Bug 31373

Summary:

rewrite fast/dom/Window/window-screen-properties to be slightly more portable

Product:

WebKit

Reporter:

Dirk Pranke <dpranke>

Component:

Tools / Tests

Assignee:

Dirk Pranke <dpranke>

Status:

RESOLVED FIXED

Severity:

Normal

CC:

commit-queue, dimich

Priority:

Version:

528+ (Nightly build)

Hardware:

OS:

OS X 10.5

Attachments:

Description	Flags
proposed rewrite of patch	dimich: review-
revised patch - corrects typo in test case, drops mistakenly included pngs	dimich: review-
drop unnecessary win expectation	none

Dirk Pranke

Reported 2009-11-11 14:07:07 PST

Right now this test will fail unless your screen is set to a single fixed colordepth (24 on the Mac). Even if we have platform-specific overrides, it means everyone on that platform has to have the same colordepth, which at least for Chromium's infrastructure isn't practical. I'd like to rewrite this to accept either 16, 24, or 32 as values. That seems like it's enough to ensure the property isn't returning garbage, but is still pretty flexible.

Attachments
proposed rewrite of patch (167.77 KB, patch) 2009-11-11 14:09 PST, Dirk Pranke	dimich: review-	Details Formatted Diff Diff
revised patch - corrects typo in test case, drops mistakenly included pngs (2.47 KB, patch) 2009-11-11 15:40 PST, Dirk Pranke	dimich: review-	Details Formatted Diff Diff
drop unnecessary win expectation (3.05 KB, patch) 2009-11-11 17:22 PST, Dirk Pranke	no flags	Details Formatted Diff Diff
Show Obsolete (2) View All Add attachment proposed patch, testcase, etc.

Dirk Pranke

Comment 1 2009-11-11 14:09:15 PST

Created attachment 43000 [details] proposed rewrite of patch

Dmitry Titov

Comment 2 2009-11-11 14:57:02 PST

Comment on attachment 43000 [details] proposed rewrite of patch The patch seems to contain unrelated png files. Also, while you are at it, does it make sense to verify that pixelDepth==colorDepth or is it not actually always the case? r- because of extra png files

Dirk Pranke

Comment 3 2009-11-11 15:38:46 PST

grr. Not only were there pngs that snuck in (I'm not sure how), there was a typo in the test file. per http://www.quirksmode.org/dom/w3c_cssom.html , there may be differences between pixeldepth and colordepth on older unix machines (where bpp <= 8). I don't know if we even support bpp <= 8, so I could certainly add that check if we wanted.

Dirk Pranke

Comment 4 2009-11-11 15:40:48 PST

Created attachment 43013 [details] revised patch - corrects typo in test case, drops mistakenly included pngs

Dmitry Titov

Comment 5 2009-11-11 16:35:47 PST

Comment on attachment 43013 [details] revised patch - corrects typo in test case, drops mistakenly included pngs Almost there. I think you also need to update platform/win/fast/dom/Window/window-screen-properties-expected.txt

Dmitry Titov

Comment 6 2009-11-11 16:39:43 PST

Err, not to update but rather delete since your updated test removes a need for a platform-specific snapshot.

Dirk Pranke

Comment 7 2009-11-11 17:22:22 PST

Created attachment 43021 [details] drop unnecessary win expectation Good catch, actually we should be able to delete it.

WebKit Commit Bot

Comment 8 2009-11-11 19:02:23 PST

Comment on attachment 43021 [details] drop unnecessary win expectation Clearing flags on attachment: 43021 Committed r50857: <http://trac.webkit.org/changeset/50857>

WebKit Commit Bot

Comment 9 2009-11-11 19:02:31 PST

All reviewed patches have been landed. Closing bug.

Note You need to log in before you can comment on or make changes to this bug.