie Parent
Modified: 2013-01-30 18:13 PST (History)
5 users (show)

See Also:


Attachments
Patch (7.05 KB, patch)
2013-01-30 11:49 PST, Julie Parent
no flags Details | Formatted Diff | Diff

Note You need to log in before you can comment on or make changes to this bug.
Description Julie Parent 2013-01-30 11:45:29 PST
Add a concept of dashboard parameters that invalidate others
Comment 1 Julie Parent 2013-01-30 11:49:29 PST
Created attachment 185529 [details]
Patch
Comment 2 Dirk Pranke 2013-01-30 12:49:12 PST
Comment on attachment 185529 [details]
Patch

I'm not sure I understand what it means for one parameter to invalidate another. Does that mean the parameters are mutually exclusive? I take it that, for example, if someone specifies testType=, you want group= to be reset to the default? So you can't actually specify both meaningfully? Is that the same thing as saying that one parameter overrides another?
Comment 3 Julie Parent 2013-01-30 13:24:01 PST
There are parameters that depend on each other.  So, for example, the list of builders available depends on the group.

Imagine I have the flakiness dashboard loaded with the group set to ToT - chromium.org and the builder set to WebKit Linux: http://test-results.appspot.com/dashboards/flakiness_dashboard.html#group=%40ToT%20-%20chromium.org&builder=WebKit%20Linux

Then, I switch the group to ToT webkit.org.  Webkit Linux is no longer a valid builder for that configuration.  Currently, the dashboard will select a new default but the query string keeps the old one, so now there is a mismatch between what is selected and what is in the url (and is incorrect in other parts of the code too).  Sometimes js errors hit due to these mismatches as well.
Comment 4 Dirk Pranke 2013-01-30 17:47:46 PST
Comment on attachment 185529 [details]
Patch

Okay. I understand what you're getting at, I think. I think the naming is a bit confusing, but I don't have any better ideas.
Comment 5 WebKit Review Bot 2013-01-30 18:13:19 PST
Comment on attachment 185529 [details]
Patch

Clearing flags on attachment: 185529

Committed r141355: <http://trac.webkit.org/changeset/141355>
Comment 6 WebKit Review Bot 2013-01-30 18:13:23 PST
All reviewed patches have been landed.  Closing bug.